{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 2406,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0012476606363069245,
      "grad_norm": 8.761757813219194,
      "learning_rate": 0.0,
      "loss": 2.2506,
      "num_tokens": 76455.0,
      "step": 1
    },
    {
      "epoch": 0.002495321272613849,
      "grad_norm": 8.78949353381417,
      "learning_rate": 1.36986301369863e-07,
      "loss": 2.2588,
      "num_tokens": 152591.0,
      "step": 2
    },
    {
      "epoch": 0.0037429819089207735,
      "grad_norm": 8.761353805328225,
      "learning_rate": 2.73972602739726e-07,
      "loss": 2.2518,
      "num_tokens": 228968.0,
      "step": 3
    },
    {
      "epoch": 0.004990642545227698,
      "grad_norm": 8.680409609244958,
      "learning_rate": 4.1095890410958903e-07,
      "loss": 2.2367,
      "num_tokens": 306355.0,
      "step": 4
    },
    {
      "epoch": 0.006238303181534623,
      "grad_norm": 8.8251922200434,
      "learning_rate": 5.47945205479452e-07,
      "loss": 2.2511,
      "num_tokens": 382390.0,
      "step": 5
    },
    {
      "epoch": 0.007485963817841547,
      "grad_norm": 8.678437336777996,
      "learning_rate": 6.849315068493151e-07,
      "loss": 2.2412,
      "num_tokens": 459014.0,
      "step": 6
    },
    {
      "epoch": 0.008733624454148471,
      "grad_norm": 8.63321084471758,
      "learning_rate": 8.219178082191781e-07,
      "loss": 2.2448,
      "num_tokens": 534258.0,
      "step": 7
    },
    {
      "epoch": 0.009981285090455396,
      "grad_norm": 8.511625206608334,
      "learning_rate": 9.589041095890411e-07,
      "loss": 2.2307,
      "num_tokens": 610291.0,
      "step": 8
    },
    {
      "epoch": 0.011228945726762321,
      "grad_norm": 8.300645950958717,
      "learning_rate": 1.095890410958904e-06,
      "loss": 2.1933,
      "num_tokens": 685797.0,
      "step": 9
    },
    {
      "epoch": 0.012476606363069246,
      "grad_norm": 7.8420598537636526,
      "learning_rate": 1.2328767123287673e-06,
      "loss": 2.1148,
      "num_tokens": 762054.0,
      "step": 10
    },
    {
      "epoch": 0.01372426699937617,
      "grad_norm": 7.620747695514949,
      "learning_rate": 1.3698630136986302e-06,
      "loss": 2.0744,
      "num_tokens": 839526.0,
      "step": 11
    },
    {
      "epoch": 0.014971927635683094,
      "grad_norm": 7.7311728817507674,
      "learning_rate": 1.5068493150684932e-06,
      "loss": 2.0946,
      "num_tokens": 915416.0,
      "step": 12
    },
    {
      "epoch": 0.016219588271990017,
      "grad_norm": 6.496044071412181,
      "learning_rate": 1.6438356164383561e-06,
      "loss": 1.8218,
      "num_tokens": 992432.0,
      "step": 13
    },
    {
      "epoch": 0.017467248908296942,
      "grad_norm": 6.466236172125798,
      "learning_rate": 1.7808219178082193e-06,
      "loss": 1.8035,
      "num_tokens": 1069191.0,
      "step": 14
    },
    {
      "epoch": 0.018714909544603867,
      "grad_norm": 6.272721593409537,
      "learning_rate": 1.9178082191780823e-06,
      "loss": 1.7429,
      "num_tokens": 1146441.0,
      "step": 15
    },
    {
      "epoch": 0.019962570180910792,
      "grad_norm": 6.294095568991284,
      "learning_rate": 2.0547945205479454e-06,
      "loss": 1.7267,
      "num_tokens": 1222771.0,
      "step": 16
    },
    {
      "epoch": 0.021210230817217717,
      "grad_norm": 6.86423665778992,
      "learning_rate": 2.191780821917808e-06,
      "loss": 1.1876,
      "num_tokens": 1298674.0,
      "step": 17
    },
    {
      "epoch": 0.022457891453524642,
      "grad_norm": 6.675922637460064,
      "learning_rate": 2.3287671232876713e-06,
      "loss": 1.1796,
      "num_tokens": 1374228.0,
      "step": 18
    },
    {
      "epoch": 0.023705552089831567,
      "grad_norm": 6.923846448363932,
      "learning_rate": 2.4657534246575345e-06,
      "loss": 1.1247,
      "num_tokens": 1451793.0,
      "step": 19
    },
    {
      "epoch": 0.024953212726138492,
      "grad_norm": 9.498586320399488,
      "learning_rate": 2.6027397260273973e-06,
      "loss": 0.9696,
      "num_tokens": 1528307.0,
      "step": 20
    },
    {
      "epoch": 0.026200873362445413,
      "grad_norm": 6.642114493263439,
      "learning_rate": 2.7397260273972604e-06,
      "loss": 0.8689,
      "num_tokens": 1604991.0,
      "step": 21
    },
    {
      "epoch": 0.02744853399875234,
      "grad_norm": 10.71018918516391,
      "learning_rate": 2.876712328767123e-06,
      "loss": 0.7623,
      "num_tokens": 1681870.0,
      "step": 22
    },
    {
      "epoch": 0.028696194635059263,
      "grad_norm": 5.930114665023217,
      "learning_rate": 3.0136986301369864e-06,
      "loss": 0.3494,
      "num_tokens": 1758838.0,
      "step": 23
    },
    {
      "epoch": 0.02994385527136619,
      "grad_norm": 2.883944763296146,
      "learning_rate": 3.1506849315068495e-06,
      "loss": 0.2483,
      "num_tokens": 1834793.0,
      "step": 24
    },
    {
      "epoch": 0.031191515907673113,
      "grad_norm": 2.6374171238506747,
      "learning_rate": 3.2876712328767123e-06,
      "loss": 0.2201,
      "num_tokens": 1911654.0,
      "step": 25
    },
    {
      "epoch": 0.032439176543980035,
      "grad_norm": 0.8688472783801062,
      "learning_rate": 3.4246575342465754e-06,
      "loss": 0.1953,
      "num_tokens": 1987054.0,
      "step": 26
    },
    {
      "epoch": 0.03368683718028696,
      "grad_norm": 0.6726353699266373,
      "learning_rate": 3.5616438356164386e-06,
      "loss": 0.1991,
      "num_tokens": 2063735.0,
      "step": 27
    },
    {
      "epoch": 0.034934497816593885,
      "grad_norm": 0.6879798139762083,
      "learning_rate": 3.6986301369863014e-06,
      "loss": 0.1865,
      "num_tokens": 2140736.0,
      "step": 28
    },
    {
      "epoch": 0.03618215845290081,
      "grad_norm": 0.5328446424800377,
      "learning_rate": 3.8356164383561645e-06,
      "loss": 0.1794,
      "num_tokens": 2215808.0,
      "step": 29
    },
    {
      "epoch": 0.037429819089207735,
      "grad_norm": 0.49572408846234306,
      "learning_rate": 3.972602739726027e-06,
      "loss": 0.1722,
      "num_tokens": 2291680.0,
      "step": 30
    },
    {
      "epoch": 0.03867747972551466,
      "grad_norm": 0.5123370045083006,
      "learning_rate": 4.109589041095891e-06,
      "loss": 0.1731,
      "num_tokens": 2367158.0,
      "step": 31
    },
    {
      "epoch": 0.039925140361821584,
      "grad_norm": 0.5664689020425665,
      "learning_rate": 4.246575342465754e-06,
      "loss": 0.1663,
      "num_tokens": 2442906.0,
      "step": 32
    },
    {
      "epoch": 0.041172800998128506,
      "grad_norm": 0.5299695266663934,
      "learning_rate": 4.383561643835616e-06,
      "loss": 0.1661,
      "num_tokens": 2518923.0,
      "step": 33
    },
    {
      "epoch": 0.042420461634435434,
      "grad_norm": 0.4311694293651315,
      "learning_rate": 4.52054794520548e-06,
      "loss": 0.1576,
      "num_tokens": 2594906.0,
      "step": 34
    },
    {
      "epoch": 0.043668122270742356,
      "grad_norm": 0.37623197449723156,
      "learning_rate": 4.657534246575343e-06,
      "loss": 0.1464,
      "num_tokens": 2670975.0,
      "step": 35
    },
    {
      "epoch": 0.044915782907049284,
      "grad_norm": 0.3043781819858105,
      "learning_rate": 4.7945205479452054e-06,
      "loss": 0.1484,
      "num_tokens": 2746676.0,
      "step": 36
    },
    {
      "epoch": 0.046163443543356206,
      "grad_norm": 0.24996202751880106,
      "learning_rate": 4.931506849315069e-06,
      "loss": 0.1467,
      "num_tokens": 2822529.0,
      "step": 37
    },
    {
      "epoch": 0.047411104179663134,
      "grad_norm": 0.2502460222418088,
      "learning_rate": 5.068493150684932e-06,
      "loss": 0.1531,
      "num_tokens": 2899091.0,
      "step": 38
    },
    {
      "epoch": 0.048658764815970056,
      "grad_norm": 0.21114797440871622,
      "learning_rate": 5.2054794520547945e-06,
      "loss": 0.1421,
      "num_tokens": 2975561.0,
      "step": 39
    },
    {
      "epoch": 0.049906425452276984,
      "grad_norm": 0.21983583172698634,
      "learning_rate": 5.342465753424658e-06,
      "loss": 0.1353,
      "num_tokens": 3051729.0,
      "step": 40
    },
    {
      "epoch": 0.051154086088583905,
      "grad_norm": 0.20644663877144454,
      "learning_rate": 5.479452054794521e-06,
      "loss": 0.1373,
      "num_tokens": 3127617.0,
      "step": 41
    },
    {
      "epoch": 0.05240174672489083,
      "grad_norm": 0.24835560615668648,
      "learning_rate": 5.6164383561643845e-06,
      "loss": 0.141,
      "num_tokens": 3204875.0,
      "step": 42
    },
    {
      "epoch": 0.053649407361197755,
      "grad_norm": 0.21591056284990343,
      "learning_rate": 5.753424657534246e-06,
      "loss": 0.1341,
      "num_tokens": 3281650.0,
      "step": 43
    },
    {
      "epoch": 0.05489706799750468,
      "grad_norm": 0.23616198046047548,
      "learning_rate": 5.89041095890411e-06,
      "loss": 0.1181,
      "num_tokens": 3356216.0,
      "step": 44
    },
    {
      "epoch": 0.056144728633811605,
      "grad_norm": 0.22200353016141697,
      "learning_rate": 6.027397260273973e-06,
      "loss": 0.139,
      "num_tokens": 3433001.0,
      "step": 45
    },
    {
      "epoch": 0.05739238927011853,
      "grad_norm": 0.25871574428667854,
      "learning_rate": 6.164383561643836e-06,
      "loss": 0.1275,
      "num_tokens": 3508082.0,
      "step": 46
    },
    {
      "epoch": 0.058640049906425455,
      "grad_norm": 0.21842107713575873,
      "learning_rate": 6.301369863013699e-06,
      "loss": 0.1265,
      "num_tokens": 3583716.0,
      "step": 47
    },
    {
      "epoch": 0.05988771054273238,
      "grad_norm": 0.24004259697156158,
      "learning_rate": 6.438356164383563e-06,
      "loss": 0.1242,
      "num_tokens": 3659753.0,
      "step": 48
    },
    {
      "epoch": 0.0611353711790393,
      "grad_norm": 0.2116092456996917,
      "learning_rate": 6.5753424657534245e-06,
      "loss": 0.1328,
      "num_tokens": 3736035.0,
      "step": 49
    },
    {
      "epoch": 0.06238303181534623,
      "grad_norm": 0.2013190481716377,
      "learning_rate": 6.712328767123288e-06,
      "loss": 0.1179,
      "num_tokens": 3811794.0,
      "step": 50
    },
    {
      "epoch": 0.06363069245165315,
      "grad_norm": 0.22264594627704687,
      "learning_rate": 6.849315068493151e-06,
      "loss": 0.1243,
      "num_tokens": 3888185.0,
      "step": 51
    },
    {
      "epoch": 0.06487835308796007,
      "grad_norm": 0.18870977013081056,
      "learning_rate": 6.9863013698630145e-06,
      "loss": 0.1216,
      "num_tokens": 3964653.0,
      "step": 52
    },
    {
      "epoch": 0.066126013724267,
      "grad_norm": 0.21848947971984553,
      "learning_rate": 7.123287671232877e-06,
      "loss": 0.1187,
      "num_tokens": 4040648.0,
      "step": 53
    },
    {
      "epoch": 0.06737367436057393,
      "grad_norm": 0.2103776859815552,
      "learning_rate": 7.260273972602741e-06,
      "loss": 0.1131,
      "num_tokens": 4117805.0,
      "step": 54
    },
    {
      "epoch": 0.06862133499688085,
      "grad_norm": 0.2005737766635463,
      "learning_rate": 7.397260273972603e-06,
      "loss": 0.1215,
      "num_tokens": 4194859.0,
      "step": 55
    },
    {
      "epoch": 0.06986899563318777,
      "grad_norm": 0.1956940878293591,
      "learning_rate": 7.534246575342466e-06,
      "loss": 0.111,
      "num_tokens": 4270378.0,
      "step": 56
    },
    {
      "epoch": 0.07111665626949469,
      "grad_norm": 0.19607223925616332,
      "learning_rate": 7.671232876712329e-06,
      "loss": 0.119,
      "num_tokens": 4346106.0,
      "step": 57
    },
    {
      "epoch": 0.07236431690580163,
      "grad_norm": 0.19170667845038655,
      "learning_rate": 7.808219178082192e-06,
      "loss": 0.1116,
      "num_tokens": 4421749.0,
      "step": 58
    },
    {
      "epoch": 0.07361197754210855,
      "grad_norm": 0.20616149253606764,
      "learning_rate": 7.945205479452055e-06,
      "loss": 0.1178,
      "num_tokens": 4499385.0,
      "step": 59
    },
    {
      "epoch": 0.07485963817841547,
      "grad_norm": 0.19654472797463204,
      "learning_rate": 8.082191780821919e-06,
      "loss": 0.1082,
      "num_tokens": 4575336.0,
      "step": 60
    },
    {
      "epoch": 0.07610729881472239,
      "grad_norm": 0.2112216421698123,
      "learning_rate": 8.219178082191782e-06,
      "loss": 0.1086,
      "num_tokens": 4650506.0,
      "step": 61
    },
    {
      "epoch": 0.07735495945102933,
      "grad_norm": 0.2076293442211523,
      "learning_rate": 8.356164383561644e-06,
      "loss": 0.1151,
      "num_tokens": 4726709.0,
      "step": 62
    },
    {
      "epoch": 0.07860262008733625,
      "grad_norm": 0.19682970614365813,
      "learning_rate": 8.493150684931507e-06,
      "loss": 0.1138,
      "num_tokens": 4802335.0,
      "step": 63
    },
    {
      "epoch": 0.07985028072364317,
      "grad_norm": 0.22630532906454626,
      "learning_rate": 8.63013698630137e-06,
      "loss": 0.1134,
      "num_tokens": 4879330.0,
      "step": 64
    },
    {
      "epoch": 0.08109794135995009,
      "grad_norm": 0.20005557766025198,
      "learning_rate": 8.767123287671233e-06,
      "loss": 0.1109,
      "num_tokens": 4956574.0,
      "step": 65
    },
    {
      "epoch": 0.08234560199625701,
      "grad_norm": 0.18919889491059635,
      "learning_rate": 8.904109589041097e-06,
      "loss": 0.1022,
      "num_tokens": 5032651.0,
      "step": 66
    },
    {
      "epoch": 0.08359326263256395,
      "grad_norm": 0.2220403365811989,
      "learning_rate": 9.04109589041096e-06,
      "loss": 0.1165,
      "num_tokens": 5110439.0,
      "step": 67
    },
    {
      "epoch": 0.08484092326887087,
      "grad_norm": 0.22886272632277932,
      "learning_rate": 9.178082191780823e-06,
      "loss": 0.0964,
      "num_tokens": 5186001.0,
      "step": 68
    },
    {
      "epoch": 0.08608858390517779,
      "grad_norm": 0.19831357939942088,
      "learning_rate": 9.315068493150685e-06,
      "loss": 0.105,
      "num_tokens": 5261891.0,
      "step": 69
    },
    {
      "epoch": 0.08733624454148471,
      "grad_norm": 0.22099512786602915,
      "learning_rate": 9.452054794520548e-06,
      "loss": 0.1008,
      "num_tokens": 5337002.0,
      "step": 70
    },
    {
      "epoch": 0.08858390517779165,
      "grad_norm": 0.18641406929882653,
      "learning_rate": 9.589041095890411e-06,
      "loss": 0.1052,
      "num_tokens": 5413796.0,
      "step": 71
    },
    {
      "epoch": 0.08983156581409857,
      "grad_norm": 0.19203725614575334,
      "learning_rate": 9.726027397260275e-06,
      "loss": 0.1012,
      "num_tokens": 5489737.0,
      "step": 72
    },
    {
      "epoch": 0.09107922645040549,
      "grad_norm": 0.20246710152992942,
      "learning_rate": 9.863013698630138e-06,
      "loss": 0.1048,
      "num_tokens": 5564974.0,
      "step": 73
    },
    {
      "epoch": 0.09232688708671241,
      "grad_norm": 0.1876916336068445,
      "learning_rate": 1e-05,
      "loss": 0.1013,
      "num_tokens": 5641815.0,
      "step": 74
    },
    {
      "epoch": 0.09357454772301933,
      "grad_norm": 0.21390292979614334,
      "learning_rate": 9.999995920069922e-06,
      "loss": 0.1029,
      "num_tokens": 5718359.0,
      "step": 75
    },
    {
      "epoch": 0.09482220835932627,
      "grad_norm": 0.2017756259599263,
      "learning_rate": 9.999983680287084e-06,
      "loss": 0.1063,
      "num_tokens": 5794874.0,
      "step": 76
    },
    {
      "epoch": 0.09606986899563319,
      "grad_norm": 0.18714097159366816,
      "learning_rate": 9.99996328067368e-06,
      "loss": 0.0957,
      "num_tokens": 5870956.0,
      "step": 77
    },
    {
      "epoch": 0.09731752963194011,
      "grad_norm": 0.18505126145277956,
      "learning_rate": 9.999934721266702e-06,
      "loss": 0.0908,
      "num_tokens": 5946585.0,
      "step": 78
    },
    {
      "epoch": 0.09856519026824703,
      "grad_norm": 0.19156875617520439,
      "learning_rate": 9.999898002117937e-06,
      "loss": 0.095,
      "num_tokens": 6021741.0,
      "step": 79
    },
    {
      "epoch": 0.09981285090455397,
      "grad_norm": 0.1929723855918604,
      "learning_rate": 9.999853123293967e-06,
      "loss": 0.1011,
      "num_tokens": 6097082.0,
      "step": 80
    },
    {
      "epoch": 0.10106051154086089,
      "grad_norm": 0.2001901241869327,
      "learning_rate": 9.99980008487617e-06,
      "loss": 0.1006,
      "num_tokens": 6173503.0,
      "step": 81
    },
    {
      "epoch": 0.10230817217716781,
      "grad_norm": 0.18306362250643257,
      "learning_rate": 9.999738886960724e-06,
      "loss": 0.0983,
      "num_tokens": 6251913.0,
      "step": 82
    },
    {
      "epoch": 0.10355583281347473,
      "grad_norm": 0.18878076652513084,
      "learning_rate": 9.999669529658596e-06,
      "loss": 0.0972,
      "num_tokens": 6327736.0,
      "step": 83
    },
    {
      "epoch": 0.10480349344978165,
      "grad_norm": 0.17758975161938698,
      "learning_rate": 9.999592013095553e-06,
      "loss": 0.0919,
      "num_tokens": 6403683.0,
      "step": 84
    },
    {
      "epoch": 0.10605115408608859,
      "grad_norm": 0.18599288991127097,
      "learning_rate": 9.999506337412157e-06,
      "loss": 0.0952,
      "num_tokens": 6479398.0,
      "step": 85
    },
    {
      "epoch": 0.10729881472239551,
      "grad_norm": 0.18920269409828955,
      "learning_rate": 9.99941250276376e-06,
      "loss": 0.0939,
      "num_tokens": 6555703.0,
      "step": 86
    },
    {
      "epoch": 0.10854647535870243,
      "grad_norm": 0.16667220843732977,
      "learning_rate": 9.999310509320518e-06,
      "loss": 0.0845,
      "num_tokens": 6631800.0,
      "step": 87
    },
    {
      "epoch": 0.10979413599500935,
      "grad_norm": 0.19074803832408102,
      "learning_rate": 9.999200357267373e-06,
      "loss": 0.0968,
      "num_tokens": 6708923.0,
      "step": 88
    },
    {
      "epoch": 0.11104179663131628,
      "grad_norm": 0.17189646671230804,
      "learning_rate": 9.999082046804062e-06,
      "loss": 0.0908,
      "num_tokens": 6784265.0,
      "step": 89
    },
    {
      "epoch": 0.11228945726762321,
      "grad_norm": 0.1887585644537061,
      "learning_rate": 9.998955578145124e-06,
      "loss": 0.1033,
      "num_tokens": 6861631.0,
      "step": 90
    },
    {
      "epoch": 0.11353711790393013,
      "grad_norm": 0.19093170392495093,
      "learning_rate": 9.998820951519877e-06,
      "loss": 0.0957,
      "num_tokens": 6937782.0,
      "step": 91
    },
    {
      "epoch": 0.11478477854023705,
      "grad_norm": 0.18068911481247474,
      "learning_rate": 9.998678167172446e-06,
      "loss": 0.0916,
      "num_tokens": 7013064.0,
      "step": 92
    },
    {
      "epoch": 0.11603243917654397,
      "grad_norm": 0.1807867354947808,
      "learning_rate": 9.99852722536174e-06,
      "loss": 0.0912,
      "num_tokens": 7089477.0,
      "step": 93
    },
    {
      "epoch": 0.11728009981285091,
      "grad_norm": 0.19214351630835524,
      "learning_rate": 9.998368126361459e-06,
      "loss": 0.089,
      "num_tokens": 7165616.0,
      "step": 94
    },
    {
      "epoch": 0.11852776044915783,
      "grad_norm": 0.1703054533344699,
      "learning_rate": 9.998200870460103e-06,
      "loss": 0.0894,
      "num_tokens": 7241528.0,
      "step": 95
    },
    {
      "epoch": 0.11977542108546475,
      "grad_norm": 0.19018854305829216,
      "learning_rate": 9.998025457960955e-06,
      "loss": 0.0907,
      "num_tokens": 7317982.0,
      "step": 96
    },
    {
      "epoch": 0.12102308172177167,
      "grad_norm": 0.18133063854407905,
      "learning_rate": 9.997841889182091e-06,
      "loss": 0.0908,
      "num_tokens": 7394173.0,
      "step": 97
    },
    {
      "epoch": 0.1222707423580786,
      "grad_norm": 0.20643507465810595,
      "learning_rate": 9.997650164456375e-06,
      "loss": 0.0959,
      "num_tokens": 7471763.0,
      "step": 98
    },
    {
      "epoch": 0.12351840299438553,
      "grad_norm": 0.19177798797637297,
      "learning_rate": 9.997450284131465e-06,
      "loss": 0.0859,
      "num_tokens": 7547218.0,
      "step": 99
    },
    {
      "epoch": 0.12476606363069245,
      "grad_norm": 0.18410460168165857,
      "learning_rate": 9.997242248569802e-06,
      "loss": 0.0929,
      "num_tokens": 7623208.0,
      "step": 100
    },
    {
      "epoch": 0.1260137242669994,
      "grad_norm": 0.17296669023164052,
      "learning_rate": 9.997026058148617e-06,
      "loss": 0.0867,
      "num_tokens": 7700131.0,
      "step": 101
    },
    {
      "epoch": 0.1272613849033063,
      "grad_norm": 0.19233260273716293,
      "learning_rate": 9.996801713259933e-06,
      "loss": 0.0911,
      "num_tokens": 7775303.0,
      "step": 102
    },
    {
      "epoch": 0.12850904553961323,
      "grad_norm": 0.1796254555341303,
      "learning_rate": 9.996569214310549e-06,
      "loss": 0.0842,
      "num_tokens": 7850968.0,
      "step": 103
    },
    {
      "epoch": 0.12975670617592014,
      "grad_norm": 0.17463731855234868,
      "learning_rate": 9.99632856172206e-06,
      "loss": 0.09,
      "num_tokens": 7928653.0,
      "step": 104
    },
    {
      "epoch": 0.13100436681222707,
      "grad_norm": 0.18296072966873897,
      "learning_rate": 9.99607975593084e-06,
      "loss": 0.093,
      "num_tokens": 8004382.0,
      "step": 105
    },
    {
      "epoch": 0.132252027448534,
      "grad_norm": 0.17517424828901534,
      "learning_rate": 9.995822797388052e-06,
      "loss": 0.0884,
      "num_tokens": 8080346.0,
      "step": 106
    },
    {
      "epoch": 0.13349968808484092,
      "grad_norm": 0.18374538686707345,
      "learning_rate": 9.995557686559635e-06,
      "loss": 0.0902,
      "num_tokens": 8157642.0,
      "step": 107
    },
    {
      "epoch": 0.13474734872114785,
      "grad_norm": 0.17640658082095845,
      "learning_rate": 9.995284423926318e-06,
      "loss": 0.0835,
      "num_tokens": 8233301.0,
      "step": 108
    },
    {
      "epoch": 0.13599500935745476,
      "grad_norm": 0.1778279324200043,
      "learning_rate": 9.995003009983608e-06,
      "loss": 0.0871,
      "num_tokens": 8308683.0,
      "step": 109
    },
    {
      "epoch": 0.1372426699937617,
      "grad_norm": 0.17643769063186895,
      "learning_rate": 9.994713445241793e-06,
      "loss": 0.0859,
      "num_tokens": 8384876.0,
      "step": 110
    },
    {
      "epoch": 0.13849033063006863,
      "grad_norm": 0.1713769310660126,
      "learning_rate": 9.994415730225943e-06,
      "loss": 0.085,
      "num_tokens": 8461189.0,
      "step": 111
    },
    {
      "epoch": 0.13973799126637554,
      "grad_norm": 0.16535183080331606,
      "learning_rate": 9.994109865475903e-06,
      "loss": 0.087,
      "num_tokens": 8536969.0,
      "step": 112
    },
    {
      "epoch": 0.14098565190268247,
      "grad_norm": 0.1713983338918515,
      "learning_rate": 9.993795851546302e-06,
      "loss": 0.0893,
      "num_tokens": 8613651.0,
      "step": 113
    },
    {
      "epoch": 0.14223331253898938,
      "grad_norm": 0.1797062763194263,
      "learning_rate": 9.993473689006538e-06,
      "loss": 0.0877,
      "num_tokens": 8689954.0,
      "step": 114
    },
    {
      "epoch": 0.14348097317529632,
      "grad_norm": 0.18257913976658624,
      "learning_rate": 9.99314337844079e-06,
      "loss": 0.0918,
      "num_tokens": 8765928.0,
      "step": 115
    },
    {
      "epoch": 0.14472863381160325,
      "grad_norm": 0.17867571616347352,
      "learning_rate": 9.992804920448013e-06,
      "loss": 0.0826,
      "num_tokens": 8841922.0,
      "step": 116
    },
    {
      "epoch": 0.14597629444791016,
      "grad_norm": 0.16778802320918987,
      "learning_rate": 9.992458315641932e-06,
      "loss": 0.0848,
      "num_tokens": 8917607.0,
      "step": 117
    },
    {
      "epoch": 0.1472239550842171,
      "grad_norm": 0.16474267413213903,
      "learning_rate": 9.992103564651048e-06,
      "loss": 0.0813,
      "num_tokens": 8993184.0,
      "step": 118
    },
    {
      "epoch": 0.14847161572052403,
      "grad_norm": 0.1841405830448607,
      "learning_rate": 9.991740668118629e-06,
      "loss": 0.0837,
      "num_tokens": 9069615.0,
      "step": 119
    },
    {
      "epoch": 0.14971927635683094,
      "grad_norm": 0.16992784576092151,
      "learning_rate": 9.991369626702717e-06,
      "loss": 0.0844,
      "num_tokens": 9146612.0,
      "step": 120
    },
    {
      "epoch": 0.15096693699313787,
      "grad_norm": 0.17131842491068652,
      "learning_rate": 9.990990441076125e-06,
      "loss": 0.0818,
      "num_tokens": 9222704.0,
      "step": 121
    },
    {
      "epoch": 0.15221459762944478,
      "grad_norm": 0.17804714059607288,
      "learning_rate": 9.990603111926424e-06,
      "loss": 0.0858,
      "num_tokens": 9298555.0,
      "step": 122
    },
    {
      "epoch": 0.15346225826575172,
      "grad_norm": 0.1729551798123775,
      "learning_rate": 9.990207639955969e-06,
      "loss": 0.0873,
      "num_tokens": 9374922.0,
      "step": 123
    },
    {
      "epoch": 0.15470991890205865,
      "grad_norm": 0.18111977631889967,
      "learning_rate": 9.989804025881862e-06,
      "loss": 0.0839,
      "num_tokens": 9450879.0,
      "step": 124
    },
    {
      "epoch": 0.15595757953836556,
      "grad_norm": 0.17388225325106033,
      "learning_rate": 9.98939227043598e-06,
      "loss": 0.088,
      "num_tokens": 9527736.0,
      "step": 125
    },
    {
      "epoch": 0.1572052401746725,
      "grad_norm": 0.17758998450950486,
      "learning_rate": 9.988972374364961e-06,
      "loss": 0.086,
      "num_tokens": 9603872.0,
      "step": 126
    },
    {
      "epoch": 0.1584529008109794,
      "grad_norm": 0.16662865144228015,
      "learning_rate": 9.988544338430203e-06,
      "loss": 0.0772,
      "num_tokens": 9679807.0,
      "step": 127
    },
    {
      "epoch": 0.15970056144728634,
      "grad_norm": 0.1734848561541322,
      "learning_rate": 9.988108163407865e-06,
      "loss": 0.0857,
      "num_tokens": 9755823.0,
      "step": 128
    },
    {
      "epoch": 0.16094822208359327,
      "grad_norm": 0.1711057297356826,
      "learning_rate": 9.987663850088862e-06,
      "loss": 0.0815,
      "num_tokens": 9832201.0,
      "step": 129
    },
    {
      "epoch": 0.16219588271990018,
      "grad_norm": 0.18609650663984184,
      "learning_rate": 9.987211399278871e-06,
      "loss": 0.0884,
      "num_tokens": 9908854.0,
      "step": 130
    },
    {
      "epoch": 0.16344354335620712,
      "grad_norm": 0.16781225049301668,
      "learning_rate": 9.98675081179832e-06,
      "loss": 0.0847,
      "num_tokens": 9985482.0,
      "step": 131
    },
    {
      "epoch": 0.16469120399251402,
      "grad_norm": 0.1666096665089819,
      "learning_rate": 9.986282088482397e-06,
      "loss": 0.082,
      "num_tokens": 10061705.0,
      "step": 132
    },
    {
      "epoch": 0.16593886462882096,
      "grad_norm": 0.16861260460290664,
      "learning_rate": 9.985805230181031e-06,
      "loss": 0.0775,
      "num_tokens": 10137791.0,
      "step": 133
    },
    {
      "epoch": 0.1671865252651279,
      "grad_norm": 0.16847932298131205,
      "learning_rate": 9.985320237758918e-06,
      "loss": 0.0788,
      "num_tokens": 10213758.0,
      "step": 134
    },
    {
      "epoch": 0.1684341859014348,
      "grad_norm": 0.16847673392353724,
      "learning_rate": 9.984827112095495e-06,
      "loss": 0.0789,
      "num_tokens": 10289192.0,
      "step": 135
    },
    {
      "epoch": 0.16968184653774174,
      "grad_norm": 0.17266317511461257,
      "learning_rate": 9.984325854084946e-06,
      "loss": 0.0792,
      "num_tokens": 10366976.0,
      "step": 136
    },
    {
      "epoch": 0.17092950717404864,
      "grad_norm": 0.17254331344081836,
      "learning_rate": 9.983816464636203e-06,
      "loss": 0.0787,
      "num_tokens": 10442519.0,
      "step": 137
    },
    {
      "epoch": 0.17217716781035558,
      "grad_norm": 0.16483644978279038,
      "learning_rate": 9.983298944672942e-06,
      "loss": 0.0761,
      "num_tokens": 10518729.0,
      "step": 138
    },
    {
      "epoch": 0.17342482844666252,
      "grad_norm": 0.16316748420591484,
      "learning_rate": 9.982773295133585e-06,
      "loss": 0.0749,
      "num_tokens": 10594664.0,
      "step": 139
    },
    {
      "epoch": 0.17467248908296942,
      "grad_norm": 0.17456574782390052,
      "learning_rate": 9.982239516971295e-06,
      "loss": 0.0826,
      "num_tokens": 10669996.0,
      "step": 140
    },
    {
      "epoch": 0.17592014971927636,
      "grad_norm": 0.17564099775879038,
      "learning_rate": 9.98169761115397e-06,
      "loss": 0.1012,
      "num_tokens": 10746016.0,
      "step": 141
    },
    {
      "epoch": 0.1771678103555833,
      "grad_norm": 0.16763817496083627,
      "learning_rate": 9.98114757866425e-06,
      "loss": 0.0732,
      "num_tokens": 10821806.0,
      "step": 142
    },
    {
      "epoch": 0.1784154709918902,
      "grad_norm": 0.179538549410821,
      "learning_rate": 9.980589420499512e-06,
      "loss": 0.0802,
      "num_tokens": 10898660.0,
      "step": 143
    },
    {
      "epoch": 0.17966313162819714,
      "grad_norm": 0.16674096787913362,
      "learning_rate": 9.980023137671862e-06,
      "loss": 0.0715,
      "num_tokens": 10974243.0,
      "step": 144
    },
    {
      "epoch": 0.18091079226450404,
      "grad_norm": 0.1822510753659331,
      "learning_rate": 9.979448731208145e-06,
      "loss": 0.078,
      "num_tokens": 11050629.0,
      "step": 145
    },
    {
      "epoch": 0.18215845290081098,
      "grad_norm": 0.16924546334327153,
      "learning_rate": 9.978866202149931e-06,
      "loss": 0.0781,
      "num_tokens": 11126941.0,
      "step": 146
    },
    {
      "epoch": 0.18340611353711792,
      "grad_norm": 0.1677901047109414,
      "learning_rate": 9.978275551553526e-06,
      "loss": 0.0799,
      "num_tokens": 11203771.0,
      "step": 147
    },
    {
      "epoch": 0.18465377417342482,
      "grad_norm": 0.18308955197895166,
      "learning_rate": 9.977676780489953e-06,
      "loss": 0.08,
      "num_tokens": 11279637.0,
      "step": 148
    },
    {
      "epoch": 0.18590143480973176,
      "grad_norm": 0.17364312210957755,
      "learning_rate": 9.977069890044965e-06,
      "loss": 0.0768,
      "num_tokens": 11354687.0,
      "step": 149
    },
    {
      "epoch": 0.18714909544603867,
      "grad_norm": 0.1574248157414426,
      "learning_rate": 9.976454881319041e-06,
      "loss": 0.0759,
      "num_tokens": 11430507.0,
      "step": 150
    },
    {
      "epoch": 0.1883967560823456,
      "grad_norm": 0.16705589510506227,
      "learning_rate": 9.975831755427376e-06,
      "loss": 0.0804,
      "num_tokens": 11509344.0,
      "step": 151
    },
    {
      "epoch": 0.18964441671865254,
      "grad_norm": 0.18306193714265942,
      "learning_rate": 9.975200513499886e-06,
      "loss": 0.08,
      "num_tokens": 11585593.0,
      "step": 152
    },
    {
      "epoch": 0.19089207735495944,
      "grad_norm": 0.1706796151807509,
      "learning_rate": 9.974561156681203e-06,
      "loss": 0.0805,
      "num_tokens": 11661285.0,
      "step": 153
    },
    {
      "epoch": 0.19213973799126638,
      "grad_norm": 0.1760002317848842,
      "learning_rate": 9.973913686130674e-06,
      "loss": 0.0785,
      "num_tokens": 11737965.0,
      "step": 154
    },
    {
      "epoch": 0.1933873986275733,
      "grad_norm": 0.17124129463592325,
      "learning_rate": 9.973258103022361e-06,
      "loss": 0.0769,
      "num_tokens": 11813675.0,
      "step": 155
    },
    {
      "epoch": 0.19463505926388022,
      "grad_norm": 0.17169440588195126,
      "learning_rate": 9.97259440854503e-06,
      "loss": 0.0809,
      "num_tokens": 11890399.0,
      "step": 156
    },
    {
      "epoch": 0.19588271990018716,
      "grad_norm": 0.16409420786702833,
      "learning_rate": 9.971922603902164e-06,
      "loss": 0.0777,
      "num_tokens": 11967839.0,
      "step": 157
    },
    {
      "epoch": 0.19713038053649407,
      "grad_norm": 0.16387759306170735,
      "learning_rate": 9.971242690311944e-06,
      "loss": 0.07,
      "num_tokens": 12042776.0,
      "step": 158
    },
    {
      "epoch": 0.198378041172801,
      "grad_norm": 0.1718471715592683,
      "learning_rate": 9.970554669007264e-06,
      "loss": 0.0765,
      "num_tokens": 12118641.0,
      "step": 159
    },
    {
      "epoch": 0.19962570180910794,
      "grad_norm": 0.17138705121466746,
      "learning_rate": 9.969858541235708e-06,
      "loss": 0.087,
      "num_tokens": 12196351.0,
      "step": 160
    },
    {
      "epoch": 0.20087336244541484,
      "grad_norm": 0.1581669843206168,
      "learning_rate": 9.969154308259572e-06,
      "loss": 0.0758,
      "num_tokens": 12272544.0,
      "step": 161
    },
    {
      "epoch": 0.20212102308172178,
      "grad_norm": 0.17516159950920784,
      "learning_rate": 9.968441971355839e-06,
      "loss": 0.0761,
      "num_tokens": 12348033.0,
      "step": 162
    },
    {
      "epoch": 0.2033686837180287,
      "grad_norm": 0.16893850093337096,
      "learning_rate": 9.967721531816194e-06,
      "loss": 0.0769,
      "num_tokens": 12424598.0,
      "step": 163
    },
    {
      "epoch": 0.20461634435433562,
      "grad_norm": 0.18632279546740235,
      "learning_rate": 9.96699299094701e-06,
      "loss": 0.0807,
      "num_tokens": 12500157.0,
      "step": 164
    },
    {
      "epoch": 0.20586400499064256,
      "grad_norm": 0.1830366676336023,
      "learning_rate": 9.966256350069355e-06,
      "loss": 0.081,
      "num_tokens": 12577080.0,
      "step": 165
    },
    {
      "epoch": 0.20711166562694946,
      "grad_norm": 0.16985754100471287,
      "learning_rate": 9.965511610518975e-06,
      "loss": 0.085,
      "num_tokens": 12654089.0,
      "step": 166
    },
    {
      "epoch": 0.2083593262632564,
      "grad_norm": 0.18253459217534665,
      "learning_rate": 9.964758773646314e-06,
      "loss": 0.072,
      "num_tokens": 12729570.0,
      "step": 167
    },
    {
      "epoch": 0.2096069868995633,
      "grad_norm": 0.1562839597111707,
      "learning_rate": 9.963997840816491e-06,
      "loss": 0.0742,
      "num_tokens": 12805646.0,
      "step": 168
    },
    {
      "epoch": 0.21085464753587024,
      "grad_norm": 0.1636269780844281,
      "learning_rate": 9.963228813409307e-06,
      "loss": 0.0705,
      "num_tokens": 12880630.0,
      "step": 169
    },
    {
      "epoch": 0.21210230817217718,
      "grad_norm": 0.1693391795899073,
      "learning_rate": 9.962451692819238e-06,
      "loss": 0.0805,
      "num_tokens": 12958382.0,
      "step": 170
    },
    {
      "epoch": 0.2133499688084841,
      "grad_norm": 0.15840163722279885,
      "learning_rate": 9.961666480455445e-06,
      "loss": 0.0714,
      "num_tokens": 13032898.0,
      "step": 171
    },
    {
      "epoch": 0.21459762944479102,
      "grad_norm": 0.18996842682963302,
      "learning_rate": 9.96087317774175e-06,
      "loss": 0.0781,
      "num_tokens": 13111574.0,
      "step": 172
    },
    {
      "epoch": 0.21584529008109793,
      "grad_norm": 0.16107199270740594,
      "learning_rate": 9.960071786116652e-06,
      "loss": 0.0751,
      "num_tokens": 13187805.0,
      "step": 173
    },
    {
      "epoch": 0.21709295071740486,
      "grad_norm": 0.1646727452686491,
      "learning_rate": 9.959262307033318e-06,
      "loss": 0.0783,
      "num_tokens": 13266711.0,
      "step": 174
    },
    {
      "epoch": 0.2183406113537118,
      "grad_norm": 0.1684533239322113,
      "learning_rate": 9.958444741959577e-06,
      "loss": 0.0733,
      "num_tokens": 13342080.0,
      "step": 175
    },
    {
      "epoch": 0.2195882719900187,
      "grad_norm": 0.16224608231089951,
      "learning_rate": 9.957619092377921e-06,
      "loss": 0.0723,
      "num_tokens": 13417353.0,
      "step": 176
    },
    {
      "epoch": 0.22083593262632564,
      "grad_norm": 0.16923571024991296,
      "learning_rate": 9.956785359785501e-06,
      "loss": 0.0714,
      "num_tokens": 13493523.0,
      "step": 177
    },
    {
      "epoch": 0.22208359326263255,
      "grad_norm": 0.1726436941795396,
      "learning_rate": 9.95594354569413e-06,
      "loss": 0.0733,
      "num_tokens": 13569377.0,
      "step": 178
    },
    {
      "epoch": 0.22333125389893949,
      "grad_norm": 0.1716907397301449,
      "learning_rate": 9.955093651630271e-06,
      "loss": 0.0745,
      "num_tokens": 13646284.0,
      "step": 179
    },
    {
      "epoch": 0.22457891453524642,
      "grad_norm": 0.17530570629038625,
      "learning_rate": 9.954235679135035e-06,
      "loss": 0.0749,
      "num_tokens": 13723722.0,
      "step": 180
    },
    {
      "epoch": 0.22582657517155333,
      "grad_norm": 0.1697074424612026,
      "learning_rate": 9.953369629764187e-06,
      "loss": 0.0736,
      "num_tokens": 13799090.0,
      "step": 181
    },
    {
      "epoch": 0.22707423580786026,
      "grad_norm": 0.16447687811649017,
      "learning_rate": 9.952495505088138e-06,
      "loss": 0.0744,
      "num_tokens": 13878148.0,
      "step": 182
    },
    {
      "epoch": 0.2283218964441672,
      "grad_norm": 0.15744871056244594,
      "learning_rate": 9.95161330669194e-06,
      "loss": 0.0716,
      "num_tokens": 13954623.0,
      "step": 183
    },
    {
      "epoch": 0.2295695570804741,
      "grad_norm": 0.17041782127698285,
      "learning_rate": 9.950723036175282e-06,
      "loss": 0.072,
      "num_tokens": 14030485.0,
      "step": 184
    },
    {
      "epoch": 0.23081721771678104,
      "grad_norm": 0.16544306500314798,
      "learning_rate": 9.9498246951525e-06,
      "loss": 0.0708,
      "num_tokens": 14106208.0,
      "step": 185
    },
    {
      "epoch": 0.23206487835308795,
      "grad_norm": 0.16364316998876874,
      "learning_rate": 9.948918285252551e-06,
      "loss": 0.0714,
      "num_tokens": 14182086.0,
      "step": 186
    },
    {
      "epoch": 0.23331253898939489,
      "grad_norm": 0.17464505187104798,
      "learning_rate": 9.948003808119034e-06,
      "loss": 0.0671,
      "num_tokens": 14257948.0,
      "step": 187
    },
    {
      "epoch": 0.23456019962570182,
      "grad_norm": 0.16183565234470615,
      "learning_rate": 9.94708126541017e-06,
      "loss": 0.0725,
      "num_tokens": 14334070.0,
      "step": 188
    },
    {
      "epoch": 0.23580786026200873,
      "grad_norm": 0.1607703903980471,
      "learning_rate": 9.94615065879881e-06,
      "loss": 0.0658,
      "num_tokens": 14408757.0,
      "step": 189
    },
    {
      "epoch": 0.23705552089831566,
      "grad_norm": 0.16763160398244317,
      "learning_rate": 9.945211989972425e-06,
      "loss": 0.0784,
      "num_tokens": 14485925.0,
      "step": 190
    },
    {
      "epoch": 0.23830318153462257,
      "grad_norm": 0.1798190055677699,
      "learning_rate": 9.944265260633105e-06,
      "loss": 0.0805,
      "num_tokens": 14561749.0,
      "step": 191
    },
    {
      "epoch": 0.2395508421709295,
      "grad_norm": 0.16757844224240012,
      "learning_rate": 9.943310472497556e-06,
      "loss": 0.0653,
      "num_tokens": 14636925.0,
      "step": 192
    },
    {
      "epoch": 0.24079850280723644,
      "grad_norm": 0.17170337264420482,
      "learning_rate": 9.942347627297095e-06,
      "loss": 0.0674,
      "num_tokens": 14712720.0,
      "step": 193
    },
    {
      "epoch": 0.24204616344354335,
      "grad_norm": 0.184755369945989,
      "learning_rate": 9.941376726777656e-06,
      "loss": 0.0771,
      "num_tokens": 14789345.0,
      "step": 194
    },
    {
      "epoch": 0.24329382407985028,
      "grad_norm": 0.16669838986318447,
      "learning_rate": 9.940397772699773e-06,
      "loss": 0.0739,
      "num_tokens": 14865183.0,
      "step": 195
    },
    {
      "epoch": 0.2445414847161572,
      "grad_norm": 0.16033230892360856,
      "learning_rate": 9.939410766838586e-06,
      "loss": 0.0689,
      "num_tokens": 14941021.0,
      "step": 196
    },
    {
      "epoch": 0.24578914535246413,
      "grad_norm": 0.15982589644947345,
      "learning_rate": 9.938415710983834e-06,
      "loss": 0.0696,
      "num_tokens": 15016682.0,
      "step": 197
    },
    {
      "epoch": 0.24703680598877106,
      "grad_norm": 0.16352167410709095,
      "learning_rate": 9.937412606939854e-06,
      "loss": 0.0744,
      "num_tokens": 15093755.0,
      "step": 198
    },
    {
      "epoch": 0.24828446662507797,
      "grad_norm": 0.16007627601527916,
      "learning_rate": 9.936401456525578e-06,
      "loss": 0.0693,
      "num_tokens": 15168241.0,
      "step": 199
    },
    {
      "epoch": 0.2495321272613849,
      "grad_norm": 0.18748749103582008,
      "learning_rate": 9.935382261574527e-06,
      "loss": 0.07,
      "num_tokens": 15243827.0,
      "step": 200
    },
    {
      "epoch": 0.25077978789769184,
      "grad_norm": 0.16246756040982152,
      "learning_rate": 9.934355023934808e-06,
      "loss": 0.0677,
      "num_tokens": 15319199.0,
      "step": 201
    },
    {
      "epoch": 0.2520274485339988,
      "grad_norm": 0.15613654661468085,
      "learning_rate": 9.933319745469117e-06,
      "loss": 0.0633,
      "num_tokens": 15394502.0,
      "step": 202
    },
    {
      "epoch": 0.25327510917030566,
      "grad_norm": 0.16230294430226105,
      "learning_rate": 9.932276428054723e-06,
      "loss": 0.0666,
      "num_tokens": 15469765.0,
      "step": 203
    },
    {
      "epoch": 0.2545227698066126,
      "grad_norm": 0.17049974841139676,
      "learning_rate": 9.931225073583476e-06,
      "loss": 0.0692,
      "num_tokens": 15546008.0,
      "step": 204
    },
    {
      "epoch": 0.2557704304429195,
      "grad_norm": 0.1791622410807636,
      "learning_rate": 9.930165683961803e-06,
      "loss": 0.0667,
      "num_tokens": 15622373.0,
      "step": 205
    },
    {
      "epoch": 0.25701809107922646,
      "grad_norm": 0.15668763943680733,
      "learning_rate": 9.929098261110694e-06,
      "loss": 0.065,
      "num_tokens": 15697707.0,
      "step": 206
    },
    {
      "epoch": 0.2582657517155334,
      "grad_norm": 0.17590506153752944,
      "learning_rate": 9.92802280696571e-06,
      "loss": 0.0691,
      "num_tokens": 15773700.0,
      "step": 207
    },
    {
      "epoch": 0.2595134123518403,
      "grad_norm": 0.17180051970570445,
      "learning_rate": 9.926939323476976e-06,
      "loss": 0.0721,
      "num_tokens": 15850214.0,
      "step": 208
    },
    {
      "epoch": 0.2607610729881472,
      "grad_norm": 0.16004289569776617,
      "learning_rate": 9.925847812609174e-06,
      "loss": 0.065,
      "num_tokens": 15925909.0,
      "step": 209
    },
    {
      "epoch": 0.26200873362445415,
      "grad_norm": 0.16249065891465686,
      "learning_rate": 9.924748276341541e-06,
      "loss": 0.0707,
      "num_tokens": 16002947.0,
      "step": 210
    },
    {
      "epoch": 0.2632563942607611,
      "grad_norm": 0.16482669351968468,
      "learning_rate": 9.923640716667872e-06,
      "loss": 0.0707,
      "num_tokens": 16079504.0,
      "step": 211
    },
    {
      "epoch": 0.264504054897068,
      "grad_norm": 0.1580233903929819,
      "learning_rate": 9.922525135596507e-06,
      "loss": 0.067,
      "num_tokens": 16154237.0,
      "step": 212
    },
    {
      "epoch": 0.2657517155333749,
      "grad_norm": 0.1721102538638486,
      "learning_rate": 9.92140153515033e-06,
      "loss": 0.0731,
      "num_tokens": 16231125.0,
      "step": 213
    },
    {
      "epoch": 0.26699937616968183,
      "grad_norm": 0.16202985302325928,
      "learning_rate": 9.92026991736677e-06,
      "loss": 0.0724,
      "num_tokens": 16307288.0,
      "step": 214
    },
    {
      "epoch": 0.26824703680598877,
      "grad_norm": 0.16552773840332874,
      "learning_rate": 9.919130284297791e-06,
      "loss": 0.0741,
      "num_tokens": 16383863.0,
      "step": 215
    },
    {
      "epoch": 0.2694946974422957,
      "grad_norm": 0.16797535769586705,
      "learning_rate": 9.917982638009891e-06,
      "loss": 0.0722,
      "num_tokens": 16459819.0,
      "step": 216
    },
    {
      "epoch": 0.27074235807860264,
      "grad_norm": 0.15652756641636162,
      "learning_rate": 9.916826980584103e-06,
      "loss": 0.0706,
      "num_tokens": 16536109.0,
      "step": 217
    },
    {
      "epoch": 0.2719900187149095,
      "grad_norm": 0.1579731989627712,
      "learning_rate": 9.91566331411598e-06,
      "loss": 0.0732,
      "num_tokens": 16612285.0,
      "step": 218
    },
    {
      "epoch": 0.27323767935121646,
      "grad_norm": 0.16458952129615206,
      "learning_rate": 9.914491640715603e-06,
      "loss": 0.069,
      "num_tokens": 16688566.0,
      "step": 219
    },
    {
      "epoch": 0.2744853399875234,
      "grad_norm": 0.17577595597550605,
      "learning_rate": 9.913311962507569e-06,
      "loss": 0.0736,
      "num_tokens": 16765400.0,
      "step": 220
    },
    {
      "epoch": 0.2757330006238303,
      "grad_norm": 0.1555054279626727,
      "learning_rate": 9.912124281630991e-06,
      "loss": 0.0663,
      "num_tokens": 16841350.0,
      "step": 221
    },
    {
      "epoch": 0.27698066126013726,
      "grad_norm": 0.16334335494111618,
      "learning_rate": 9.910928600239493e-06,
      "loss": 0.0702,
      "num_tokens": 16917039.0,
      "step": 222
    },
    {
      "epoch": 0.27822832189644414,
      "grad_norm": 0.168630513058605,
      "learning_rate": 9.909724920501207e-06,
      "loss": 0.072,
      "num_tokens": 16993479.0,
      "step": 223
    },
    {
      "epoch": 0.2794759825327511,
      "grad_norm": 0.1601670962324957,
      "learning_rate": 9.90851324459877e-06,
      "loss": 0.0712,
      "num_tokens": 17070088.0,
      "step": 224
    },
    {
      "epoch": 0.280723643169058,
      "grad_norm": 0.15618610848645934,
      "learning_rate": 9.907293574729317e-06,
      "loss": 0.0665,
      "num_tokens": 17146961.0,
      "step": 225
    },
    {
      "epoch": 0.28197130380536495,
      "grad_norm": 0.16271573902028957,
      "learning_rate": 9.906065913104474e-06,
      "loss": 0.0676,
      "num_tokens": 17222177.0,
      "step": 226
    },
    {
      "epoch": 0.2832189644416719,
      "grad_norm": 0.16139454951559776,
      "learning_rate": 9.904830261950366e-06,
      "loss": 0.0609,
      "num_tokens": 17297238.0,
      "step": 227
    },
    {
      "epoch": 0.28446662507797876,
      "grad_norm": 0.15660531460344787,
      "learning_rate": 9.903586623507603e-06,
      "loss": 0.0652,
      "num_tokens": 17372069.0,
      "step": 228
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.14936446527678607,
      "learning_rate": 9.902335000031273e-06,
      "loss": 0.0661,
      "num_tokens": 17447408.0,
      "step": 229
    },
    {
      "epoch": 0.28696194635059263,
      "grad_norm": 0.1718394050631786,
      "learning_rate": 9.901075393790953e-06,
      "loss": 0.0691,
      "num_tokens": 17524605.0,
      "step": 230
    },
    {
      "epoch": 0.28820960698689957,
      "grad_norm": 0.1614819161412718,
      "learning_rate": 9.899807807070684e-06,
      "loss": 0.0637,
      "num_tokens": 17600453.0,
      "step": 231
    },
    {
      "epoch": 0.2894572676232065,
      "grad_norm": 0.16194959667978912,
      "learning_rate": 9.898532242168987e-06,
      "loss": 0.0663,
      "num_tokens": 17675636.0,
      "step": 232
    },
    {
      "epoch": 0.29070492825951344,
      "grad_norm": 0.2044575062693707,
      "learning_rate": 9.897248701398848e-06,
      "loss": 0.0717,
      "num_tokens": 17751533.0,
      "step": 233
    },
    {
      "epoch": 0.2919525888958203,
      "grad_norm": 0.17517502746183747,
      "learning_rate": 9.895957187087713e-06,
      "loss": 0.0675,
      "num_tokens": 17827827.0,
      "step": 234
    },
    {
      "epoch": 0.29320024953212726,
      "grad_norm": 0.1615899115066392,
      "learning_rate": 9.894657701577488e-06,
      "loss": 0.0706,
      "num_tokens": 17903877.0,
      "step": 235
    },
    {
      "epoch": 0.2944479101684342,
      "grad_norm": 0.1574811806997747,
      "learning_rate": 9.893350247224532e-06,
      "loss": 0.0652,
      "num_tokens": 17979658.0,
      "step": 236
    },
    {
      "epoch": 0.2956955708047411,
      "grad_norm": 0.16378983393360064,
      "learning_rate": 9.892034826399657e-06,
      "loss": 0.0676,
      "num_tokens": 18055261.0,
      "step": 237
    },
    {
      "epoch": 0.29694323144104806,
      "grad_norm": 0.15078063149117402,
      "learning_rate": 9.890711441488117e-06,
      "loss": 0.0653,
      "num_tokens": 18131915.0,
      "step": 238
    },
    {
      "epoch": 0.29819089207735494,
      "grad_norm": 0.171614702706413,
      "learning_rate": 9.889380094889609e-06,
      "loss": 0.0687,
      "num_tokens": 18208182.0,
      "step": 239
    },
    {
      "epoch": 0.2994385527136619,
      "grad_norm": 0.1550942059023908,
      "learning_rate": 9.888040789018267e-06,
      "loss": 0.0651,
      "num_tokens": 18283562.0,
      "step": 240
    },
    {
      "epoch": 0.3006862133499688,
      "grad_norm": 0.1564914167240467,
      "learning_rate": 9.886693526302657e-06,
      "loss": 0.0676,
      "num_tokens": 18360629.0,
      "step": 241
    },
    {
      "epoch": 0.30193387398627575,
      "grad_norm": 0.15787622574156354,
      "learning_rate": 9.885338309185775e-06,
      "loss": 0.0662,
      "num_tokens": 18437479.0,
      "step": 242
    },
    {
      "epoch": 0.3031815346225827,
      "grad_norm": 0.1630718774763793,
      "learning_rate": 9.883975140125035e-06,
      "loss": 0.0644,
      "num_tokens": 18514069.0,
      "step": 243
    },
    {
      "epoch": 0.30442919525888956,
      "grad_norm": 0.16940710112136453,
      "learning_rate": 9.88260402159228e-06,
      "loss": 0.064,
      "num_tokens": 18590589.0,
      "step": 244
    },
    {
      "epoch": 0.3056768558951965,
      "grad_norm": 0.15453236589867414,
      "learning_rate": 9.88122495607376e-06,
      "loss": 0.0603,
      "num_tokens": 18666004.0,
      "step": 245
    },
    {
      "epoch": 0.30692451653150343,
      "grad_norm": 0.15529247967534632,
      "learning_rate": 9.879837946070138e-06,
      "loss": 0.0644,
      "num_tokens": 18743017.0,
      "step": 246
    },
    {
      "epoch": 0.30817217716781037,
      "grad_norm": 0.1696878843200201,
      "learning_rate": 9.878442994096481e-06,
      "loss": 0.0691,
      "num_tokens": 18820248.0,
      "step": 247
    },
    {
      "epoch": 0.3094198378041173,
      "grad_norm": 0.1607452517295558,
      "learning_rate": 9.87704010268226e-06,
      "loss": 0.0614,
      "num_tokens": 18895528.0,
      "step": 248
    },
    {
      "epoch": 0.3106674984404242,
      "grad_norm": 0.15937732385503722,
      "learning_rate": 9.87562927437134e-06,
      "loss": 0.0652,
      "num_tokens": 18971471.0,
      "step": 249
    },
    {
      "epoch": 0.3119151590767311,
      "grad_norm": 0.15886046706962062,
      "learning_rate": 9.87421051172198e-06,
      "loss": 0.0667,
      "num_tokens": 19048325.0,
      "step": 250
    },
    {
      "epoch": 0.31316281971303805,
      "grad_norm": 0.14717021560879917,
      "learning_rate": 9.872783817306827e-06,
      "loss": 0.0625,
      "num_tokens": 19124561.0,
      "step": 251
    },
    {
      "epoch": 0.314410480349345,
      "grad_norm": 0.17052344641660577,
      "learning_rate": 9.871349193712905e-06,
      "loss": 0.0726,
      "num_tokens": 19201931.0,
      "step": 252
    },
    {
      "epoch": 0.3156581409856519,
      "grad_norm": 0.15865459451726951,
      "learning_rate": 9.869906643541625e-06,
      "loss": 0.0676,
      "num_tokens": 19278610.0,
      "step": 253
    },
    {
      "epoch": 0.3169058016219588,
      "grad_norm": 0.15293715680282027,
      "learning_rate": 9.868456169408763e-06,
      "loss": 0.0625,
      "num_tokens": 19354612.0,
      "step": 254
    },
    {
      "epoch": 0.31815346225826574,
      "grad_norm": 0.1550509113871186,
      "learning_rate": 9.866997773944469e-06,
      "loss": 0.0614,
      "num_tokens": 19430341.0,
      "step": 255
    },
    {
      "epoch": 0.3194011228945727,
      "grad_norm": 0.16470912732464232,
      "learning_rate": 9.865531459793254e-06,
      "loss": 0.0681,
      "num_tokens": 19510053.0,
      "step": 256
    },
    {
      "epoch": 0.3206487835308796,
      "grad_norm": 0.1725567554413264,
      "learning_rate": 9.864057229613988e-06,
      "loss": 0.0646,
      "num_tokens": 19587458.0,
      "step": 257
    },
    {
      "epoch": 0.32189644416718655,
      "grad_norm": 0.1649145739650761,
      "learning_rate": 9.862575086079897e-06,
      "loss": 0.0594,
      "num_tokens": 19663013.0,
      "step": 258
    },
    {
      "epoch": 0.3231441048034934,
      "grad_norm": 0.15548449511521806,
      "learning_rate": 9.861085031878556e-06,
      "loss": 0.0678,
      "num_tokens": 19738422.0,
      "step": 259
    },
    {
      "epoch": 0.32439176543980036,
      "grad_norm": 0.18027689043783654,
      "learning_rate": 9.859587069711883e-06,
      "loss": 0.0709,
      "num_tokens": 19815943.0,
      "step": 260
    },
    {
      "epoch": 0.3256394260761073,
      "grad_norm": 0.17995034139406604,
      "learning_rate": 9.858081202296133e-06,
      "loss": 0.0649,
      "num_tokens": 19891473.0,
      "step": 261
    },
    {
      "epoch": 0.32688708671241423,
      "grad_norm": 0.16879224411874624,
      "learning_rate": 9.856567432361903e-06,
      "loss": 0.0602,
      "num_tokens": 19966715.0,
      "step": 262
    },
    {
      "epoch": 0.32813474734872117,
      "grad_norm": 0.1616050002907175,
      "learning_rate": 9.855045762654115e-06,
      "loss": 0.0658,
      "num_tokens": 20043692.0,
      "step": 263
    },
    {
      "epoch": 0.32938240798502805,
      "grad_norm": 0.15471903513941546,
      "learning_rate": 9.853516195932014e-06,
      "loss": 0.0648,
      "num_tokens": 20120146.0,
      "step": 264
    },
    {
      "epoch": 0.330630068621335,
      "grad_norm": 0.16554661598465897,
      "learning_rate": 9.851978734969168e-06,
      "loss": 0.0674,
      "num_tokens": 20197768.0,
      "step": 265
    },
    {
      "epoch": 0.3318777292576419,
      "grad_norm": 0.16044112115366635,
      "learning_rate": 9.850433382553457e-06,
      "loss": 0.0644,
      "num_tokens": 20274631.0,
      "step": 266
    },
    {
      "epoch": 0.33312538989394885,
      "grad_norm": 0.15623389055861656,
      "learning_rate": 9.848880141487076e-06,
      "loss": 0.0586,
      "num_tokens": 20349379.0,
      "step": 267
    },
    {
      "epoch": 0.3343730505302558,
      "grad_norm": 0.16320331916068953,
      "learning_rate": 9.847319014586517e-06,
      "loss": 0.068,
      "num_tokens": 20425517.0,
      "step": 268
    },
    {
      "epoch": 0.33562071116656267,
      "grad_norm": 0.17410549807002865,
      "learning_rate": 9.845750004682576e-06,
      "loss": 0.0704,
      "num_tokens": 20502237.0,
      "step": 269
    },
    {
      "epoch": 0.3368683718028696,
      "grad_norm": 0.16317259612786264,
      "learning_rate": 9.844173114620342e-06,
      "loss": 0.0621,
      "num_tokens": 20577801.0,
      "step": 270
    },
    {
      "epoch": 0.33811603243917654,
      "grad_norm": 0.18438429356646444,
      "learning_rate": 9.842588347259192e-06,
      "loss": 0.0702,
      "num_tokens": 20656419.0,
      "step": 271
    },
    {
      "epoch": 0.3393636930754835,
      "grad_norm": 0.16727261944128552,
      "learning_rate": 9.84099570547279e-06,
      "loss": 0.0628,
      "num_tokens": 20731879.0,
      "step": 272
    },
    {
      "epoch": 0.3406113537117904,
      "grad_norm": 0.1568241324741727,
      "learning_rate": 9.839395192149077e-06,
      "loss": 0.0622,
      "num_tokens": 20808391.0,
      "step": 273
    },
    {
      "epoch": 0.3418590143480973,
      "grad_norm": 0.16750714409000023,
      "learning_rate": 9.837786810190268e-06,
      "loss": 0.0742,
      "num_tokens": 20886624.0,
      "step": 274
    },
    {
      "epoch": 0.3431066749844042,
      "grad_norm": 0.1624035418734388,
      "learning_rate": 9.836170562512844e-06,
      "loss": 0.0597,
      "num_tokens": 20961132.0,
      "step": 275
    },
    {
      "epoch": 0.34435433562071116,
      "grad_norm": 0.1612541838233848,
      "learning_rate": 9.83454645204755e-06,
      "loss": 0.0602,
      "num_tokens": 21037133.0,
      "step": 276
    },
    {
      "epoch": 0.3456019962570181,
      "grad_norm": 0.14663015839501767,
      "learning_rate": 9.832914481739391e-06,
      "loss": 0.0594,
      "num_tokens": 21113052.0,
      "step": 277
    },
    {
      "epoch": 0.34684965689332503,
      "grad_norm": 0.15397436093532624,
      "learning_rate": 9.831274654547623e-06,
      "loss": 0.0631,
      "num_tokens": 21189025.0,
      "step": 278
    },
    {
      "epoch": 0.34809731752963197,
      "grad_norm": 0.16081389996426831,
      "learning_rate": 9.829626973445745e-06,
      "loss": 0.0612,
      "num_tokens": 21264149.0,
      "step": 279
    },
    {
      "epoch": 0.34934497816593885,
      "grad_norm": 0.16965553989510368,
      "learning_rate": 9.827971441421504e-06,
      "loss": 0.0651,
      "num_tokens": 21341066.0,
      "step": 280
    },
    {
      "epoch": 0.3505926388022458,
      "grad_norm": 0.1565695656314987,
      "learning_rate": 9.826308061476878e-06,
      "loss": 0.0658,
      "num_tokens": 21417561.0,
      "step": 281
    },
    {
      "epoch": 0.3518402994385527,
      "grad_norm": 0.16773412549123703,
      "learning_rate": 9.824636836628078e-06,
      "loss": 0.0598,
      "num_tokens": 21492874.0,
      "step": 282
    },
    {
      "epoch": 0.35308796007485965,
      "grad_norm": 0.14945204955447638,
      "learning_rate": 9.822957769905544e-06,
      "loss": 0.0598,
      "num_tokens": 21568772.0,
      "step": 283
    },
    {
      "epoch": 0.3543356207111666,
      "grad_norm": 0.15223520296820806,
      "learning_rate": 9.821270864353924e-06,
      "loss": 0.064,
      "num_tokens": 21645126.0,
      "step": 284
    },
    {
      "epoch": 0.35558328134747347,
      "grad_norm": 0.15032291291240013,
      "learning_rate": 9.819576123032092e-06,
      "loss": 0.0623,
      "num_tokens": 21720674.0,
      "step": 285
    },
    {
      "epoch": 0.3568309419837804,
      "grad_norm": 0.15628452605672147,
      "learning_rate": 9.817873549013127e-06,
      "loss": 0.0618,
      "num_tokens": 21797336.0,
      "step": 286
    },
    {
      "epoch": 0.35807860262008734,
      "grad_norm": 0.16221897267977384,
      "learning_rate": 9.816163145384308e-06,
      "loss": 0.0694,
      "num_tokens": 21875109.0,
      "step": 287
    },
    {
      "epoch": 0.3593262632563943,
      "grad_norm": 0.16324022520446585,
      "learning_rate": 9.814444915247115e-06,
      "loss": 0.0591,
      "num_tokens": 21951197.0,
      "step": 288
    },
    {
      "epoch": 0.3605739238927012,
      "grad_norm": 0.16795042723920559,
      "learning_rate": 9.81271886171722e-06,
      "loss": 0.0648,
      "num_tokens": 22028072.0,
      "step": 289
    },
    {
      "epoch": 0.3618215845290081,
      "grad_norm": 0.16591459092130173,
      "learning_rate": 9.810984987924477e-06,
      "loss": 0.0645,
      "num_tokens": 22104892.0,
      "step": 290
    },
    {
      "epoch": 0.363069245165315,
      "grad_norm": 0.17711965952141914,
      "learning_rate": 9.809243297012923e-06,
      "loss": 0.0646,
      "num_tokens": 22181330.0,
      "step": 291
    },
    {
      "epoch": 0.36431690580162196,
      "grad_norm": 0.16517263676833355,
      "learning_rate": 9.807493792140774e-06,
      "loss": 0.0592,
      "num_tokens": 22256931.0,
      "step": 292
    },
    {
      "epoch": 0.3655645664379289,
      "grad_norm": 0.16613266514267386,
      "learning_rate": 9.805736476480407e-06,
      "loss": 0.083,
      "num_tokens": 22333756.0,
      "step": 293
    },
    {
      "epoch": 0.36681222707423583,
      "grad_norm": 0.17374361101676075,
      "learning_rate": 9.803971353218367e-06,
      "loss": 0.065,
      "num_tokens": 22409725.0,
      "step": 294
    },
    {
      "epoch": 0.3680598877105427,
      "grad_norm": 0.16484309467609493,
      "learning_rate": 9.802198425555358e-06,
      "loss": 0.0644,
      "num_tokens": 22485868.0,
      "step": 295
    },
    {
      "epoch": 0.36930754834684965,
      "grad_norm": 0.15491884230683195,
      "learning_rate": 9.800417696706234e-06,
      "loss": 0.0631,
      "num_tokens": 22562139.0,
      "step": 296
    },
    {
      "epoch": 0.3705552089831566,
      "grad_norm": 0.16556597167488193,
      "learning_rate": 9.798629169899992e-06,
      "loss": 0.0625,
      "num_tokens": 22638781.0,
      "step": 297
    },
    {
      "epoch": 0.3718028696194635,
      "grad_norm": 0.15406768450216785,
      "learning_rate": 9.796832848379775e-06,
      "loss": 0.065,
      "num_tokens": 22715605.0,
      "step": 298
    },
    {
      "epoch": 0.37305053025577045,
      "grad_norm": 0.15504153849950547,
      "learning_rate": 9.795028735402853e-06,
      "loss": 0.062,
      "num_tokens": 22791830.0,
      "step": 299
    },
    {
      "epoch": 0.37429819089207733,
      "grad_norm": 0.15412320004002014,
      "learning_rate": 9.79321683424063e-06,
      "loss": 0.0615,
      "num_tokens": 22867531.0,
      "step": 300
    },
    {
      "epoch": 0.37554585152838427,
      "grad_norm": 0.15756901782309857,
      "learning_rate": 9.791397148178632e-06,
      "loss": 0.0553,
      "num_tokens": 22943631.0,
      "step": 301
    },
    {
      "epoch": 0.3767935121646912,
      "grad_norm": 0.1531898630456781,
      "learning_rate": 9.789569680516497e-06,
      "loss": 0.0637,
      "num_tokens": 23019403.0,
      "step": 302
    },
    {
      "epoch": 0.37804117280099814,
      "grad_norm": 0.1515173653286233,
      "learning_rate": 9.78773443456798e-06,
      "loss": 0.0622,
      "num_tokens": 23095187.0,
      "step": 303
    },
    {
      "epoch": 0.3792888334373051,
      "grad_norm": 0.15625736804819645,
      "learning_rate": 9.785891413660931e-06,
      "loss": 0.0589,
      "num_tokens": 23170661.0,
      "step": 304
    },
    {
      "epoch": 0.38053649407361195,
      "grad_norm": 0.15883217445897935,
      "learning_rate": 9.784040621137308e-06,
      "loss": 0.0599,
      "num_tokens": 23247363.0,
      "step": 305
    },
    {
      "epoch": 0.3817841547099189,
      "grad_norm": 0.14990678103187605,
      "learning_rate": 9.78218206035316e-06,
      "loss": 0.0565,
      "num_tokens": 23322963.0,
      "step": 306
    },
    {
      "epoch": 0.3830318153462258,
      "grad_norm": 0.15000063559454646,
      "learning_rate": 9.780315734678612e-06,
      "loss": 0.0555,
      "num_tokens": 23398550.0,
      "step": 307
    },
    {
      "epoch": 0.38427947598253276,
      "grad_norm": 0.15831644899679806,
      "learning_rate": 9.778441647497882e-06,
      "loss": 0.0618,
      "num_tokens": 23474279.0,
      "step": 308
    },
    {
      "epoch": 0.3855271366188397,
      "grad_norm": 0.15900356422471373,
      "learning_rate": 9.776559802209255e-06,
      "loss": 0.0598,
      "num_tokens": 23551400.0,
      "step": 309
    },
    {
      "epoch": 0.3867747972551466,
      "grad_norm": 0.16055646774167992,
      "learning_rate": 9.774670202225084e-06,
      "loss": 0.0604,
      "num_tokens": 23627330.0,
      "step": 310
    },
    {
      "epoch": 0.3880224578914535,
      "grad_norm": 0.1591984868303217,
      "learning_rate": 9.772772850971788e-06,
      "loss": 0.0605,
      "num_tokens": 23704675.0,
      "step": 311
    },
    {
      "epoch": 0.38927011852776044,
      "grad_norm": 0.1503919917904232,
      "learning_rate": 9.770867751889837e-06,
      "loss": 0.0621,
      "num_tokens": 23781138.0,
      "step": 312
    },
    {
      "epoch": 0.3905177791640674,
      "grad_norm": 0.15795257667206367,
      "learning_rate": 9.76895490843375e-06,
      "loss": 0.0611,
      "num_tokens": 23856629.0,
      "step": 313
    },
    {
      "epoch": 0.3917654398003743,
      "grad_norm": 0.15149709424372163,
      "learning_rate": 9.767034324072091e-06,
      "loss": 0.0567,
      "num_tokens": 23932154.0,
      "step": 314
    },
    {
      "epoch": 0.3930131004366812,
      "grad_norm": 0.14733318425088304,
      "learning_rate": 9.76510600228746e-06,
      "loss": 0.0571,
      "num_tokens": 24007368.0,
      "step": 315
    },
    {
      "epoch": 0.39426076107298813,
      "grad_norm": 0.15325605310711474,
      "learning_rate": 9.763169946576488e-06,
      "loss": 0.0621,
      "num_tokens": 24083790.0,
      "step": 316
    },
    {
      "epoch": 0.39550842170929507,
      "grad_norm": 0.16166497874502006,
      "learning_rate": 9.76122616044983e-06,
      "loss": 0.061,
      "num_tokens": 24159562.0,
      "step": 317
    },
    {
      "epoch": 0.396756082345602,
      "grad_norm": 0.17868899974287344,
      "learning_rate": 9.759274647432156e-06,
      "loss": 0.0766,
      "num_tokens": 24237537.0,
      "step": 318
    },
    {
      "epoch": 0.39800374298190894,
      "grad_norm": 0.14894383219801544,
      "learning_rate": 9.75731541106215e-06,
      "loss": 0.0575,
      "num_tokens": 24314398.0,
      "step": 319
    },
    {
      "epoch": 0.39925140361821587,
      "grad_norm": 0.16352349044368675,
      "learning_rate": 9.755348454892498e-06,
      "loss": 0.0653,
      "num_tokens": 24390811.0,
      "step": 320
    },
    {
      "epoch": 0.40049906425452275,
      "grad_norm": 0.1538556452099385,
      "learning_rate": 9.753373782489887e-06,
      "loss": 0.0618,
      "num_tokens": 24467181.0,
      "step": 321
    },
    {
      "epoch": 0.4017467248908297,
      "grad_norm": 0.15876023490938346,
      "learning_rate": 9.751391397434996e-06,
      "loss": 0.0635,
      "num_tokens": 24543387.0,
      "step": 322
    },
    {
      "epoch": 0.4029943855271366,
      "grad_norm": 0.1715301439726509,
      "learning_rate": 9.74940130332249e-06,
      "loss": 0.0595,
      "num_tokens": 24620711.0,
      "step": 323
    },
    {
      "epoch": 0.40424204616344356,
      "grad_norm": 0.1563923945268615,
      "learning_rate": 9.747403503761006e-06,
      "loss": 0.0623,
      "num_tokens": 24697604.0,
      "step": 324
    },
    {
      "epoch": 0.4054897067997505,
      "grad_norm": 0.16071610421961047,
      "learning_rate": 9.74539800237316e-06,
      "loss": 0.0626,
      "num_tokens": 24774256.0,
      "step": 325
    },
    {
      "epoch": 0.4067373674360574,
      "grad_norm": 0.14769896169660982,
      "learning_rate": 9.743384802795535e-06,
      "loss": 0.0584,
      "num_tokens": 24849764.0,
      "step": 326
    },
    {
      "epoch": 0.4079850280723643,
      "grad_norm": 0.16996418544343941,
      "learning_rate": 9.741363908678669e-06,
      "loss": 0.0658,
      "num_tokens": 24926572.0,
      "step": 327
    },
    {
      "epoch": 0.40923268870867124,
      "grad_norm": 0.14800046537258346,
      "learning_rate": 9.739335323687052e-06,
      "loss": 0.0587,
      "num_tokens": 25002238.0,
      "step": 328
    },
    {
      "epoch": 0.4104803493449782,
      "grad_norm": 0.16207435268349787,
      "learning_rate": 9.737299051499125e-06,
      "loss": 0.0668,
      "num_tokens": 25083450.0,
      "step": 329
    },
    {
      "epoch": 0.4117280099812851,
      "grad_norm": 0.15391056645444173,
      "learning_rate": 9.735255095807263e-06,
      "loss": 0.0645,
      "num_tokens": 25160430.0,
      "step": 330
    },
    {
      "epoch": 0.412975670617592,
      "grad_norm": 0.14555280500195275,
      "learning_rate": 9.733203460317777e-06,
      "loss": 0.0599,
      "num_tokens": 25236437.0,
      "step": 331
    },
    {
      "epoch": 0.41422333125389893,
      "grad_norm": 0.15489125686680358,
      "learning_rate": 9.731144148750898e-06,
      "loss": 0.0616,
      "num_tokens": 25312474.0,
      "step": 332
    },
    {
      "epoch": 0.41547099189020587,
      "grad_norm": 0.15739674799028386,
      "learning_rate": 9.729077164840784e-06,
      "loss": 0.0632,
      "num_tokens": 25387735.0,
      "step": 333
    },
    {
      "epoch": 0.4167186525265128,
      "grad_norm": 0.16457084948558323,
      "learning_rate": 9.727002512335502e-06,
      "loss": 0.0604,
      "num_tokens": 25465301.0,
      "step": 334
    },
    {
      "epoch": 0.41796631316281974,
      "grad_norm": 0.15500114590858863,
      "learning_rate": 9.724920194997022e-06,
      "loss": 0.0582,
      "num_tokens": 25541452.0,
      "step": 335
    },
    {
      "epoch": 0.4192139737991266,
      "grad_norm": 0.1586434395647764,
      "learning_rate": 9.722830216601217e-06,
      "loss": 0.0624,
      "num_tokens": 25618060.0,
      "step": 336
    },
    {
      "epoch": 0.42046163443543355,
      "grad_norm": 0.1635042500886567,
      "learning_rate": 9.720732580937848e-06,
      "loss": 0.064,
      "num_tokens": 25695664.0,
      "step": 337
    },
    {
      "epoch": 0.4217092950717405,
      "grad_norm": 0.17358072634477886,
      "learning_rate": 9.718627291810561e-06,
      "loss": 0.0666,
      "num_tokens": 25772767.0,
      "step": 338
    },
    {
      "epoch": 0.4229569557080474,
      "grad_norm": 0.1497956417489676,
      "learning_rate": 9.716514353036884e-06,
      "loss": 0.0582,
      "num_tokens": 25851734.0,
      "step": 339
    },
    {
      "epoch": 0.42420461634435436,
      "grad_norm": 0.15852636906702566,
      "learning_rate": 9.714393768448214e-06,
      "loss": 0.0602,
      "num_tokens": 25928318.0,
      "step": 340
    },
    {
      "epoch": 0.42545227698066124,
      "grad_norm": 0.1561377489499093,
      "learning_rate": 9.712265541889809e-06,
      "loss": 0.0539,
      "num_tokens": 26003907.0,
      "step": 341
    },
    {
      "epoch": 0.4266999376169682,
      "grad_norm": 0.16140448999169976,
      "learning_rate": 9.710129677220788e-06,
      "loss": 0.061,
      "num_tokens": 26079673.0,
      "step": 342
    },
    {
      "epoch": 0.4279475982532751,
      "grad_norm": 0.14579047647672982,
      "learning_rate": 9.707986178314123e-06,
      "loss": 0.0519,
      "num_tokens": 26154655.0,
      "step": 343
    },
    {
      "epoch": 0.42919525888958204,
      "grad_norm": 0.153370210574673,
      "learning_rate": 9.705835049056621e-06,
      "loss": 0.0564,
      "num_tokens": 26230489.0,
      "step": 344
    },
    {
      "epoch": 0.430442919525889,
      "grad_norm": 0.14963027725660538,
      "learning_rate": 9.70367629334893e-06,
      "loss": 0.0556,
      "num_tokens": 26306723.0,
      "step": 345
    },
    {
      "epoch": 0.43169058016219586,
      "grad_norm": 0.15481215278421856,
      "learning_rate": 9.701509915105527e-06,
      "loss": 0.0582,
      "num_tokens": 26382427.0,
      "step": 346
    },
    {
      "epoch": 0.4329382407985028,
      "grad_norm": 0.15019442485278475,
      "learning_rate": 9.699335918254714e-06,
      "loss": 0.0586,
      "num_tokens": 26457992.0,
      "step": 347
    },
    {
      "epoch": 0.43418590143480973,
      "grad_norm": 0.14883402591779346,
      "learning_rate": 9.6971543067386e-06,
      "loss": 0.0536,
      "num_tokens": 26533426.0,
      "step": 348
    },
    {
      "epoch": 0.43543356207111666,
      "grad_norm": 0.1493537080558344,
      "learning_rate": 9.694965084513106e-06,
      "loss": 0.0541,
      "num_tokens": 26608589.0,
      "step": 349
    },
    {
      "epoch": 0.4366812227074236,
      "grad_norm": 0.1548264803254639,
      "learning_rate": 9.692768255547957e-06,
      "loss": 0.0592,
      "num_tokens": 26685043.0,
      "step": 350
    },
    {
      "epoch": 0.4379288833437305,
      "grad_norm": 0.15333844028749696,
      "learning_rate": 9.690563823826666e-06,
      "loss": 0.0554,
      "num_tokens": 26760976.0,
      "step": 351
    },
    {
      "epoch": 0.4391765439800374,
      "grad_norm": 0.1482621937184213,
      "learning_rate": 9.688351793346533e-06,
      "loss": 0.0544,
      "num_tokens": 26837215.0,
      "step": 352
    },
    {
      "epoch": 0.44042420461634435,
      "grad_norm": 0.15640660305026827,
      "learning_rate": 9.68613216811864e-06,
      "loss": 0.0575,
      "num_tokens": 26913730.0,
      "step": 353
    },
    {
      "epoch": 0.4416718652526513,
      "grad_norm": 0.15020500164794995,
      "learning_rate": 9.683904952167837e-06,
      "loss": 0.0559,
      "num_tokens": 26989823.0,
      "step": 354
    },
    {
      "epoch": 0.4429195258889582,
      "grad_norm": 0.17001807801907493,
      "learning_rate": 9.681670149532739e-06,
      "loss": 0.0641,
      "num_tokens": 27066112.0,
      "step": 355
    },
    {
      "epoch": 0.4441671865252651,
      "grad_norm": 0.1498412004993274,
      "learning_rate": 9.67942776426572e-06,
      "loss": 0.0625,
      "num_tokens": 27144273.0,
      "step": 356
    },
    {
      "epoch": 0.44541484716157204,
      "grad_norm": 0.1534522416368002,
      "learning_rate": 9.677177800432903e-06,
      "loss": 0.0622,
      "num_tokens": 27220942.0,
      "step": 357
    },
    {
      "epoch": 0.44666250779787897,
      "grad_norm": 0.1550843398492664,
      "learning_rate": 9.67492026211415e-06,
      "loss": 0.0616,
      "num_tokens": 27297340.0,
      "step": 358
    },
    {
      "epoch": 0.4479101684341859,
      "grad_norm": 0.15391171863206848,
      "learning_rate": 9.672655153403064e-06,
      "loss": 0.0601,
      "num_tokens": 27373562.0,
      "step": 359
    },
    {
      "epoch": 0.44915782907049284,
      "grad_norm": 0.15222482131910586,
      "learning_rate": 9.670382478406967e-06,
      "loss": 0.0631,
      "num_tokens": 27450810.0,
      "step": 360
    },
    {
      "epoch": 0.4504054897067998,
      "grad_norm": 0.15561420752022959,
      "learning_rate": 9.66810224124691e-06,
      "loss": 0.056,
      "num_tokens": 27526478.0,
      "step": 361
    },
    {
      "epoch": 0.45165315034310666,
      "grad_norm": 0.16603159970455272,
      "learning_rate": 9.665814446057652e-06,
      "loss": 0.0556,
      "num_tokens": 27602783.0,
      "step": 362
    },
    {
      "epoch": 0.4529008109794136,
      "grad_norm": 0.14679941146709885,
      "learning_rate": 9.663519096987653e-06,
      "loss": 0.0553,
      "num_tokens": 27678758.0,
      "step": 363
    },
    {
      "epoch": 0.45414847161572053,
      "grad_norm": 0.15949940287162023,
      "learning_rate": 9.661216198199078e-06,
      "loss": 0.0614,
      "num_tokens": 27755052.0,
      "step": 364
    },
    {
      "epoch": 0.45539613225202746,
      "grad_norm": 0.1524486410430796,
      "learning_rate": 9.658905753867778e-06,
      "loss": 0.0642,
      "num_tokens": 27832939.0,
      "step": 365
    },
    {
      "epoch": 0.4566437928883344,
      "grad_norm": 0.16992806811708352,
      "learning_rate": 9.656587768183287e-06,
      "loss": 0.0578,
      "num_tokens": 27909228.0,
      "step": 366
    },
    {
      "epoch": 0.4578914535246413,
      "grad_norm": 0.14352236801233487,
      "learning_rate": 9.654262245348813e-06,
      "loss": 0.0571,
      "num_tokens": 27985198.0,
      "step": 367
    },
    {
      "epoch": 0.4591391141609482,
      "grad_norm": 0.15126724066365793,
      "learning_rate": 9.651929189581233e-06,
      "loss": 0.0589,
      "num_tokens": 28061265.0,
      "step": 368
    },
    {
      "epoch": 0.46038677479725515,
      "grad_norm": 0.17699813009849166,
      "learning_rate": 9.649588605111082e-06,
      "loss": 0.0607,
      "num_tokens": 28139144.0,
      "step": 369
    },
    {
      "epoch": 0.4616344354335621,
      "grad_norm": 0.152592012188265,
      "learning_rate": 9.647240496182545e-06,
      "loss": 0.0547,
      "num_tokens": 28216085.0,
      "step": 370
    },
    {
      "epoch": 0.462882096069869,
      "grad_norm": 0.1443972817699674,
      "learning_rate": 9.644884867053455e-06,
      "loss": 0.0546,
      "num_tokens": 28291162.0,
      "step": 371
    },
    {
      "epoch": 0.4641297567061759,
      "grad_norm": 0.14799715036949382,
      "learning_rate": 9.64252172199528e-06,
      "loss": 0.0575,
      "num_tokens": 28367050.0,
      "step": 372
    },
    {
      "epoch": 0.46537741734248284,
      "grad_norm": 0.15809469352500546,
      "learning_rate": 9.640151065293117e-06,
      "loss": 0.0759,
      "num_tokens": 28442851.0,
      "step": 373
    },
    {
      "epoch": 0.46662507797878977,
      "grad_norm": 0.14924075699579795,
      "learning_rate": 9.63777290124568e-06,
      "loss": 0.0536,
      "num_tokens": 28518374.0,
      "step": 374
    },
    {
      "epoch": 0.4678727386150967,
      "grad_norm": 0.14464935799941636,
      "learning_rate": 9.635387234165303e-06,
      "loss": 0.0558,
      "num_tokens": 28594304.0,
      "step": 375
    },
    {
      "epoch": 0.46912039925140364,
      "grad_norm": 0.15717484410743035,
      "learning_rate": 9.632994068377916e-06,
      "loss": 0.0602,
      "num_tokens": 28670659.0,
      "step": 376
    },
    {
      "epoch": 0.4703680598877105,
      "grad_norm": 0.14844579724039525,
      "learning_rate": 9.63059340822306e-06,
      "loss": 0.0598,
      "num_tokens": 28746845.0,
      "step": 377
    },
    {
      "epoch": 0.47161572052401746,
      "grad_norm": 0.1594886034213672,
      "learning_rate": 9.628185258053852e-06,
      "loss": 0.0609,
      "num_tokens": 28822799.0,
      "step": 378
    },
    {
      "epoch": 0.4728633811603244,
      "grad_norm": 0.14548531881469184,
      "learning_rate": 9.625769622236995e-06,
      "loss": 0.0572,
      "num_tokens": 28899534.0,
      "step": 379
    },
    {
      "epoch": 0.4741110417966313,
      "grad_norm": 0.15452383568551192,
      "learning_rate": 9.623346505152771e-06,
      "loss": 0.0588,
      "num_tokens": 28975820.0,
      "step": 380
    },
    {
      "epoch": 0.47535870243293826,
      "grad_norm": 0.15002724218298902,
      "learning_rate": 9.620915911195021e-06,
      "loss": 0.0588,
      "num_tokens": 29053347.0,
      "step": 381
    },
    {
      "epoch": 0.47660636306924514,
      "grad_norm": 0.15471346976150332,
      "learning_rate": 9.618477844771147e-06,
      "loss": 0.0557,
      "num_tokens": 29128816.0,
      "step": 382
    },
    {
      "epoch": 0.4778540237055521,
      "grad_norm": 0.1550002737473579,
      "learning_rate": 9.6160323103021e-06,
      "loss": 0.0591,
      "num_tokens": 29204981.0,
      "step": 383
    },
    {
      "epoch": 0.479101684341859,
      "grad_norm": 0.15149835324532465,
      "learning_rate": 9.613579312222377e-06,
      "loss": 0.0542,
      "num_tokens": 29280430.0,
      "step": 384
    },
    {
      "epoch": 0.48034934497816595,
      "grad_norm": 0.16201311454989606,
      "learning_rate": 9.611118854979998e-06,
      "loss": 0.0597,
      "num_tokens": 29357489.0,
      "step": 385
    },
    {
      "epoch": 0.4815970056144729,
      "grad_norm": 0.14974683349924134,
      "learning_rate": 9.608650943036522e-06,
      "loss": 0.0557,
      "num_tokens": 29433257.0,
      "step": 386
    },
    {
      "epoch": 0.48284466625077976,
      "grad_norm": 0.1559608713414317,
      "learning_rate": 9.606175580867016e-06,
      "loss": 0.0567,
      "num_tokens": 29509590.0,
      "step": 387
    },
    {
      "epoch": 0.4840923268870867,
      "grad_norm": 0.15336009260629543,
      "learning_rate": 9.60369277296006e-06,
      "loss": 0.0561,
      "num_tokens": 29587782.0,
      "step": 388
    },
    {
      "epoch": 0.48533998752339363,
      "grad_norm": 0.15936225446679803,
      "learning_rate": 9.601202523817735e-06,
      "loss": 0.0525,
      "num_tokens": 29662613.0,
      "step": 389
    },
    {
      "epoch": 0.48658764815970057,
      "grad_norm": 0.1472806374581374,
      "learning_rate": 9.598704837955618e-06,
      "loss": 0.0546,
      "num_tokens": 29739245.0,
      "step": 390
    },
    {
      "epoch": 0.4878353087960075,
      "grad_norm": 0.16550022474863885,
      "learning_rate": 9.596199719902765e-06,
      "loss": 0.0616,
      "num_tokens": 29816275.0,
      "step": 391
    },
    {
      "epoch": 0.4890829694323144,
      "grad_norm": 0.16013358324619242,
      "learning_rate": 9.593687174201715e-06,
      "loss": 0.0566,
      "num_tokens": 29893157.0,
      "step": 392
    },
    {
      "epoch": 0.4903306300686213,
      "grad_norm": 0.1516369697660201,
      "learning_rate": 9.59116720540847e-06,
      "loss": 0.0547,
      "num_tokens": 29969852.0,
      "step": 393
    },
    {
      "epoch": 0.49157829070492826,
      "grad_norm": 0.15400623283535445,
      "learning_rate": 9.588639818092498e-06,
      "loss": 0.0574,
      "num_tokens": 30046961.0,
      "step": 394
    },
    {
      "epoch": 0.4928259513412352,
      "grad_norm": 0.14891759374811117,
      "learning_rate": 9.586105016836713e-06,
      "loss": 0.057,
      "num_tokens": 30123438.0,
      "step": 395
    },
    {
      "epoch": 0.4940736119775421,
      "grad_norm": 0.1543457655390604,
      "learning_rate": 9.58356280623748e-06,
      "loss": 0.0568,
      "num_tokens": 30199647.0,
      "step": 396
    },
    {
      "epoch": 0.495321272613849,
      "grad_norm": 0.16009856686730312,
      "learning_rate": 9.58101319090459e-06,
      "loss": 0.056,
      "num_tokens": 30276427.0,
      "step": 397
    },
    {
      "epoch": 0.49656893325015594,
      "grad_norm": 0.14536110498931284,
      "learning_rate": 9.578456175461272e-06,
      "loss": 0.0539,
      "num_tokens": 30353184.0,
      "step": 398
    },
    {
      "epoch": 0.4978165938864629,
      "grad_norm": 0.15472828164087113,
      "learning_rate": 9.575891764544162e-06,
      "loss": 0.0587,
      "num_tokens": 30429481.0,
      "step": 399
    },
    {
      "epoch": 0.4990642545227698,
      "grad_norm": 0.1600147319756837,
      "learning_rate": 9.573319962803317e-06,
      "loss": 0.0564,
      "num_tokens": 30505682.0,
      "step": 400
    },
    {
      "epoch": 0.5003119151590767,
      "grad_norm": 0.14645724562558174,
      "learning_rate": 9.570740774902189e-06,
      "loss": 0.0533,
      "num_tokens": 30581573.0,
      "step": 401
    },
    {
      "epoch": 0.5015595757953837,
      "grad_norm": 0.15711265382750586,
      "learning_rate": 9.568154205517623e-06,
      "loss": 0.0577,
      "num_tokens": 30657816.0,
      "step": 402
    },
    {
      "epoch": 0.5028072364316906,
      "grad_norm": 0.15283501905181346,
      "learning_rate": 9.565560259339856e-06,
      "loss": 0.0609,
      "num_tokens": 30733986.0,
      "step": 403
    },
    {
      "epoch": 0.5040548970679976,
      "grad_norm": 0.16340162533953054,
      "learning_rate": 9.562958941072491e-06,
      "loss": 0.0639,
      "num_tokens": 30812920.0,
      "step": 404
    },
    {
      "epoch": 0.5053025577043044,
      "grad_norm": 0.14460855761419414,
      "learning_rate": 9.560350255432508e-06,
      "loss": 0.0533,
      "num_tokens": 30889729.0,
      "step": 405
    },
    {
      "epoch": 0.5065502183406113,
      "grad_norm": 0.15949124398271095,
      "learning_rate": 9.557734207150243e-06,
      "loss": 0.0581,
      "num_tokens": 30965816.0,
      "step": 406
    },
    {
      "epoch": 0.5077978789769183,
      "grad_norm": 0.13801699119897257,
      "learning_rate": 9.55511080096938e-06,
      "loss": 0.0538,
      "num_tokens": 31041927.0,
      "step": 407
    },
    {
      "epoch": 0.5090455396132252,
      "grad_norm": 0.13912012303471716,
      "learning_rate": 9.552480041646949e-06,
      "loss": 0.0531,
      "num_tokens": 31117687.0,
      "step": 408
    },
    {
      "epoch": 0.5102932002495322,
      "grad_norm": 0.1629842644162694,
      "learning_rate": 9.549841933953308e-06,
      "loss": 0.0589,
      "num_tokens": 31194057.0,
      "step": 409
    },
    {
      "epoch": 0.511540860885839,
      "grad_norm": 0.15827973713090576,
      "learning_rate": 9.547196482672148e-06,
      "loss": 0.0588,
      "num_tokens": 31270979.0,
      "step": 410
    },
    {
      "epoch": 0.5127885215221459,
      "grad_norm": 0.15401143305786705,
      "learning_rate": 9.544543692600473e-06,
      "loss": 0.0572,
      "num_tokens": 31347683.0,
      "step": 411
    },
    {
      "epoch": 0.5140361821584529,
      "grad_norm": 0.1658146704954777,
      "learning_rate": 9.541883568548588e-06,
      "loss": 0.0596,
      "num_tokens": 31424509.0,
      "step": 412
    },
    {
      "epoch": 0.5152838427947598,
      "grad_norm": 0.14602113674983888,
      "learning_rate": 9.539216115340106e-06,
      "loss": 0.0546,
      "num_tokens": 31500007.0,
      "step": 413
    },
    {
      "epoch": 0.5165315034310668,
      "grad_norm": 0.1568068093453694,
      "learning_rate": 9.536541337811923e-06,
      "loss": 0.0567,
      "num_tokens": 31576786.0,
      "step": 414
    },
    {
      "epoch": 0.5177791640673737,
      "grad_norm": 0.14427631505975017,
      "learning_rate": 9.533859240814221e-06,
      "loss": 0.0555,
      "num_tokens": 31652782.0,
      "step": 415
    },
    {
      "epoch": 0.5190268247036806,
      "grad_norm": 0.13640071991365751,
      "learning_rate": 9.531169829210452e-06,
      "loss": 0.0499,
      "num_tokens": 31728711.0,
      "step": 416
    },
    {
      "epoch": 0.5202744853399875,
      "grad_norm": 0.15127104928479138,
      "learning_rate": 9.528473107877333e-06,
      "loss": 0.0566,
      "num_tokens": 31806014.0,
      "step": 417
    },
    {
      "epoch": 0.5215221459762944,
      "grad_norm": 0.14309135833657355,
      "learning_rate": 9.525769081704835e-06,
      "loss": 0.0489,
      "num_tokens": 31880741.0,
      "step": 418
    },
    {
      "epoch": 0.5227698066126014,
      "grad_norm": 0.15539877823691686,
      "learning_rate": 9.523057755596174e-06,
      "loss": 0.0529,
      "num_tokens": 31956173.0,
      "step": 419
    },
    {
      "epoch": 0.5240174672489083,
      "grad_norm": 0.17228729683993205,
      "learning_rate": 9.520339134467803e-06,
      "loss": 0.0547,
      "num_tokens": 32032399.0,
      "step": 420
    },
    {
      "epoch": 0.5252651278852152,
      "grad_norm": 0.1583895665904868,
      "learning_rate": 9.517613223249402e-06,
      "loss": 0.0533,
      "num_tokens": 32108181.0,
      "step": 421
    },
    {
      "epoch": 0.5265127885215222,
      "grad_norm": 0.15998682022689042,
      "learning_rate": 9.514880026883877e-06,
      "loss": 0.0591,
      "num_tokens": 32184396.0,
      "step": 422
    },
    {
      "epoch": 0.527760449157829,
      "grad_norm": 0.1565903843630229,
      "learning_rate": 9.512139550327338e-06,
      "loss": 0.0549,
      "num_tokens": 32259958.0,
      "step": 423
    },
    {
      "epoch": 0.529008109794136,
      "grad_norm": 0.15825635032436505,
      "learning_rate": 9.509391798549091e-06,
      "loss": 0.0499,
      "num_tokens": 32335779.0,
      "step": 424
    },
    {
      "epoch": 0.5302557704304429,
      "grad_norm": 0.15151788807905717,
      "learning_rate": 9.50663677653165e-06,
      "loss": 0.0613,
      "num_tokens": 32412537.0,
      "step": 425
    },
    {
      "epoch": 0.5315034310667498,
      "grad_norm": 0.14537670619366175,
      "learning_rate": 9.503874489270697e-06,
      "loss": 0.0546,
      "num_tokens": 32489973.0,
      "step": 426
    },
    {
      "epoch": 0.5327510917030568,
      "grad_norm": 0.14119646580930728,
      "learning_rate": 9.501104941775094e-06,
      "loss": 0.0523,
      "num_tokens": 32566189.0,
      "step": 427
    },
    {
      "epoch": 0.5339987523393637,
      "grad_norm": 0.15182301317440733,
      "learning_rate": 9.49832813906687e-06,
      "loss": 0.054,
      "num_tokens": 32642863.0,
      "step": 428
    },
    {
      "epoch": 0.5352464129756707,
      "grad_norm": 0.16184584308403277,
      "learning_rate": 9.495544086181204e-06,
      "loss": 0.0513,
      "num_tokens": 32718435.0,
      "step": 429
    },
    {
      "epoch": 0.5364940736119775,
      "grad_norm": 0.14525021676931954,
      "learning_rate": 9.49275278816643e-06,
      "loss": 0.0566,
      "num_tokens": 32794713.0,
      "step": 430
    },
    {
      "epoch": 0.5377417342482844,
      "grad_norm": 0.14780472666005015,
      "learning_rate": 9.489954250084011e-06,
      "loss": 0.0536,
      "num_tokens": 32872091.0,
      "step": 431
    },
    {
      "epoch": 0.5389893948845914,
      "grad_norm": 0.15645539651072504,
      "learning_rate": 9.487148477008545e-06,
      "loss": 0.0538,
      "num_tokens": 32947802.0,
      "step": 432
    },
    {
      "epoch": 0.5402370555208983,
      "grad_norm": 0.14181033938548343,
      "learning_rate": 9.484335474027744e-06,
      "loss": 0.0536,
      "num_tokens": 33025155.0,
      "step": 433
    },
    {
      "epoch": 0.5414847161572053,
      "grad_norm": 0.15415741993862173,
      "learning_rate": 9.481515246242435e-06,
      "loss": 0.0561,
      "num_tokens": 33101169.0,
      "step": 434
    },
    {
      "epoch": 0.5427323767935122,
      "grad_norm": 0.1490145272995484,
      "learning_rate": 9.478687798766544e-06,
      "loss": 0.0549,
      "num_tokens": 33178375.0,
      "step": 435
    },
    {
      "epoch": 0.543980037429819,
      "grad_norm": 0.15402426723128798,
      "learning_rate": 9.475853136727086e-06,
      "loss": 0.0546,
      "num_tokens": 33253461.0,
      "step": 436
    },
    {
      "epoch": 0.545227698066126,
      "grad_norm": 0.15545840379696368,
      "learning_rate": 9.473011265264159e-06,
      "loss": 0.0536,
      "num_tokens": 33329434.0,
      "step": 437
    },
    {
      "epoch": 0.5464753587024329,
      "grad_norm": 0.14007805927580225,
      "learning_rate": 9.470162189530938e-06,
      "loss": 0.0582,
      "num_tokens": 33405693.0,
      "step": 438
    },
    {
      "epoch": 0.5477230193387399,
      "grad_norm": 0.14677878355239657,
      "learning_rate": 9.467305914693658e-06,
      "loss": 0.0546,
      "num_tokens": 33481868.0,
      "step": 439
    },
    {
      "epoch": 0.5489706799750468,
      "grad_norm": 0.13752170718015977,
      "learning_rate": 9.464442445931605e-06,
      "loss": 0.051,
      "num_tokens": 33557214.0,
      "step": 440
    },
    {
      "epoch": 0.5502183406113537,
      "grad_norm": 0.15178088205554877,
      "learning_rate": 9.461571788437119e-06,
      "loss": 0.0576,
      "num_tokens": 33632980.0,
      "step": 441
    },
    {
      "epoch": 0.5514660012476607,
      "grad_norm": 0.14398063856278132,
      "learning_rate": 9.458693947415564e-06,
      "loss": 0.0535,
      "num_tokens": 33710087.0,
      "step": 442
    },
    {
      "epoch": 0.5527136618839675,
      "grad_norm": 0.14539327976112087,
      "learning_rate": 9.455808928085339e-06,
      "loss": 0.0519,
      "num_tokens": 33785845.0,
      "step": 443
    },
    {
      "epoch": 0.5539613225202745,
      "grad_norm": 0.1450732701024791,
      "learning_rate": 9.452916735677857e-06,
      "loss": 0.0556,
      "num_tokens": 33861967.0,
      "step": 444
    },
    {
      "epoch": 0.5552089831565814,
      "grad_norm": 0.13986151085445409,
      "learning_rate": 9.450017375437534e-06,
      "loss": 0.0501,
      "num_tokens": 33937887.0,
      "step": 445
    },
    {
      "epoch": 0.5564566437928883,
      "grad_norm": 0.16219894351016226,
      "learning_rate": 9.44711085262179e-06,
      "loss": 0.0573,
      "num_tokens": 34014701.0,
      "step": 446
    },
    {
      "epoch": 0.5577043044291953,
      "grad_norm": 0.15868863153077056,
      "learning_rate": 9.444197172501025e-06,
      "loss": 0.0559,
      "num_tokens": 34090915.0,
      "step": 447
    },
    {
      "epoch": 0.5589519650655022,
      "grad_norm": 0.15055297288103786,
      "learning_rate": 9.441276340358624e-06,
      "loss": 0.0515,
      "num_tokens": 34167324.0,
      "step": 448
    },
    {
      "epoch": 0.5601996257018091,
      "grad_norm": 0.15550675387011964,
      "learning_rate": 9.438348361490938e-06,
      "loss": 0.056,
      "num_tokens": 34243404.0,
      "step": 449
    },
    {
      "epoch": 0.561447286338116,
      "grad_norm": 0.14516804951124496,
      "learning_rate": 9.43541324120728e-06,
      "loss": 0.0521,
      "num_tokens": 34319033.0,
      "step": 450
    },
    {
      "epoch": 0.5626949469744229,
      "grad_norm": 0.14219558052085499,
      "learning_rate": 9.432470984829908e-06,
      "loss": 0.0582,
      "num_tokens": 34397287.0,
      "step": 451
    },
    {
      "epoch": 0.5639426076107299,
      "grad_norm": 0.14782808411123538,
      "learning_rate": 9.429521597694023e-06,
      "loss": 0.0537,
      "num_tokens": 34473491.0,
      "step": 452
    },
    {
      "epoch": 0.5651902682470368,
      "grad_norm": 0.14684446756874775,
      "learning_rate": 9.426565085147755e-06,
      "loss": 0.055,
      "num_tokens": 34550056.0,
      "step": 453
    },
    {
      "epoch": 0.5664379288833438,
      "grad_norm": 0.1476528606816408,
      "learning_rate": 9.423601452552153e-06,
      "loss": 0.0528,
      "num_tokens": 34626574.0,
      "step": 454
    },
    {
      "epoch": 0.5676855895196506,
      "grad_norm": 0.15032224550644455,
      "learning_rate": 9.420630705281182e-06,
      "loss": 0.0521,
      "num_tokens": 34702812.0,
      "step": 455
    },
    {
      "epoch": 0.5689332501559575,
      "grad_norm": 0.15507617585697697,
      "learning_rate": 9.417652848721704e-06,
      "loss": 0.0555,
      "num_tokens": 34779202.0,
      "step": 456
    },
    {
      "epoch": 0.5701809107922645,
      "grad_norm": 0.13946889304429896,
      "learning_rate": 9.41466788827347e-06,
      "loss": 0.0493,
      "num_tokens": 34854641.0,
      "step": 457
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.15804626765822227,
      "learning_rate": 9.411675829349119e-06,
      "loss": 0.0542,
      "num_tokens": 34931173.0,
      "step": 458
    },
    {
      "epoch": 0.5726762320648784,
      "grad_norm": 0.15864241799013198,
      "learning_rate": 9.408676677374158e-06,
      "loss": 0.0554,
      "num_tokens": 35007346.0,
      "step": 459
    },
    {
      "epoch": 0.5739238927011853,
      "grad_norm": 0.1486978082092752,
      "learning_rate": 9.405670437786953e-06,
      "loss": 0.0503,
      "num_tokens": 35082953.0,
      "step": 460
    },
    {
      "epoch": 0.5751715533374921,
      "grad_norm": 0.15936338978994993,
      "learning_rate": 9.402657116038728e-06,
      "loss": 0.0557,
      "num_tokens": 35159654.0,
      "step": 461
    },
    {
      "epoch": 0.5764192139737991,
      "grad_norm": 0.1419878547063181,
      "learning_rate": 9.399636717593545e-06,
      "loss": 0.0518,
      "num_tokens": 35234785.0,
      "step": 462
    },
    {
      "epoch": 0.577666874610106,
      "grad_norm": 0.15864527117251057,
      "learning_rate": 9.3966092479283e-06,
      "loss": 0.0552,
      "num_tokens": 35310616.0,
      "step": 463
    },
    {
      "epoch": 0.578914535246413,
      "grad_norm": 0.1503705578277098,
      "learning_rate": 9.39357471253271e-06,
      "loss": 0.0505,
      "num_tokens": 35385435.0,
      "step": 464
    },
    {
      "epoch": 0.5801621958827199,
      "grad_norm": 0.16117755152514393,
      "learning_rate": 9.390533116909305e-06,
      "loss": 0.0579,
      "num_tokens": 35463241.0,
      "step": 465
    },
    {
      "epoch": 0.5814098565190269,
      "grad_norm": 0.13917139410745755,
      "learning_rate": 9.387484466573417e-06,
      "loss": 0.0501,
      "num_tokens": 35538049.0,
      "step": 466
    },
    {
      "epoch": 0.5826575171553338,
      "grad_norm": 0.15269748868147934,
      "learning_rate": 9.38442876705317e-06,
      "loss": 0.0545,
      "num_tokens": 35613417.0,
      "step": 467
    },
    {
      "epoch": 0.5839051777916406,
      "grad_norm": 0.14011033067669287,
      "learning_rate": 9.381366023889475e-06,
      "loss": 0.0507,
      "num_tokens": 35689560.0,
      "step": 468
    },
    {
      "epoch": 0.5851528384279476,
      "grad_norm": 0.1605155443204243,
      "learning_rate": 9.378296242636002e-06,
      "loss": 0.0511,
      "num_tokens": 35764777.0,
      "step": 469
    },
    {
      "epoch": 0.5864004990642545,
      "grad_norm": 0.14318225311143812,
      "learning_rate": 9.375219428859202e-06,
      "loss": 0.0523,
      "num_tokens": 35840524.0,
      "step": 470
    },
    {
      "epoch": 0.5876481597005615,
      "grad_norm": 0.15081095543187126,
      "learning_rate": 9.372135588138262e-06,
      "loss": 0.0541,
      "num_tokens": 35919754.0,
      "step": 471
    },
    {
      "epoch": 0.5888958203368684,
      "grad_norm": 0.167227205565762,
      "learning_rate": 9.369044726065121e-06,
      "loss": 0.0524,
      "num_tokens": 35995857.0,
      "step": 472
    },
    {
      "epoch": 0.5901434809731753,
      "grad_norm": 0.14110886145114976,
      "learning_rate": 9.365946848244445e-06,
      "loss": 0.0512,
      "num_tokens": 36072256.0,
      "step": 473
    },
    {
      "epoch": 0.5913911416094823,
      "grad_norm": 0.1566742798779281,
      "learning_rate": 9.362841960293622e-06,
      "loss": 0.0546,
      "num_tokens": 36149372.0,
      "step": 474
    },
    {
      "epoch": 0.5926388022457891,
      "grad_norm": 0.1457390981400815,
      "learning_rate": 9.359730067842753e-06,
      "loss": 0.0539,
      "num_tokens": 36225637.0,
      "step": 475
    },
    {
      "epoch": 0.5938864628820961,
      "grad_norm": 0.14052219234375607,
      "learning_rate": 9.35661117653464e-06,
      "loss": 0.0491,
      "num_tokens": 36300751.0,
      "step": 476
    },
    {
      "epoch": 0.595134123518403,
      "grad_norm": 0.14585199639575464,
      "learning_rate": 9.353485292024775e-06,
      "loss": 0.0515,
      "num_tokens": 36377358.0,
      "step": 477
    },
    {
      "epoch": 0.5963817841547099,
      "grad_norm": 0.13602504917204367,
      "learning_rate": 9.35035241998133e-06,
      "loss": 0.0483,
      "num_tokens": 36452580.0,
      "step": 478
    },
    {
      "epoch": 0.5976294447910169,
      "grad_norm": 0.16038197120224465,
      "learning_rate": 9.347212566085153e-06,
      "loss": 0.0537,
      "num_tokens": 36527842.0,
      "step": 479
    },
    {
      "epoch": 0.5988771054273238,
      "grad_norm": 0.14394564870839627,
      "learning_rate": 9.344065736029746e-06,
      "loss": 0.0502,
      "num_tokens": 36603703.0,
      "step": 480
    },
    {
      "epoch": 0.6001247660636307,
      "grad_norm": 0.14127804811329153,
      "learning_rate": 9.34091193552126e-06,
      "loss": 0.0539,
      "num_tokens": 36680671.0,
      "step": 481
    },
    {
      "epoch": 0.6013724266999376,
      "grad_norm": 0.14612515735227102,
      "learning_rate": 9.337751170278495e-06,
      "loss": 0.0532,
      "num_tokens": 36757176.0,
      "step": 482
    },
    {
      "epoch": 0.6026200873362445,
      "grad_norm": 0.16129333056420253,
      "learning_rate": 9.334583446032866e-06,
      "loss": 0.0528,
      "num_tokens": 36834365.0,
      "step": 483
    },
    {
      "epoch": 0.6038677479725515,
      "grad_norm": 0.14048550830562176,
      "learning_rate": 9.331408768528423e-06,
      "loss": 0.0505,
      "num_tokens": 36911009.0,
      "step": 484
    },
    {
      "epoch": 0.6051154086088584,
      "grad_norm": 0.1401119811356599,
      "learning_rate": 9.328227143521809e-06,
      "loss": 0.0515,
      "num_tokens": 36986609.0,
      "step": 485
    },
    {
      "epoch": 0.6063630692451654,
      "grad_norm": 0.1393551599655052,
      "learning_rate": 9.325038576782275e-06,
      "loss": 0.0528,
      "num_tokens": 37063346.0,
      "step": 486
    },
    {
      "epoch": 0.6076107298814722,
      "grad_norm": 0.14547884377830422,
      "learning_rate": 9.321843074091654e-06,
      "loss": 0.0516,
      "num_tokens": 37140308.0,
      "step": 487
    },
    {
      "epoch": 0.6088583905177791,
      "grad_norm": 0.15034063695478633,
      "learning_rate": 9.318640641244362e-06,
      "loss": 0.0546,
      "num_tokens": 37216032.0,
      "step": 488
    },
    {
      "epoch": 0.6101060511540861,
      "grad_norm": 0.13651732647664802,
      "learning_rate": 9.315431284047375e-06,
      "loss": 0.0485,
      "num_tokens": 37291131.0,
      "step": 489
    },
    {
      "epoch": 0.611353711790393,
      "grad_norm": 0.13875747516393921,
      "learning_rate": 9.312215008320228e-06,
      "loss": 0.0544,
      "num_tokens": 37366741.0,
      "step": 490
    },
    {
      "epoch": 0.6126013724267,
      "grad_norm": 0.15178182686688055,
      "learning_rate": 9.308991819895001e-06,
      "loss": 0.0482,
      "num_tokens": 37441390.0,
      "step": 491
    },
    {
      "epoch": 0.6138490330630069,
      "grad_norm": 0.15247440606751309,
      "learning_rate": 9.30576172461631e-06,
      "loss": 0.0525,
      "num_tokens": 37518266.0,
      "step": 492
    },
    {
      "epoch": 0.6150966936993137,
      "grad_norm": 0.14750601213608294,
      "learning_rate": 9.302524728341292e-06,
      "loss": 0.0535,
      "num_tokens": 37593756.0,
      "step": 493
    },
    {
      "epoch": 0.6163443543356207,
      "grad_norm": 0.15128573856082372,
      "learning_rate": 9.299280836939602e-06,
      "loss": 0.0511,
      "num_tokens": 37670443.0,
      "step": 494
    },
    {
      "epoch": 0.6175920149719276,
      "grad_norm": 0.14776419777529295,
      "learning_rate": 9.296030056293394e-06,
      "loss": 0.0534,
      "num_tokens": 37746034.0,
      "step": 495
    },
    {
      "epoch": 0.6188396756082346,
      "grad_norm": 0.13724958742040025,
      "learning_rate": 9.292772392297316e-06,
      "loss": 0.0498,
      "num_tokens": 37821112.0,
      "step": 496
    },
    {
      "epoch": 0.6200873362445415,
      "grad_norm": 0.1465443892025746,
      "learning_rate": 9.289507850858498e-06,
      "loss": 0.0529,
      "num_tokens": 37898299.0,
      "step": 497
    },
    {
      "epoch": 0.6213349968808484,
      "grad_norm": 0.13641348218349508,
      "learning_rate": 9.286236437896538e-06,
      "loss": 0.0514,
      "num_tokens": 37974347.0,
      "step": 498
    },
    {
      "epoch": 0.6225826575171554,
      "grad_norm": 0.14507500345532376,
      "learning_rate": 9.282958159343502e-06,
      "loss": 0.0501,
      "num_tokens": 38050616.0,
      "step": 499
    },
    {
      "epoch": 0.6238303181534622,
      "grad_norm": 0.15639183855577438,
      "learning_rate": 9.279673021143895e-06,
      "loss": 0.0522,
      "num_tokens": 38127554.0,
      "step": 500
    },
    {
      "epoch": 0.6250779787897692,
      "grad_norm": 0.14274508358776897,
      "learning_rate": 9.276381029254668e-06,
      "loss": 0.0511,
      "num_tokens": 38203466.0,
      "step": 501
    },
    {
      "epoch": 0.6263256394260761,
      "grad_norm": 0.157102895207002,
      "learning_rate": 9.273082189645197e-06,
      "loss": 0.049,
      "num_tokens": 38278972.0,
      "step": 502
    },
    {
      "epoch": 0.627573300062383,
      "grad_norm": 0.16591536991200595,
      "learning_rate": 9.269776508297272e-06,
      "loss": 0.0484,
      "num_tokens": 38354884.0,
      "step": 503
    },
    {
      "epoch": 0.62882096069869,
      "grad_norm": 0.1422390165622962,
      "learning_rate": 9.266463991205096e-06,
      "loss": 0.0505,
      "num_tokens": 38430978.0,
      "step": 504
    },
    {
      "epoch": 0.6300686213349969,
      "grad_norm": 0.15110469574389213,
      "learning_rate": 9.263144644375264e-06,
      "loss": 0.0564,
      "num_tokens": 38506717.0,
      "step": 505
    },
    {
      "epoch": 0.6313162819713038,
      "grad_norm": 0.14145594461673772,
      "learning_rate": 9.259818473826753e-06,
      "loss": 0.0522,
      "num_tokens": 38582821.0,
      "step": 506
    },
    {
      "epoch": 0.6325639426076107,
      "grad_norm": 0.1433778713312024,
      "learning_rate": 9.256485485590916e-06,
      "loss": 0.0514,
      "num_tokens": 38658439.0,
      "step": 507
    },
    {
      "epoch": 0.6338116032439176,
      "grad_norm": 0.14504254404583986,
      "learning_rate": 9.25314568571147e-06,
      "loss": 0.0496,
      "num_tokens": 38734314.0,
      "step": 508
    },
    {
      "epoch": 0.6350592638802246,
      "grad_norm": 0.16233069854732574,
      "learning_rate": 9.24979908024448e-06,
      "loss": 0.0533,
      "num_tokens": 38810766.0,
      "step": 509
    },
    {
      "epoch": 0.6363069245165315,
      "grad_norm": 0.13653654499950715,
      "learning_rate": 9.246445675258353e-06,
      "loss": 0.0485,
      "num_tokens": 38886088.0,
      "step": 510
    },
    {
      "epoch": 0.6375545851528385,
      "grad_norm": 0.1446725493124663,
      "learning_rate": 9.243085476833827e-06,
      "loss": 0.0511,
      "num_tokens": 38963707.0,
      "step": 511
    },
    {
      "epoch": 0.6388022457891454,
      "grad_norm": 0.15005179375811173,
      "learning_rate": 9.239718491063956e-06,
      "loss": 0.0532,
      "num_tokens": 39039002.0,
      "step": 512
    },
    {
      "epoch": 0.6400499064254522,
      "grad_norm": 0.14813266791028454,
      "learning_rate": 9.236344724054104e-06,
      "loss": 0.0555,
      "num_tokens": 39115410.0,
      "step": 513
    },
    {
      "epoch": 0.6412975670617592,
      "grad_norm": 0.13508668746062624,
      "learning_rate": 9.232964181921931e-06,
      "loss": 0.0485,
      "num_tokens": 39190976.0,
      "step": 514
    },
    {
      "epoch": 0.6425452276980661,
      "grad_norm": 0.137197292476883,
      "learning_rate": 9.22957687079738e-06,
      "loss": 0.0551,
      "num_tokens": 39268014.0,
      "step": 515
    },
    {
      "epoch": 0.6437928883343731,
      "grad_norm": 0.14185888406488098,
      "learning_rate": 9.22618279682267e-06,
      "loss": 0.0538,
      "num_tokens": 39343390.0,
      "step": 516
    },
    {
      "epoch": 0.64504054897068,
      "grad_norm": 0.15637160791998272,
      "learning_rate": 9.222781966152284e-06,
      "loss": 0.0546,
      "num_tokens": 39419891.0,
      "step": 517
    },
    {
      "epoch": 0.6462882096069869,
      "grad_norm": 0.15136329659724634,
      "learning_rate": 9.219374384952955e-06,
      "loss": 0.0518,
      "num_tokens": 39496489.0,
      "step": 518
    },
    {
      "epoch": 0.6475358702432938,
      "grad_norm": 0.14065294824454894,
      "learning_rate": 9.215960059403657e-06,
      "loss": 0.0471,
      "num_tokens": 39572325.0,
      "step": 519
    },
    {
      "epoch": 0.6487835308796007,
      "grad_norm": 0.1500553366493087,
      "learning_rate": 9.212538995695597e-06,
      "loss": 0.053,
      "num_tokens": 39648939.0,
      "step": 520
    },
    {
      "epoch": 0.6500311915159077,
      "grad_norm": 0.142043093516072,
      "learning_rate": 9.209111200032197e-06,
      "loss": 0.0512,
      "num_tokens": 39724946.0,
      "step": 521
    },
    {
      "epoch": 0.6512788521522146,
      "grad_norm": 0.14617390218690907,
      "learning_rate": 9.205676678629084e-06,
      "loss": 0.0494,
      "num_tokens": 39800626.0,
      "step": 522
    },
    {
      "epoch": 0.6525265127885215,
      "grad_norm": 0.13478534135148082,
      "learning_rate": 9.202235437714085e-06,
      "loss": 0.0492,
      "num_tokens": 39876911.0,
      "step": 523
    },
    {
      "epoch": 0.6537741734248285,
      "grad_norm": 0.13717863504687297,
      "learning_rate": 9.198787483527211e-06,
      "loss": 0.0502,
      "num_tokens": 39953220.0,
      "step": 524
    },
    {
      "epoch": 0.6550218340611353,
      "grad_norm": 0.15373758129813972,
      "learning_rate": 9.195332822320643e-06,
      "loss": 0.0507,
      "num_tokens": 40028637.0,
      "step": 525
    },
    {
      "epoch": 0.6562694946974423,
      "grad_norm": 0.13761979854431663,
      "learning_rate": 9.191871460358727e-06,
      "loss": 0.0507,
      "num_tokens": 40104825.0,
      "step": 526
    },
    {
      "epoch": 0.6575171553337492,
      "grad_norm": 0.14149107473356692,
      "learning_rate": 9.188403403917959e-06,
      "loss": 0.0477,
      "num_tokens": 40180413.0,
      "step": 527
    },
    {
      "epoch": 0.6587648159700561,
      "grad_norm": 0.14081135806377892,
      "learning_rate": 9.184928659286972e-06,
      "loss": 0.0502,
      "num_tokens": 40257223.0,
      "step": 528
    },
    {
      "epoch": 0.6600124766063631,
      "grad_norm": 0.14440533546913695,
      "learning_rate": 9.181447232766531e-06,
      "loss": 0.0528,
      "num_tokens": 40333215.0,
      "step": 529
    },
    {
      "epoch": 0.66126013724267,
      "grad_norm": 0.14033132815670718,
      "learning_rate": 9.177959130669512e-06,
      "loss": 0.052,
      "num_tokens": 40410361.0,
      "step": 530
    },
    {
      "epoch": 0.662507797878977,
      "grad_norm": 0.14487287487973752,
      "learning_rate": 9.174464359320898e-06,
      "loss": 0.0505,
      "num_tokens": 40485973.0,
      "step": 531
    },
    {
      "epoch": 0.6637554585152838,
      "grad_norm": 0.14004587102147162,
      "learning_rate": 9.170962925057769e-06,
      "loss": 0.0509,
      "num_tokens": 40561925.0,
      "step": 532
    },
    {
      "epoch": 0.6650031191515907,
      "grad_norm": 0.13979887669681615,
      "learning_rate": 9.167454834229281e-06,
      "loss": 0.0525,
      "num_tokens": 40640581.0,
      "step": 533
    },
    {
      "epoch": 0.6662507797878977,
      "grad_norm": 0.13186354100825973,
      "learning_rate": 9.163940093196663e-06,
      "loss": 0.0476,
      "num_tokens": 40716574.0,
      "step": 534
    },
    {
      "epoch": 0.6674984404242046,
      "grad_norm": 0.15358376736324988,
      "learning_rate": 9.160418708333203e-06,
      "loss": 0.0581,
      "num_tokens": 40793125.0,
      "step": 535
    },
    {
      "epoch": 0.6687461010605116,
      "grad_norm": 0.14215032384807272,
      "learning_rate": 9.156890686024239e-06,
      "loss": 0.051,
      "num_tokens": 40869118.0,
      "step": 536
    },
    {
      "epoch": 0.6699937616968185,
      "grad_norm": 0.14579188342681432,
      "learning_rate": 9.153356032667138e-06,
      "loss": 0.0495,
      "num_tokens": 40946005.0,
      "step": 537
    },
    {
      "epoch": 0.6712414223331253,
      "grad_norm": 0.1392706758978059,
      "learning_rate": 9.149814754671296e-06,
      "loss": 0.0539,
      "num_tokens": 41023290.0,
      "step": 538
    },
    {
      "epoch": 0.6724890829694323,
      "grad_norm": 0.13919076738970143,
      "learning_rate": 9.14626685845812e-06,
      "loss": 0.0507,
      "num_tokens": 41100482.0,
      "step": 539
    },
    {
      "epoch": 0.6737367436057392,
      "grad_norm": 0.14007643948823942,
      "learning_rate": 9.142712350461021e-06,
      "loss": 0.0482,
      "num_tokens": 41175700.0,
      "step": 540
    },
    {
      "epoch": 0.6749844042420462,
      "grad_norm": 0.142886288926492,
      "learning_rate": 9.139151237125393e-06,
      "loss": 0.0517,
      "num_tokens": 41251124.0,
      "step": 541
    },
    {
      "epoch": 0.6762320648783531,
      "grad_norm": 0.1328562647414474,
      "learning_rate": 9.135583524908614e-06,
      "loss": 0.0498,
      "num_tokens": 41328351.0,
      "step": 542
    },
    {
      "epoch": 0.67747972551466,
      "grad_norm": 0.14304761262946025,
      "learning_rate": 9.132009220280021e-06,
      "loss": 0.0545,
      "num_tokens": 41404625.0,
      "step": 543
    },
    {
      "epoch": 0.678727386150967,
      "grad_norm": 0.14231842272684792,
      "learning_rate": 9.128428329720911e-06,
      "loss": 0.052,
      "num_tokens": 41480638.0,
      "step": 544
    },
    {
      "epoch": 0.6799750467872738,
      "grad_norm": 0.14373073570586645,
      "learning_rate": 9.12484085972452e-06,
      "loss": 0.049,
      "num_tokens": 41556512.0,
      "step": 545
    },
    {
      "epoch": 0.6812227074235808,
      "grad_norm": 0.1418611549255363,
      "learning_rate": 9.121246816796017e-06,
      "loss": 0.0468,
      "num_tokens": 41633777.0,
      "step": 546
    },
    {
      "epoch": 0.6824703680598877,
      "grad_norm": 0.14393902575169928,
      "learning_rate": 9.117646207452487e-06,
      "loss": 0.0551,
      "num_tokens": 41710870.0,
      "step": 547
    },
    {
      "epoch": 0.6837180286961946,
      "grad_norm": 0.128685650110848,
      "learning_rate": 9.114039038222922e-06,
      "loss": 0.0505,
      "num_tokens": 41787983.0,
      "step": 548
    },
    {
      "epoch": 0.6849656893325016,
      "grad_norm": 0.14694897850884,
      "learning_rate": 9.110425315648212e-06,
      "loss": 0.0518,
      "num_tokens": 41863855.0,
      "step": 549
    },
    {
      "epoch": 0.6862133499688085,
      "grad_norm": 0.1326421204130351,
      "learning_rate": 9.106805046281127e-06,
      "loss": 0.0482,
      "num_tokens": 41940301.0,
      "step": 550
    },
    {
      "epoch": 0.6874610106051154,
      "grad_norm": 0.13892809572479203,
      "learning_rate": 9.103178236686309e-06,
      "loss": 0.0503,
      "num_tokens": 42015296.0,
      "step": 551
    },
    {
      "epoch": 0.6887086712414223,
      "grad_norm": 0.138296019206815,
      "learning_rate": 9.099544893440265e-06,
      "loss": 0.0477,
      "num_tokens": 42090560.0,
      "step": 552
    },
    {
      "epoch": 0.6899563318777293,
      "grad_norm": 0.13688585206848972,
      "learning_rate": 9.095905023131337e-06,
      "loss": 0.051,
      "num_tokens": 42166613.0,
      "step": 553
    },
    {
      "epoch": 0.6912039925140362,
      "grad_norm": 0.14866506305162613,
      "learning_rate": 9.092258632359714e-06,
      "loss": 0.0518,
      "num_tokens": 42243060.0,
      "step": 554
    },
    {
      "epoch": 0.6924516531503431,
      "grad_norm": 0.13540311621936235,
      "learning_rate": 9.088605727737405e-06,
      "loss": 0.0499,
      "num_tokens": 42319484.0,
      "step": 555
    },
    {
      "epoch": 0.6936993137866501,
      "grad_norm": 0.13923746025516218,
      "learning_rate": 9.08494631588823e-06,
      "loss": 0.0517,
      "num_tokens": 42396278.0,
      "step": 556
    },
    {
      "epoch": 0.6949469744229569,
      "grad_norm": 0.1305168808594344,
      "learning_rate": 9.08128040344781e-06,
      "loss": 0.0461,
      "num_tokens": 42471074.0,
      "step": 557
    },
    {
      "epoch": 0.6961946350592639,
      "grad_norm": 0.13963165538464697,
      "learning_rate": 9.077607997063546e-06,
      "loss": 0.0496,
      "num_tokens": 42546608.0,
      "step": 558
    },
    {
      "epoch": 0.6974422956955708,
      "grad_norm": 0.14171018648487066,
      "learning_rate": 9.073929103394627e-06,
      "loss": 0.0511,
      "num_tokens": 42622791.0,
      "step": 559
    },
    {
      "epoch": 0.6986899563318777,
      "grad_norm": 0.1482616284034211,
      "learning_rate": 9.070243729111998e-06,
      "loss": 0.0536,
      "num_tokens": 42698266.0,
      "step": 560
    },
    {
      "epoch": 0.6999376169681847,
      "grad_norm": 0.1389694479982638,
      "learning_rate": 9.066551880898356e-06,
      "loss": 0.0502,
      "num_tokens": 42775035.0,
      "step": 561
    },
    {
      "epoch": 0.7011852776044916,
      "grad_norm": 0.14647295565403837,
      "learning_rate": 9.062853565448137e-06,
      "loss": 0.0508,
      "num_tokens": 42851775.0,
      "step": 562
    },
    {
      "epoch": 0.7024329382407986,
      "grad_norm": 0.14191889661269946,
      "learning_rate": 9.059148789467508e-06,
      "loss": 0.0531,
      "num_tokens": 42928574.0,
      "step": 563
    },
    {
      "epoch": 0.7036805988771054,
      "grad_norm": 0.1483489925428368,
      "learning_rate": 9.055437559674343e-06,
      "loss": 0.0518,
      "num_tokens": 43005534.0,
      "step": 564
    },
    {
      "epoch": 0.7049282595134123,
      "grad_norm": 0.14255136898330478,
      "learning_rate": 9.051719882798226e-06,
      "loss": 0.049,
      "num_tokens": 43081297.0,
      "step": 565
    },
    {
      "epoch": 0.7061759201497193,
      "grad_norm": 0.13615790668479794,
      "learning_rate": 9.047995765580428e-06,
      "loss": 0.0474,
      "num_tokens": 43157741.0,
      "step": 566
    },
    {
      "epoch": 0.7074235807860262,
      "grad_norm": 0.1348869319148483,
      "learning_rate": 9.044265214773901e-06,
      "loss": 0.0487,
      "num_tokens": 43234575.0,
      "step": 567
    },
    {
      "epoch": 0.7086712414223332,
      "grad_norm": 0.1345129640129995,
      "learning_rate": 9.040528237143258e-06,
      "loss": 0.0519,
      "num_tokens": 43311435.0,
      "step": 568
    },
    {
      "epoch": 0.7099189020586401,
      "grad_norm": 0.1424294334295593,
      "learning_rate": 9.036784839464771e-06,
      "loss": 0.0507,
      "num_tokens": 43388039.0,
      "step": 569
    },
    {
      "epoch": 0.7111665626949469,
      "grad_norm": 0.13921104210577145,
      "learning_rate": 9.033035028526352e-06,
      "loss": 0.0478,
      "num_tokens": 43463609.0,
      "step": 570
    },
    {
      "epoch": 0.7124142233312539,
      "grad_norm": 0.13815837263264233,
      "learning_rate": 9.029278811127539e-06,
      "loss": 0.0482,
      "num_tokens": 43540358.0,
      "step": 571
    },
    {
      "epoch": 0.7136618839675608,
      "grad_norm": 0.14661609406068885,
      "learning_rate": 9.025516194079493e-06,
      "loss": 0.0527,
      "num_tokens": 43616703.0,
      "step": 572
    },
    {
      "epoch": 0.7149095446038678,
      "grad_norm": 0.1420210259390237,
      "learning_rate": 9.021747184204974e-06,
      "loss": 0.0482,
      "num_tokens": 43692473.0,
      "step": 573
    },
    {
      "epoch": 0.7161572052401747,
      "grad_norm": 0.15807055160763936,
      "learning_rate": 9.017971788338338e-06,
      "loss": 0.0572,
      "num_tokens": 43770031.0,
      "step": 574
    },
    {
      "epoch": 0.7174048658764816,
      "grad_norm": 0.14478525891433547,
      "learning_rate": 9.014190013325514e-06,
      "loss": 0.0516,
      "num_tokens": 43846662.0,
      "step": 575
    },
    {
      "epoch": 0.7186525265127885,
      "grad_norm": 0.14868363097251575,
      "learning_rate": 9.010401866024007e-06,
      "loss": 0.0492,
      "num_tokens": 43922348.0,
      "step": 576
    },
    {
      "epoch": 0.7199001871490954,
      "grad_norm": 0.13523798397269673,
      "learning_rate": 9.006607353302874e-06,
      "loss": 0.0478,
      "num_tokens": 43997815.0,
      "step": 577
    },
    {
      "epoch": 0.7211478477854024,
      "grad_norm": 0.1322796061870224,
      "learning_rate": 9.00280648204271e-06,
      "loss": 0.0449,
      "num_tokens": 44074133.0,
      "step": 578
    },
    {
      "epoch": 0.7223955084217093,
      "grad_norm": 0.13851692878820665,
      "learning_rate": 8.998999259135648e-06,
      "loss": 0.0474,
      "num_tokens": 44148719.0,
      "step": 579
    },
    {
      "epoch": 0.7236431690580162,
      "grad_norm": 0.17724864437281473,
      "learning_rate": 8.99518569148533e-06,
      "loss": 0.0513,
      "num_tokens": 44224599.0,
      "step": 580
    },
    {
      "epoch": 0.7248908296943232,
      "grad_norm": 0.15309429572852976,
      "learning_rate": 8.991365786006908e-06,
      "loss": 0.0529,
      "num_tokens": 44300530.0,
      "step": 581
    },
    {
      "epoch": 0.72613849033063,
      "grad_norm": 0.14474973461404464,
      "learning_rate": 8.987539549627026e-06,
      "loss": 0.0526,
      "num_tokens": 44377954.0,
      "step": 582
    },
    {
      "epoch": 0.727386150966937,
      "grad_norm": 0.14258419787145118,
      "learning_rate": 8.983706989283804e-06,
      "loss": 0.0474,
      "num_tokens": 44453567.0,
      "step": 583
    },
    {
      "epoch": 0.7286338116032439,
      "grad_norm": 0.13275146203069824,
      "learning_rate": 8.979868111926836e-06,
      "loss": 0.0471,
      "num_tokens": 44529975.0,
      "step": 584
    },
    {
      "epoch": 0.7298814722395508,
      "grad_norm": 0.14399620167955793,
      "learning_rate": 8.976022924517167e-06,
      "loss": 0.0516,
      "num_tokens": 44607134.0,
      "step": 585
    },
    {
      "epoch": 0.7311291328758578,
      "grad_norm": 0.14627346228999338,
      "learning_rate": 8.972171434027283e-06,
      "loss": 0.0519,
      "num_tokens": 44683151.0,
      "step": 586
    },
    {
      "epoch": 0.7323767935121647,
      "grad_norm": 0.14811956571593604,
      "learning_rate": 8.968313647441098e-06,
      "loss": 0.055,
      "num_tokens": 44759260.0,
      "step": 587
    },
    {
      "epoch": 0.7336244541484717,
      "grad_norm": 0.13835530298124146,
      "learning_rate": 8.964449571753949e-06,
      "loss": 0.05,
      "num_tokens": 44835552.0,
      "step": 588
    },
    {
      "epoch": 0.7348721147847785,
      "grad_norm": 0.1341138944001767,
      "learning_rate": 8.96057921397257e-06,
      "loss": 0.0504,
      "num_tokens": 44912366.0,
      "step": 589
    },
    {
      "epoch": 0.7361197754210854,
      "grad_norm": 0.1495197595834141,
      "learning_rate": 8.95670258111509e-06,
      "loss": 0.053,
      "num_tokens": 44990082.0,
      "step": 590
    },
    {
      "epoch": 0.7373674360573924,
      "grad_norm": 0.14011781606379067,
      "learning_rate": 8.95281968021102e-06,
      "loss": 0.0481,
      "num_tokens": 45065731.0,
      "step": 591
    },
    {
      "epoch": 0.7386150966936993,
      "grad_norm": 0.14888004483940512,
      "learning_rate": 8.948930518301228e-06,
      "loss": 0.0546,
      "num_tokens": 45143212.0,
      "step": 592
    },
    {
      "epoch": 0.7398627573300063,
      "grad_norm": 0.1434778875987488,
      "learning_rate": 8.945035102437943e-06,
      "loss": 0.0533,
      "num_tokens": 45219875.0,
      "step": 593
    },
    {
      "epoch": 0.7411104179663132,
      "grad_norm": 0.1382641241537832,
      "learning_rate": 8.94113343968473e-06,
      "loss": 0.0525,
      "num_tokens": 45297056.0,
      "step": 594
    },
    {
      "epoch": 0.74235807860262,
      "grad_norm": 0.14167216666261945,
      "learning_rate": 8.937225537116482e-06,
      "loss": 0.0537,
      "num_tokens": 45374106.0,
      "step": 595
    },
    {
      "epoch": 0.743605739238927,
      "grad_norm": 0.1438734355396013,
      "learning_rate": 8.93331140181941e-06,
      "loss": 0.0524,
      "num_tokens": 45449504.0,
      "step": 596
    },
    {
      "epoch": 0.7448533998752339,
      "grad_norm": 0.14488167117333944,
      "learning_rate": 8.929391040891022e-06,
      "loss": 0.0493,
      "num_tokens": 45526160.0,
      "step": 597
    },
    {
      "epoch": 0.7461010605115409,
      "grad_norm": 0.13888422862308172,
      "learning_rate": 8.92546446144012e-06,
      "loss": 0.0462,
      "num_tokens": 45601707.0,
      "step": 598
    },
    {
      "epoch": 0.7473487211478478,
      "grad_norm": 0.14288450463132527,
      "learning_rate": 8.921531670586778e-06,
      "loss": 0.0582,
      "num_tokens": 45678093.0,
      "step": 599
    },
    {
      "epoch": 0.7485963817841547,
      "grad_norm": 0.134670825844891,
      "learning_rate": 8.917592675462333e-06,
      "loss": 0.05,
      "num_tokens": 45754558.0,
      "step": 600
    },
    {
      "epoch": 0.7498440424204617,
      "grad_norm": 0.136217251833068,
      "learning_rate": 8.913647483209376e-06,
      "loss": 0.0477,
      "num_tokens": 45830631.0,
      "step": 601
    },
    {
      "epoch": 0.7510917030567685,
      "grad_norm": 0.1452373399840719,
      "learning_rate": 8.909696100981734e-06,
      "loss": 0.0543,
      "num_tokens": 45908434.0,
      "step": 602
    },
    {
      "epoch": 0.7523393636930755,
      "grad_norm": 0.14921298014946266,
      "learning_rate": 8.905738535944453e-06,
      "loss": 0.0485,
      "num_tokens": 45984136.0,
      "step": 603
    },
    {
      "epoch": 0.7535870243293824,
      "grad_norm": 0.14271019380781036,
      "learning_rate": 8.901774795273799e-06,
      "loss": 0.0503,
      "num_tokens": 46060807.0,
      "step": 604
    },
    {
      "epoch": 0.7548346849656893,
      "grad_norm": 0.1374279559284211,
      "learning_rate": 8.897804886157229e-06,
      "loss": 0.0487,
      "num_tokens": 46136263.0,
      "step": 605
    },
    {
      "epoch": 0.7560823456019963,
      "grad_norm": 0.1275634793402176,
      "learning_rate": 8.893828815793389e-06,
      "loss": 0.0481,
      "num_tokens": 46212176.0,
      "step": 606
    },
    {
      "epoch": 0.7573300062383032,
      "grad_norm": 0.13808428299635023,
      "learning_rate": 8.889846591392097e-06,
      "loss": 0.0485,
      "num_tokens": 46287974.0,
      "step": 607
    },
    {
      "epoch": 0.7585776668746101,
      "grad_norm": 0.12657074289437703,
      "learning_rate": 8.88585822017433e-06,
      "loss": 0.048,
      "num_tokens": 46364270.0,
      "step": 608
    },
    {
      "epoch": 0.759825327510917,
      "grad_norm": 0.1286073078080098,
      "learning_rate": 8.881863709372207e-06,
      "loss": 0.045,
      "num_tokens": 46441185.0,
      "step": 609
    },
    {
      "epoch": 0.7610729881472239,
      "grad_norm": 0.15397840232569265,
      "learning_rate": 8.877863066228987e-06,
      "loss": 0.0463,
      "num_tokens": 46516606.0,
      "step": 610
    },
    {
      "epoch": 0.7623206487835309,
      "grad_norm": 0.14476818404924205,
      "learning_rate": 8.873856297999045e-06,
      "loss": 0.0493,
      "num_tokens": 46592201.0,
      "step": 611
    },
    {
      "epoch": 0.7635683094198378,
      "grad_norm": 0.1405957000571124,
      "learning_rate": 8.869843411947862e-06,
      "loss": 0.0476,
      "num_tokens": 46668676.0,
      "step": 612
    },
    {
      "epoch": 0.7648159700561448,
      "grad_norm": 0.13555365965845964,
      "learning_rate": 8.865824415352014e-06,
      "loss": 0.0478,
      "num_tokens": 46744569.0,
      "step": 613
    },
    {
      "epoch": 0.7660636306924516,
      "grad_norm": 0.1469494770314817,
      "learning_rate": 8.861799315499157e-06,
      "loss": 0.0478,
      "num_tokens": 46820151.0,
      "step": 614
    },
    {
      "epoch": 0.7673112913287585,
      "grad_norm": 0.12973149282283644,
      "learning_rate": 8.85776811968801e-06,
      "loss": 0.0473,
      "num_tokens": 46895382.0,
      "step": 615
    },
    {
      "epoch": 0.7685589519650655,
      "grad_norm": 0.1378044671519852,
      "learning_rate": 8.853730835228354e-06,
      "loss": 0.0531,
      "num_tokens": 46971400.0,
      "step": 616
    },
    {
      "epoch": 0.7698066126013724,
      "grad_norm": 0.14644478526342583,
      "learning_rate": 8.849687469441003e-06,
      "loss": 0.0531,
      "num_tokens": 47048651.0,
      "step": 617
    },
    {
      "epoch": 0.7710542732376794,
      "grad_norm": 0.14707076641241534,
      "learning_rate": 8.845638029657804e-06,
      "loss": 0.0468,
      "num_tokens": 47124698.0,
      "step": 618
    },
    {
      "epoch": 0.7723019338739863,
      "grad_norm": 0.1314006053285289,
      "learning_rate": 8.841582523221614e-06,
      "loss": 0.0517,
      "num_tokens": 47201140.0,
      "step": 619
    },
    {
      "epoch": 0.7735495945102931,
      "grad_norm": 0.14004488885132432,
      "learning_rate": 8.83752095748629e-06,
      "loss": 0.0489,
      "num_tokens": 47277704.0,
      "step": 620
    },
    {
      "epoch": 0.7747972551466001,
      "grad_norm": 0.1468794488076061,
      "learning_rate": 8.833453339816682e-06,
      "loss": 0.0501,
      "num_tokens": 47353136.0,
      "step": 621
    },
    {
      "epoch": 0.776044915782907,
      "grad_norm": 0.13530186455081034,
      "learning_rate": 8.829379677588607e-06,
      "loss": 0.0451,
      "num_tokens": 47428571.0,
      "step": 622
    },
    {
      "epoch": 0.777292576419214,
      "grad_norm": 0.13276087320331062,
      "learning_rate": 8.825299978188847e-06,
      "loss": 0.0468,
      "num_tokens": 47504497.0,
      "step": 623
    },
    {
      "epoch": 0.7785402370555209,
      "grad_norm": 0.14105954588118347,
      "learning_rate": 8.821214249015133e-06,
      "loss": 0.052,
      "num_tokens": 47580142.0,
      "step": 624
    },
    {
      "epoch": 0.7797878976918278,
      "grad_norm": 0.13687964960635024,
      "learning_rate": 8.817122497476122e-06,
      "loss": 0.0463,
      "num_tokens": 47655481.0,
      "step": 625
    },
    {
      "epoch": 0.7810355583281348,
      "grad_norm": 0.12988897107395347,
      "learning_rate": 8.8130247309914e-06,
      "loss": 0.0457,
      "num_tokens": 47730729.0,
      "step": 626
    },
    {
      "epoch": 0.7822832189644416,
      "grad_norm": 0.14026915526349465,
      "learning_rate": 8.808920956991455e-06,
      "loss": 0.051,
      "num_tokens": 47807189.0,
      "step": 627
    },
    {
      "epoch": 0.7835308796007486,
      "grad_norm": 0.142295896643215,
      "learning_rate": 8.80481118291767e-06,
      "loss": 0.0487,
      "num_tokens": 47882622.0,
      "step": 628
    },
    {
      "epoch": 0.7847785402370555,
      "grad_norm": 0.1370740658768669,
      "learning_rate": 8.800695416222305e-06,
      "loss": 0.0495,
      "num_tokens": 47958319.0,
      "step": 629
    },
    {
      "epoch": 0.7860262008733624,
      "grad_norm": 0.13452523773803834,
      "learning_rate": 8.796573664368492e-06,
      "loss": 0.0483,
      "num_tokens": 48034149.0,
      "step": 630
    },
    {
      "epoch": 0.7872738615096694,
      "grad_norm": 0.14079938056246274,
      "learning_rate": 8.792445934830215e-06,
      "loss": 0.0478,
      "num_tokens": 48110857.0,
      "step": 631
    },
    {
      "epoch": 0.7885215221459763,
      "grad_norm": 0.14941160855473518,
      "learning_rate": 8.78831223509229e-06,
      "loss": 0.0509,
      "num_tokens": 48186687.0,
      "step": 632
    },
    {
      "epoch": 0.7897691827822833,
      "grad_norm": 0.12938950186652354,
      "learning_rate": 8.784172572650366e-06,
      "loss": 0.0462,
      "num_tokens": 48262209.0,
      "step": 633
    },
    {
      "epoch": 0.7910168434185901,
      "grad_norm": 0.13822393920620288,
      "learning_rate": 8.780026955010903e-06,
      "loss": 0.0499,
      "num_tokens": 48338358.0,
      "step": 634
    },
    {
      "epoch": 0.7922645040548971,
      "grad_norm": 0.14421310117588615,
      "learning_rate": 8.77587538969116e-06,
      "loss": 0.0492,
      "num_tokens": 48414450.0,
      "step": 635
    },
    {
      "epoch": 0.793512164691204,
      "grad_norm": 0.15091469224096057,
      "learning_rate": 8.771717884219177e-06,
      "loss": 0.0499,
      "num_tokens": 48490141.0,
      "step": 636
    },
    {
      "epoch": 0.7947598253275109,
      "grad_norm": 0.14201341470304255,
      "learning_rate": 8.767554446133771e-06,
      "loss": 0.0529,
      "num_tokens": 48566985.0,
      "step": 637
    },
    {
      "epoch": 0.7960074859638179,
      "grad_norm": 0.1286122639124437,
      "learning_rate": 8.763385082984511e-06,
      "loss": 0.0448,
      "num_tokens": 48641855.0,
      "step": 638
    },
    {
      "epoch": 0.7972551466001248,
      "grad_norm": 0.1548475688653674,
      "learning_rate": 8.759209802331714e-06,
      "loss": 0.0514,
      "num_tokens": 48717988.0,
      "step": 639
    },
    {
      "epoch": 0.7985028072364317,
      "grad_norm": 0.12966093282997393,
      "learning_rate": 8.755028611746426e-06,
      "loss": 0.0476,
      "num_tokens": 48794607.0,
      "step": 640
    },
    {
      "epoch": 0.7997504678727386,
      "grad_norm": 0.13188968216215208,
      "learning_rate": 8.750841518810407e-06,
      "loss": 0.0481,
      "num_tokens": 48871059.0,
      "step": 641
    },
    {
      "epoch": 0.8009981285090455,
      "grad_norm": 0.13664684497120977,
      "learning_rate": 8.746648531116126e-06,
      "loss": 0.0446,
      "num_tokens": 48948505.0,
      "step": 642
    },
    {
      "epoch": 0.8022457891453525,
      "grad_norm": 0.14317884582014548,
      "learning_rate": 8.742449656266733e-06,
      "loss": 0.0504,
      "num_tokens": 49024812.0,
      "step": 643
    },
    {
      "epoch": 0.8034934497816594,
      "grad_norm": 0.1347554523074578,
      "learning_rate": 8.738244901876061e-06,
      "loss": 0.0473,
      "num_tokens": 49101044.0,
      "step": 644
    },
    {
      "epoch": 0.8047411104179664,
      "grad_norm": 0.13636883319815177,
      "learning_rate": 8.7340342755686e-06,
      "loss": 0.0467,
      "num_tokens": 49177145.0,
      "step": 645
    },
    {
      "epoch": 0.8059887710542732,
      "grad_norm": 0.1275910557926321,
      "learning_rate": 8.729817784979485e-06,
      "loss": 0.0495,
      "num_tokens": 49254377.0,
      "step": 646
    },
    {
      "epoch": 0.8072364316905801,
      "grad_norm": 0.14394843837797267,
      "learning_rate": 8.725595437754489e-06,
      "loss": 0.0519,
      "num_tokens": 49330881.0,
      "step": 647
    },
    {
      "epoch": 0.8084840923268871,
      "grad_norm": 0.1417831430199472,
      "learning_rate": 8.721367241550007e-06,
      "loss": 0.0477,
      "num_tokens": 49406788.0,
      "step": 648
    },
    {
      "epoch": 0.809731752963194,
      "grad_norm": 0.12920116740596488,
      "learning_rate": 8.717133204033034e-06,
      "loss": 0.0455,
      "num_tokens": 49482606.0,
      "step": 649
    },
    {
      "epoch": 0.810979413599501,
      "grad_norm": 0.14287675644499612,
      "learning_rate": 8.71289333288116e-06,
      "loss": 0.0498,
      "num_tokens": 49559132.0,
      "step": 650
    },
    {
      "epoch": 0.8122270742358079,
      "grad_norm": 0.14254507159222335,
      "learning_rate": 8.708647635782553e-06,
      "loss": 0.0484,
      "num_tokens": 49635261.0,
      "step": 651
    },
    {
      "epoch": 0.8134747348721147,
      "grad_norm": 0.1419100200993809,
      "learning_rate": 8.704396120435944e-06,
      "loss": 0.0537,
      "num_tokens": 49712659.0,
      "step": 652
    },
    {
      "epoch": 0.8147223955084217,
      "grad_norm": 0.12941282820277408,
      "learning_rate": 8.700138794550617e-06,
      "loss": 0.0491,
      "num_tokens": 49788525.0,
      "step": 653
    },
    {
      "epoch": 0.8159700561447286,
      "grad_norm": 0.13450455701515368,
      "learning_rate": 8.695875665846392e-06,
      "loss": 0.0497,
      "num_tokens": 49864781.0,
      "step": 654
    },
    {
      "epoch": 0.8172177167810356,
      "grad_norm": 0.13125571555637372,
      "learning_rate": 8.691606742053608e-06,
      "loss": 0.0479,
      "num_tokens": 49941093.0,
      "step": 655
    },
    {
      "epoch": 0.8184653774173425,
      "grad_norm": 0.14213099235813234,
      "learning_rate": 8.687332030913114e-06,
      "loss": 0.0467,
      "num_tokens": 50016325.0,
      "step": 656
    },
    {
      "epoch": 0.8197130380536494,
      "grad_norm": 0.1390710924527652,
      "learning_rate": 8.683051540176252e-06,
      "loss": 0.0495,
      "num_tokens": 50091945.0,
      "step": 657
    },
    {
      "epoch": 0.8209606986899564,
      "grad_norm": 0.12939775374003082,
      "learning_rate": 8.67876527760485e-06,
      "loss": 0.0452,
      "num_tokens": 50166912.0,
      "step": 658
    },
    {
      "epoch": 0.8222083593262632,
      "grad_norm": 0.1442490674566971,
      "learning_rate": 8.674473250971194e-06,
      "loss": 0.0495,
      "num_tokens": 50242363.0,
      "step": 659
    },
    {
      "epoch": 0.8234560199625702,
      "grad_norm": 0.12532738571426058,
      "learning_rate": 8.670175468058027e-06,
      "loss": 0.0455,
      "num_tokens": 50318669.0,
      "step": 660
    },
    {
      "epoch": 0.8247036805988771,
      "grad_norm": 0.1482963807294505,
      "learning_rate": 8.665871936658525e-06,
      "loss": 0.0515,
      "num_tokens": 50395068.0,
      "step": 661
    },
    {
      "epoch": 0.825951341235184,
      "grad_norm": 0.14537684974160384,
      "learning_rate": 8.661562664576297e-06,
      "loss": 0.0502,
      "num_tokens": 50471770.0,
      "step": 662
    },
    {
      "epoch": 0.827199001871491,
      "grad_norm": 0.14127973389090612,
      "learning_rate": 8.65724765962535e-06,
      "loss": 0.0512,
      "num_tokens": 50548248.0,
      "step": 663
    },
    {
      "epoch": 0.8284466625077979,
      "grad_norm": 0.13467898135586287,
      "learning_rate": 8.652926929630097e-06,
      "loss": 0.0455,
      "num_tokens": 50623214.0,
      "step": 664
    },
    {
      "epoch": 0.8296943231441049,
      "grad_norm": 0.1381037134217272,
      "learning_rate": 8.648600482425325e-06,
      "loss": 0.051,
      "num_tokens": 50699609.0,
      "step": 665
    },
    {
      "epoch": 0.8309419837804117,
      "grad_norm": 0.14193032487419316,
      "learning_rate": 8.644268325856193e-06,
      "loss": 0.0564,
      "num_tokens": 50778673.0,
      "step": 666
    },
    {
      "epoch": 0.8321896444167186,
      "grad_norm": 0.13611323916808238,
      "learning_rate": 8.639930467778206e-06,
      "loss": 0.0521,
      "num_tokens": 50855609.0,
      "step": 667
    },
    {
      "epoch": 0.8334373050530256,
      "grad_norm": 0.13531201611201343,
      "learning_rate": 8.635586916057214e-06,
      "loss": 0.0506,
      "num_tokens": 50931303.0,
      "step": 668
    },
    {
      "epoch": 0.8346849656893325,
      "grad_norm": 0.1303221088261824,
      "learning_rate": 8.631237678569391e-06,
      "loss": 0.0441,
      "num_tokens": 51006600.0,
      "step": 669
    },
    {
      "epoch": 0.8359326263256395,
      "grad_norm": 0.13622877089659338,
      "learning_rate": 8.626882763201215e-06,
      "loss": 0.0484,
      "num_tokens": 51082648.0,
      "step": 670
    },
    {
      "epoch": 0.8371802869619464,
      "grad_norm": 0.13249433552081497,
      "learning_rate": 8.62252217784947e-06,
      "loss": 0.0462,
      "num_tokens": 51158853.0,
      "step": 671
    },
    {
      "epoch": 0.8384279475982532,
      "grad_norm": 0.128012742550011,
      "learning_rate": 8.61815593042121e-06,
      "loss": 0.0465,
      "num_tokens": 51235008.0,
      "step": 672
    },
    {
      "epoch": 0.8396756082345602,
      "grad_norm": 0.13676039074631796,
      "learning_rate": 8.61378402883376e-06,
      "loss": 0.0484,
      "num_tokens": 51312341.0,
      "step": 673
    },
    {
      "epoch": 0.8409232688708671,
      "grad_norm": 0.13896442644172538,
      "learning_rate": 8.609406481014704e-06,
      "loss": 0.0538,
      "num_tokens": 51389560.0,
      "step": 674
    },
    {
      "epoch": 0.8421709295071741,
      "grad_norm": 0.13178732543595786,
      "learning_rate": 8.605023294901857e-06,
      "loss": 0.0492,
      "num_tokens": 51467043.0,
      "step": 675
    },
    {
      "epoch": 0.843418590143481,
      "grad_norm": 0.13530364899028144,
      "learning_rate": 8.600634478443262e-06,
      "loss": 0.0504,
      "num_tokens": 51544362.0,
      "step": 676
    },
    {
      "epoch": 0.8446662507797879,
      "grad_norm": 0.14451569461184943,
      "learning_rate": 8.596240039597168e-06,
      "loss": 0.049,
      "num_tokens": 51620148.0,
      "step": 677
    },
    {
      "epoch": 0.8459139114160948,
      "grad_norm": 0.12951759840418117,
      "learning_rate": 8.59183998633202e-06,
      "loss": 0.0474,
      "num_tokens": 51696229.0,
      "step": 678
    },
    {
      "epoch": 0.8471615720524017,
      "grad_norm": 0.13526546179585314,
      "learning_rate": 8.587434326626446e-06,
      "loss": 0.0484,
      "num_tokens": 51772533.0,
      "step": 679
    },
    {
      "epoch": 0.8484092326887087,
      "grad_norm": 0.12302067491500918,
      "learning_rate": 8.58302306846924e-06,
      "loss": 0.043,
      "num_tokens": 51848768.0,
      "step": 680
    },
    {
      "epoch": 0.8496568933250156,
      "grad_norm": 0.1457279157996279,
      "learning_rate": 8.57860621985934e-06,
      "loss": 0.0487,
      "num_tokens": 51925423.0,
      "step": 681
    },
    {
      "epoch": 0.8509045539613225,
      "grad_norm": 0.13073888393774447,
      "learning_rate": 8.574183788805838e-06,
      "loss": 0.0437,
      "num_tokens": 52001711.0,
      "step": 682
    },
    {
      "epoch": 0.8521522145976295,
      "grad_norm": 0.1416231531180916,
      "learning_rate": 8.56975578332793e-06,
      "loss": 0.0487,
      "num_tokens": 52078103.0,
      "step": 683
    },
    {
      "epoch": 0.8533998752339363,
      "grad_norm": 0.1443155543381733,
      "learning_rate": 8.56532221145493e-06,
      "loss": 0.0474,
      "num_tokens": 52154193.0,
      "step": 684
    },
    {
      "epoch": 0.8546475358702433,
      "grad_norm": 0.13662087765610023,
      "learning_rate": 8.560883081226246e-06,
      "loss": 0.0471,
      "num_tokens": 52230298.0,
      "step": 685
    },
    {
      "epoch": 0.8558951965065502,
      "grad_norm": 0.12505249076114922,
      "learning_rate": 8.55643840069136e-06,
      "loss": 0.0467,
      "num_tokens": 52306515.0,
      "step": 686
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.1310843454403954,
      "learning_rate": 8.551988177909825e-06,
      "loss": 0.0489,
      "num_tokens": 52383062.0,
      "step": 687
    },
    {
      "epoch": 0.8583905177791641,
      "grad_norm": 0.1320466168637091,
      "learning_rate": 8.547532420951236e-06,
      "loss": 0.0442,
      "num_tokens": 52458704.0,
      "step": 688
    },
    {
      "epoch": 0.859638178415471,
      "grad_norm": 0.13717646106028553,
      "learning_rate": 8.543071137895231e-06,
      "loss": 0.0489,
      "num_tokens": 52536401.0,
      "step": 689
    },
    {
      "epoch": 0.860885839051778,
      "grad_norm": 0.13096083678794693,
      "learning_rate": 8.538604336831463e-06,
      "loss": 0.0477,
      "num_tokens": 52612783.0,
      "step": 690
    },
    {
      "epoch": 0.8621334996880848,
      "grad_norm": 0.13066288819408423,
      "learning_rate": 8.53413202585959e-06,
      "loss": 0.0484,
      "num_tokens": 52688848.0,
      "step": 691
    },
    {
      "epoch": 0.8633811603243917,
      "grad_norm": 0.13458486228791905,
      "learning_rate": 8.529654213089266e-06,
      "loss": 0.0463,
      "num_tokens": 52765192.0,
      "step": 692
    },
    {
      "epoch": 0.8646288209606987,
      "grad_norm": 0.12641960186422305,
      "learning_rate": 8.52517090664012e-06,
      "loss": 0.0433,
      "num_tokens": 52840184.0,
      "step": 693
    },
    {
      "epoch": 0.8658764815970056,
      "grad_norm": 0.139222442092495,
      "learning_rate": 8.520682114641739e-06,
      "loss": 0.0485,
      "num_tokens": 52915968.0,
      "step": 694
    },
    {
      "epoch": 0.8671241422333126,
      "grad_norm": 0.1310804758808409,
      "learning_rate": 8.51618784523366e-06,
      "loss": 0.0459,
      "num_tokens": 52991860.0,
      "step": 695
    },
    {
      "epoch": 0.8683718028696195,
      "grad_norm": 0.14181865718671147,
      "learning_rate": 8.511688106565356e-06,
      "loss": 0.0488,
      "num_tokens": 53068597.0,
      "step": 696
    },
    {
      "epoch": 0.8696194635059263,
      "grad_norm": 0.13954957893418474,
      "learning_rate": 8.507182906796209e-06,
      "loss": 0.0537,
      "num_tokens": 53146323.0,
      "step": 697
    },
    {
      "epoch": 0.8708671241422333,
      "grad_norm": 0.13072333860570592,
      "learning_rate": 8.50267225409551e-06,
      "loss": 0.0439,
      "num_tokens": 53221723.0,
      "step": 698
    },
    {
      "epoch": 0.8721147847785402,
      "grad_norm": 0.13924135218825956,
      "learning_rate": 8.498156156642434e-06,
      "loss": 0.0485,
      "num_tokens": 53297446.0,
      "step": 699
    },
    {
      "epoch": 0.8733624454148472,
      "grad_norm": 0.1277280295683365,
      "learning_rate": 8.493634622626031e-06,
      "loss": 0.0446,
      "num_tokens": 53373744.0,
      "step": 700
    },
    {
      "epoch": 0.8746101060511541,
      "grad_norm": 0.13143765975583496,
      "learning_rate": 8.489107660245208e-06,
      "loss": 0.0422,
      "num_tokens": 53448470.0,
      "step": 701
    },
    {
      "epoch": 0.875857766687461,
      "grad_norm": 0.13289670713727797,
      "learning_rate": 8.484575277708718e-06,
      "loss": 0.0478,
      "num_tokens": 53526159.0,
      "step": 702
    },
    {
      "epoch": 0.877105427323768,
      "grad_norm": 0.14509962938350404,
      "learning_rate": 8.480037483235142e-06,
      "loss": 0.0475,
      "num_tokens": 53602483.0,
      "step": 703
    },
    {
      "epoch": 0.8783530879600748,
      "grad_norm": 0.1282228253361154,
      "learning_rate": 8.475494285052873e-06,
      "loss": 0.0497,
      "num_tokens": 53679362.0,
      "step": 704
    },
    {
      "epoch": 0.8796007485963818,
      "grad_norm": 0.12663678209388982,
      "learning_rate": 8.470945691400095e-06,
      "loss": 0.0455,
      "num_tokens": 53756359.0,
      "step": 705
    },
    {
      "epoch": 0.8808484092326887,
      "grad_norm": 0.1387150438279793,
      "learning_rate": 8.466391710524792e-06,
      "loss": 0.0486,
      "num_tokens": 53832649.0,
      "step": 706
    },
    {
      "epoch": 0.8820960698689956,
      "grad_norm": 0.13224121719631327,
      "learning_rate": 8.461832350684701e-06,
      "loss": 0.0468,
      "num_tokens": 53908486.0,
      "step": 707
    },
    {
      "epoch": 0.8833437305053026,
      "grad_norm": 0.12567190952105986,
      "learning_rate": 8.457267620147326e-06,
      "loss": 0.047,
      "num_tokens": 53984289.0,
      "step": 708
    },
    {
      "epoch": 0.8845913911416095,
      "grad_norm": 0.11477635822823083,
      "learning_rate": 8.452697527189901e-06,
      "loss": 0.0428,
      "num_tokens": 54060920.0,
      "step": 709
    },
    {
      "epoch": 0.8858390517779164,
      "grad_norm": 0.12985982634867996,
      "learning_rate": 8.448122080099384e-06,
      "loss": 0.0506,
      "num_tokens": 54137223.0,
      "step": 710
    },
    {
      "epoch": 0.8870867124142233,
      "grad_norm": 0.1303891811928819,
      "learning_rate": 8.443541287172443e-06,
      "loss": 0.0484,
      "num_tokens": 54214088.0,
      "step": 711
    },
    {
      "epoch": 0.8883343730505302,
      "grad_norm": 0.12282561290472704,
      "learning_rate": 8.438955156715443e-06,
      "loss": 0.0419,
      "num_tokens": 54289595.0,
      "step": 712
    },
    {
      "epoch": 0.8895820336868372,
      "grad_norm": 0.13256356712192674,
      "learning_rate": 8.434363697044423e-06,
      "loss": 0.0474,
      "num_tokens": 54365777.0,
      "step": 713
    },
    {
      "epoch": 0.8908296943231441,
      "grad_norm": 0.1295150884171825,
      "learning_rate": 8.429766916485087e-06,
      "loss": 0.0482,
      "num_tokens": 54442444.0,
      "step": 714
    },
    {
      "epoch": 0.8920773549594511,
      "grad_norm": 0.1354661630576582,
      "learning_rate": 8.42516482337279e-06,
      "loss": 0.0474,
      "num_tokens": 54517968.0,
      "step": 715
    },
    {
      "epoch": 0.8933250155957579,
      "grad_norm": 0.12661066471963617,
      "learning_rate": 8.420557426052513e-06,
      "loss": 0.0477,
      "num_tokens": 54594152.0,
      "step": 716
    },
    {
      "epoch": 0.8945726762320648,
      "grad_norm": 0.13003061445709582,
      "learning_rate": 8.415944732878863e-06,
      "loss": 0.0462,
      "num_tokens": 54670044.0,
      "step": 717
    },
    {
      "epoch": 0.8958203368683718,
      "grad_norm": 0.1347890412511373,
      "learning_rate": 8.411326752216048e-06,
      "loss": 0.0477,
      "num_tokens": 54745643.0,
      "step": 718
    },
    {
      "epoch": 0.8970679975046787,
      "grad_norm": 0.13806299761156438,
      "learning_rate": 8.406703492437863e-06,
      "loss": 0.0462,
      "num_tokens": 54820709.0,
      "step": 719
    },
    {
      "epoch": 0.8983156581409857,
      "grad_norm": 0.15064441560708958,
      "learning_rate": 8.402074961927674e-06,
      "loss": 0.0506,
      "num_tokens": 54896955.0,
      "step": 720
    },
    {
      "epoch": 0.8995633187772926,
      "grad_norm": 0.12826416085523926,
      "learning_rate": 8.397441169078404e-06,
      "loss": 0.0468,
      "num_tokens": 54974212.0,
      "step": 721
    },
    {
      "epoch": 0.9008109794135996,
      "grad_norm": 0.12699295137628364,
      "learning_rate": 8.392802122292522e-06,
      "loss": 0.0451,
      "num_tokens": 55050113.0,
      "step": 722
    },
    {
      "epoch": 0.9020586400499064,
      "grad_norm": 0.1401057811483236,
      "learning_rate": 8.388157829982023e-06,
      "loss": 0.045,
      "num_tokens": 55126297.0,
      "step": 723
    },
    {
      "epoch": 0.9033063006862133,
      "grad_norm": 0.13188033083997214,
      "learning_rate": 8.383508300568409e-06,
      "loss": 0.0527,
      "num_tokens": 55203060.0,
      "step": 724
    },
    {
      "epoch": 0.9045539613225203,
      "grad_norm": 0.1272260308340507,
      "learning_rate": 8.378853542482687e-06,
      "loss": 0.0444,
      "num_tokens": 55278770.0,
      "step": 725
    },
    {
      "epoch": 0.9058016219588272,
      "grad_norm": 0.14369524671244313,
      "learning_rate": 8.374193564165338e-06,
      "loss": 0.0444,
      "num_tokens": 55354950.0,
      "step": 726
    },
    {
      "epoch": 0.9070492825951342,
      "grad_norm": 0.13648519325941133,
      "learning_rate": 8.36952837406631e-06,
      "loss": 0.0473,
      "num_tokens": 55431358.0,
      "step": 727
    },
    {
      "epoch": 0.9082969432314411,
      "grad_norm": 0.12922616460751848,
      "learning_rate": 8.364857980645006e-06,
      "loss": 0.0454,
      "num_tokens": 55507747.0,
      "step": 728
    },
    {
      "epoch": 0.9095446038677479,
      "grad_norm": 0.127875634016502,
      "learning_rate": 8.360182392370258e-06,
      "loss": 0.0481,
      "num_tokens": 55585197.0,
      "step": 729
    },
    {
      "epoch": 0.9107922645040549,
      "grad_norm": 0.13104378719838006,
      "learning_rate": 8.355501617720321e-06,
      "loss": 0.0439,
      "num_tokens": 55660574.0,
      "step": 730
    },
    {
      "epoch": 0.9120399251403618,
      "grad_norm": 0.12926845838420958,
      "learning_rate": 8.350815665182855e-06,
      "loss": 0.0441,
      "num_tokens": 55736377.0,
      "step": 731
    },
    {
      "epoch": 0.9132875857766688,
      "grad_norm": 0.1270523993265075,
      "learning_rate": 8.34612454325491e-06,
      "loss": 0.0471,
      "num_tokens": 55812894.0,
      "step": 732
    },
    {
      "epoch": 0.9145352464129757,
      "grad_norm": 0.12394117715946282,
      "learning_rate": 8.341428260442907e-06,
      "loss": 0.0447,
      "num_tokens": 55888948.0,
      "step": 733
    },
    {
      "epoch": 0.9157829070492826,
      "grad_norm": 0.12072645194274476,
      "learning_rate": 8.336726825262622e-06,
      "loss": 0.0451,
      "num_tokens": 55964442.0,
      "step": 734
    },
    {
      "epoch": 0.9170305676855895,
      "grad_norm": 0.12672993586675507,
      "learning_rate": 8.332020246239183e-06,
      "loss": 0.0465,
      "num_tokens": 56041127.0,
      "step": 735
    },
    {
      "epoch": 0.9182782283218964,
      "grad_norm": 0.13657796012109533,
      "learning_rate": 8.327308531907039e-06,
      "loss": 0.0482,
      "num_tokens": 56119235.0,
      "step": 736
    },
    {
      "epoch": 0.9195258889582034,
      "grad_norm": 0.13041755901650912,
      "learning_rate": 8.322591690809952e-06,
      "loss": 0.0469,
      "num_tokens": 56195147.0,
      "step": 737
    },
    {
      "epoch": 0.9207735495945103,
      "grad_norm": 0.12361588139014848,
      "learning_rate": 8.317869731500981e-06,
      "loss": 0.0557,
      "num_tokens": 56271734.0,
      "step": 738
    },
    {
      "epoch": 0.9220212102308172,
      "grad_norm": 0.1272077384011452,
      "learning_rate": 8.313142662542465e-06,
      "loss": 0.0462,
      "num_tokens": 56348987.0,
      "step": 739
    },
    {
      "epoch": 0.9232688708671242,
      "grad_norm": 0.1399474323212219,
      "learning_rate": 8.30841049250601e-06,
      "loss": 0.0455,
      "num_tokens": 56426136.0,
      "step": 740
    },
    {
      "epoch": 0.924516531503431,
      "grad_norm": 0.13812075909506322,
      "learning_rate": 8.303673229972468e-06,
      "loss": 0.0496,
      "num_tokens": 56503751.0,
      "step": 741
    },
    {
      "epoch": 0.925764192139738,
      "grad_norm": 0.13599501450992818,
      "learning_rate": 8.298930883531932e-06,
      "loss": 0.0524,
      "num_tokens": 56581417.0,
      "step": 742
    },
    {
      "epoch": 0.9270118527760449,
      "grad_norm": 0.1325812107893234,
      "learning_rate": 8.294183461783704e-06,
      "loss": 0.0471,
      "num_tokens": 56656845.0,
      "step": 743
    },
    {
      "epoch": 0.9282595134123518,
      "grad_norm": 0.12712734428173242,
      "learning_rate": 8.2894309733363e-06,
      "loss": 0.0454,
      "num_tokens": 56733635.0,
      "step": 744
    },
    {
      "epoch": 0.9295071740486588,
      "grad_norm": 0.14005206065485848,
      "learning_rate": 8.284673426807413e-06,
      "loss": 0.0463,
      "num_tokens": 56809680.0,
      "step": 745
    },
    {
      "epoch": 0.9307548346849657,
      "grad_norm": 0.14087300313657175,
      "learning_rate": 8.279910830823917e-06,
      "loss": 0.0459,
      "num_tokens": 56884969.0,
      "step": 746
    },
    {
      "epoch": 0.9320024953212727,
      "grad_norm": 0.1352822553422503,
      "learning_rate": 8.275143194021837e-06,
      "loss": 0.0426,
      "num_tokens": 56960412.0,
      "step": 747
    },
    {
      "epoch": 0.9332501559575795,
      "grad_norm": 0.1248091276939536,
      "learning_rate": 8.270370525046338e-06,
      "loss": 0.0442,
      "num_tokens": 57035943.0,
      "step": 748
    },
    {
      "epoch": 0.9344978165938864,
      "grad_norm": 0.14141827065829213,
      "learning_rate": 8.265592832551714e-06,
      "loss": 0.0499,
      "num_tokens": 57112296.0,
      "step": 749
    },
    {
      "epoch": 0.9357454772301934,
      "grad_norm": 0.15528789441951488,
      "learning_rate": 8.260810125201363e-06,
      "loss": 0.0523,
      "num_tokens": 57189546.0,
      "step": 750
    },
    {
      "epoch": 0.9369931378665003,
      "grad_norm": 0.1444757644146868,
      "learning_rate": 8.25602241166778e-06,
      "loss": 0.0477,
      "num_tokens": 57266306.0,
      "step": 751
    },
    {
      "epoch": 0.9382407985028073,
      "grad_norm": 0.12551880038959676,
      "learning_rate": 8.251229700632536e-06,
      "loss": 0.0459,
      "num_tokens": 57341686.0,
      "step": 752
    },
    {
      "epoch": 0.9394884591391142,
      "grad_norm": 0.13610947168319384,
      "learning_rate": 8.246432000786267e-06,
      "loss": 0.0466,
      "num_tokens": 57418946.0,
      "step": 753
    },
    {
      "epoch": 0.940736119775421,
      "grad_norm": 0.1418391670800807,
      "learning_rate": 8.241629320828652e-06,
      "loss": 0.0436,
      "num_tokens": 57494409.0,
      "step": 754
    },
    {
      "epoch": 0.941983780411728,
      "grad_norm": 0.14475322007531208,
      "learning_rate": 8.2368216694684e-06,
      "loss": 0.0474,
      "num_tokens": 57572009.0,
      "step": 755
    },
    {
      "epoch": 0.9432314410480349,
      "grad_norm": 0.13184602873921808,
      "learning_rate": 8.232009055423236e-06,
      "loss": 0.0423,
      "num_tokens": 57646117.0,
      "step": 756
    },
    {
      "epoch": 0.9444791016843419,
      "grad_norm": 0.12812576418278393,
      "learning_rate": 8.227191487419887e-06,
      "loss": 0.044,
      "num_tokens": 57721560.0,
      "step": 757
    },
    {
      "epoch": 0.9457267623206488,
      "grad_norm": 0.12654952839318306,
      "learning_rate": 8.222368974194057e-06,
      "loss": 0.0461,
      "num_tokens": 57798213.0,
      "step": 758
    },
    {
      "epoch": 0.9469744229569557,
      "grad_norm": 0.12934305800781434,
      "learning_rate": 8.217541524490422e-06,
      "loss": 0.0478,
      "num_tokens": 57875229.0,
      "step": 759
    },
    {
      "epoch": 0.9482220835932627,
      "grad_norm": 0.13295210650526854,
      "learning_rate": 8.212709147062604e-06,
      "loss": 0.0444,
      "num_tokens": 57950833.0,
      "step": 760
    },
    {
      "epoch": 0.9494697442295695,
      "grad_norm": 0.12300137067417298,
      "learning_rate": 8.207871850673168e-06,
      "loss": 0.0465,
      "num_tokens": 58027155.0,
      "step": 761
    },
    {
      "epoch": 0.9507174048658765,
      "grad_norm": 0.12846718774897786,
      "learning_rate": 8.203029644093593e-06,
      "loss": 0.0485,
      "num_tokens": 58103747.0,
      "step": 762
    },
    {
      "epoch": 0.9519650655021834,
      "grad_norm": 0.12166615887497252,
      "learning_rate": 8.198182536104262e-06,
      "loss": 0.0444,
      "num_tokens": 58179264.0,
      "step": 763
    },
    {
      "epoch": 0.9532127261384903,
      "grad_norm": 0.13359004503185487,
      "learning_rate": 8.193330535494448e-06,
      "loss": 0.0459,
      "num_tokens": 58255206.0,
      "step": 764
    },
    {
      "epoch": 0.9544603867747973,
      "grad_norm": 0.13069849304221773,
      "learning_rate": 8.188473651062296e-06,
      "loss": 0.0455,
      "num_tokens": 58330290.0,
      "step": 765
    },
    {
      "epoch": 0.9557080474111042,
      "grad_norm": 0.1376163273913286,
      "learning_rate": 8.183611891614803e-06,
      "loss": 0.0443,
      "num_tokens": 58405956.0,
      "step": 766
    },
    {
      "epoch": 0.9569557080474111,
      "grad_norm": 0.13686003984330788,
      "learning_rate": 8.178745265967808e-06,
      "loss": 0.0464,
      "num_tokens": 58482233.0,
      "step": 767
    },
    {
      "epoch": 0.958203368683718,
      "grad_norm": 0.12627060136965676,
      "learning_rate": 8.173873782945976e-06,
      "loss": 0.047,
      "num_tokens": 58559969.0,
      "step": 768
    },
    {
      "epoch": 0.9594510293200249,
      "grad_norm": 0.12879510165555993,
      "learning_rate": 8.168997451382778e-06,
      "loss": 0.0482,
      "num_tokens": 58635674.0,
      "step": 769
    },
    {
      "epoch": 0.9606986899563319,
      "grad_norm": 0.1331376647953998,
      "learning_rate": 8.164116280120478e-06,
      "loss": 0.0442,
      "num_tokens": 58712224.0,
      "step": 770
    },
    {
      "epoch": 0.9619463505926388,
      "grad_norm": 0.13246529667252444,
      "learning_rate": 8.159230278010113e-06,
      "loss": 0.0464,
      "num_tokens": 58788204.0,
      "step": 771
    },
    {
      "epoch": 0.9631940112289458,
      "grad_norm": 0.13752849551124316,
      "learning_rate": 8.154339453911483e-06,
      "loss": 0.0453,
      "num_tokens": 58864120.0,
      "step": 772
    },
    {
      "epoch": 0.9644416718652526,
      "grad_norm": 0.118632440035503,
      "learning_rate": 8.14944381669313e-06,
      "loss": 0.0401,
      "num_tokens": 58938337.0,
      "step": 773
    },
    {
      "epoch": 0.9656893325015595,
      "grad_norm": 0.11478561278311256,
      "learning_rate": 8.144543375232322e-06,
      "loss": 0.0442,
      "num_tokens": 59013332.0,
      "step": 774
    },
    {
      "epoch": 0.9669369931378665,
      "grad_norm": 0.13782337191599275,
      "learning_rate": 8.139638138415041e-06,
      "loss": 0.0462,
      "num_tokens": 59088997.0,
      "step": 775
    },
    {
      "epoch": 0.9681846537741734,
      "grad_norm": 0.12839918012353083,
      "learning_rate": 8.134728115135967e-06,
      "loss": 0.0484,
      "num_tokens": 59164665.0,
      "step": 776
    },
    {
      "epoch": 0.9694323144104804,
      "grad_norm": 0.12847654480275927,
      "learning_rate": 8.129813314298457e-06,
      "loss": 0.0469,
      "num_tokens": 59241148.0,
      "step": 777
    },
    {
      "epoch": 0.9706799750467873,
      "grad_norm": 0.12255793939033695,
      "learning_rate": 8.124893744814524e-06,
      "loss": 0.0471,
      "num_tokens": 59318320.0,
      "step": 778
    },
    {
      "epoch": 0.9719276356830941,
      "grad_norm": 0.1254272675633237,
      "learning_rate": 8.11996941560484e-06,
      "loss": 0.042,
      "num_tokens": 59393045.0,
      "step": 779
    },
    {
      "epoch": 0.9731752963194011,
      "grad_norm": 0.1273761325986967,
      "learning_rate": 8.115040335598701e-06,
      "loss": 0.046,
      "num_tokens": 59472300.0,
      "step": 780
    },
    {
      "epoch": 0.974422956955708,
      "grad_norm": 0.12602861483850217,
      "learning_rate": 8.110106513734019e-06,
      "loss": 0.0429,
      "num_tokens": 59548496.0,
      "step": 781
    },
    {
      "epoch": 0.975670617592015,
      "grad_norm": 0.1363028683424439,
      "learning_rate": 8.105167958957302e-06,
      "loss": 0.0493,
      "num_tokens": 59626492.0,
      "step": 782
    },
    {
      "epoch": 0.9769182782283219,
      "grad_norm": 0.12803093679059996,
      "learning_rate": 8.100224680223647e-06,
      "loss": 0.0538,
      "num_tokens": 59702812.0,
      "step": 783
    },
    {
      "epoch": 0.9781659388646288,
      "grad_norm": 0.1316321034732895,
      "learning_rate": 8.09527668649671e-06,
      "loss": 0.0462,
      "num_tokens": 59779318.0,
      "step": 784
    },
    {
      "epoch": 0.9794135995009358,
      "grad_norm": 0.13441963780619764,
      "learning_rate": 8.090323986748696e-06,
      "loss": 0.0485,
      "num_tokens": 59855866.0,
      "step": 785
    },
    {
      "epoch": 0.9806612601372426,
      "grad_norm": 0.13172648789937602,
      "learning_rate": 8.085366589960353e-06,
      "loss": 0.045,
      "num_tokens": 59933364.0,
      "step": 786
    },
    {
      "epoch": 0.9819089207735496,
      "grad_norm": 0.13013258573518577,
      "learning_rate": 8.080404505120936e-06,
      "loss": 0.0441,
      "num_tokens": 60009279.0,
      "step": 787
    },
    {
      "epoch": 0.9831565814098565,
      "grad_norm": 0.1319062787442704,
      "learning_rate": 8.075437741228205e-06,
      "loss": 0.0464,
      "num_tokens": 60085623.0,
      "step": 788
    },
    {
      "epoch": 0.9844042420461634,
      "grad_norm": 0.1308809048521274,
      "learning_rate": 8.070466307288404e-06,
      "loss": 0.0456,
      "num_tokens": 60161774.0,
      "step": 789
    },
    {
      "epoch": 0.9856519026824704,
      "grad_norm": 0.12602659288322934,
      "learning_rate": 8.065490212316245e-06,
      "loss": 0.0439,
      "num_tokens": 60238406.0,
      "step": 790
    },
    {
      "epoch": 0.9868995633187773,
      "grad_norm": 0.13344938845482,
      "learning_rate": 8.060509465334895e-06,
      "loss": 0.0473,
      "num_tokens": 60314716.0,
      "step": 791
    },
    {
      "epoch": 0.9881472239550843,
      "grad_norm": 0.1357885856390025,
      "learning_rate": 8.055524075375951e-06,
      "loss": 0.047,
      "num_tokens": 60391513.0,
      "step": 792
    },
    {
      "epoch": 0.9893948845913911,
      "grad_norm": 0.13692794199262653,
      "learning_rate": 8.050534051479432e-06,
      "loss": 0.0488,
      "num_tokens": 60468385.0,
      "step": 793
    },
    {
      "epoch": 0.990642545227698,
      "grad_norm": 0.1337920475499864,
      "learning_rate": 8.045539402693759e-06,
      "loss": 0.0478,
      "num_tokens": 60545686.0,
      "step": 794
    },
    {
      "epoch": 0.991890205864005,
      "grad_norm": 0.12891045119563382,
      "learning_rate": 8.040540138075743e-06,
      "loss": 0.0484,
      "num_tokens": 60622263.0,
      "step": 795
    },
    {
      "epoch": 0.9931378665003119,
      "grad_norm": 0.12475341972299739,
      "learning_rate": 8.035536266690561e-06,
      "loss": 0.0453,
      "num_tokens": 60698072.0,
      "step": 796
    },
    {
      "epoch": 0.9943855271366189,
      "grad_norm": 0.12557072799605964,
      "learning_rate": 8.030527797611742e-06,
      "loss": 0.0473,
      "num_tokens": 60773991.0,
      "step": 797
    },
    {
      "epoch": 0.9956331877729258,
      "grad_norm": 0.1326655960570811,
      "learning_rate": 8.025514739921155e-06,
      "loss": 0.0471,
      "num_tokens": 60850174.0,
      "step": 798
    },
    {
      "epoch": 0.9968808484092326,
      "grad_norm": 0.14443419491894904,
      "learning_rate": 8.02049710270899e-06,
      "loss": 0.0487,
      "num_tokens": 60926616.0,
      "step": 799
    },
    {
      "epoch": 0.9981285090455396,
      "grad_norm": 0.13979837723347358,
      "learning_rate": 8.015474895073739e-06,
      "loss": 0.0489,
      "num_tokens": 61003563.0,
      "step": 800
    },
    {
      "epoch": 0.9993761696818465,
      "grad_norm": 0.13287610948560982,
      "learning_rate": 8.010448126122183e-06,
      "loss": 0.0519,
      "num_tokens": 61080948.0,
      "step": 801
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.13287610948560982,
      "learning_rate": 8.005416804969374e-06,
      "loss": 0.0464,
      "num_tokens": 61119461.0,
      "step": 802
    },
    {
      "epoch": 1.001247660636307,
      "grad_norm": 0.21428891616070306,
      "learning_rate": 8.000380940738616e-06,
      "loss": 0.0394,
      "num_tokens": 61194181.0,
      "step": 803
    },
    {
      "epoch": 1.0024953212726138,
      "grad_norm": 0.11463542811424311,
      "learning_rate": 7.995340542561453e-06,
      "loss": 0.0382,
      "num_tokens": 61269860.0,
      "step": 804
    },
    {
      "epoch": 1.0037429819089208,
      "grad_norm": 0.11296312584424995,
      "learning_rate": 7.990295619577653e-06,
      "loss": 0.0393,
      "num_tokens": 61346977.0,
      "step": 805
    },
    {
      "epoch": 1.0049906425452277,
      "grad_norm": 0.12142614458207387,
      "learning_rate": 7.985246180935184e-06,
      "loss": 0.0426,
      "num_tokens": 61423778.0,
      "step": 806
    },
    {
      "epoch": 1.0062383031815347,
      "grad_norm": 0.124283410642958,
      "learning_rate": 7.980192235790207e-06,
      "loss": 0.04,
      "num_tokens": 61500104.0,
      "step": 807
    },
    {
      "epoch": 1.0074859638178415,
      "grad_norm": 0.11514734690512161,
      "learning_rate": 7.97513379330705e-06,
      "loss": 0.0409,
      "num_tokens": 61577699.0,
      "step": 808
    },
    {
      "epoch": 1.0087336244541485,
      "grad_norm": 0.12772053062936375,
      "learning_rate": 7.970070862658198e-06,
      "loss": 0.0385,
      "num_tokens": 61654726.0,
      "step": 809
    },
    {
      "epoch": 1.0099812850904555,
      "grad_norm": 0.1420848581428031,
      "learning_rate": 7.965003453024273e-06,
      "loss": 0.0402,
      "num_tokens": 61731992.0,
      "step": 810
    },
    {
      "epoch": 1.0112289457267623,
      "grad_norm": 0.12101709378925632,
      "learning_rate": 7.959931573594025e-06,
      "loss": 0.0376,
      "num_tokens": 61809566.0,
      "step": 811
    },
    {
      "epoch": 1.0124766063630692,
      "grad_norm": 0.12379290639805463,
      "learning_rate": 7.954855233564301e-06,
      "loss": 0.0388,
      "num_tokens": 61886528.0,
      "step": 812
    },
    {
      "epoch": 1.0137242669993762,
      "grad_norm": 0.1310115238681384,
      "learning_rate": 7.949774442140043e-06,
      "loss": 0.0385,
      "num_tokens": 61962417.0,
      "step": 813
    },
    {
      "epoch": 1.014971927635683,
      "grad_norm": 0.13401797553444206,
      "learning_rate": 7.944689208534257e-06,
      "loss": 0.0383,
      "num_tokens": 62038796.0,
      "step": 814
    },
    {
      "epoch": 1.01621958827199,
      "grad_norm": 0.12765079817825717,
      "learning_rate": 7.939599541968012e-06,
      "loss": 0.0442,
      "num_tokens": 62114741.0,
      "step": 815
    },
    {
      "epoch": 1.017467248908297,
      "grad_norm": 0.1469418814720509,
      "learning_rate": 7.93450545167041e-06,
      "loss": 0.0379,
      "num_tokens": 62191755.0,
      "step": 816
    },
    {
      "epoch": 1.018714909544604,
      "grad_norm": 0.1369016694262358,
      "learning_rate": 7.929406946878576e-06,
      "loss": 0.0369,
      "num_tokens": 62266358.0,
      "step": 817
    },
    {
      "epoch": 1.0199625701809107,
      "grad_norm": 0.12983077233787918,
      "learning_rate": 7.924304036837643e-06,
      "loss": 0.0387,
      "num_tokens": 62341490.0,
      "step": 818
    },
    {
      "epoch": 1.0212102308172177,
      "grad_norm": 0.13162911142976996,
      "learning_rate": 7.919196730800727e-06,
      "loss": 0.0387,
      "num_tokens": 62418083.0,
      "step": 819
    },
    {
      "epoch": 1.0224578914535247,
      "grad_norm": 0.1358666256385649,
      "learning_rate": 7.914085038028918e-06,
      "loss": 0.041,
      "num_tokens": 62494660.0,
      "step": 820
    },
    {
      "epoch": 1.0237055520898315,
      "grad_norm": 0.13858343996922898,
      "learning_rate": 7.908968967791262e-06,
      "loss": 0.0424,
      "num_tokens": 62571070.0,
      "step": 821
    },
    {
      "epoch": 1.0249532127261385,
      "grad_norm": 0.11844680272321946,
      "learning_rate": 7.903848529364738e-06,
      "loss": 0.0405,
      "num_tokens": 62648820.0,
      "step": 822
    },
    {
      "epoch": 1.0262008733624455,
      "grad_norm": 0.1252962753765857,
      "learning_rate": 7.89872373203425e-06,
      "loss": 0.0398,
      "num_tokens": 62726276.0,
      "step": 823
    },
    {
      "epoch": 1.0274485339987522,
      "grad_norm": 0.11847161903518785,
      "learning_rate": 7.893594585092601e-06,
      "loss": 0.0415,
      "num_tokens": 62802667.0,
      "step": 824
    },
    {
      "epoch": 1.0286961946350592,
      "grad_norm": 0.13434324948235268,
      "learning_rate": 7.888461097840494e-06,
      "loss": 0.041,
      "num_tokens": 62878138.0,
      "step": 825
    },
    {
      "epoch": 1.0299438552713662,
      "grad_norm": 0.12675768390987047,
      "learning_rate": 7.883323279586483e-06,
      "loss": 0.0393,
      "num_tokens": 62954633.0,
      "step": 826
    },
    {
      "epoch": 1.0311915159076732,
      "grad_norm": 0.11827927529499944,
      "learning_rate": 7.87818113964699e-06,
      "loss": 0.0391,
      "num_tokens": 63030685.0,
      "step": 827
    },
    {
      "epoch": 1.03243917654398,
      "grad_norm": 0.13125937485650266,
      "learning_rate": 7.873034687346268e-06,
      "loss": 0.0372,
      "num_tokens": 63106146.0,
      "step": 828
    },
    {
      "epoch": 1.033686837180287,
      "grad_norm": 0.11652636577512891,
      "learning_rate": 7.86788393201639e-06,
      "loss": 0.0379,
      "num_tokens": 63181401.0,
      "step": 829
    },
    {
      "epoch": 1.034934497816594,
      "grad_norm": 0.1341701967742822,
      "learning_rate": 7.862728882997236e-06,
      "loss": 0.0423,
      "num_tokens": 63257458.0,
      "step": 830
    },
    {
      "epoch": 1.0361821584529007,
      "grad_norm": 0.1364539989781808,
      "learning_rate": 7.857569549636462e-06,
      "loss": 0.0388,
      "num_tokens": 63334805.0,
      "step": 831
    },
    {
      "epoch": 1.0374298190892077,
      "grad_norm": 0.12536924010011452,
      "learning_rate": 7.852405941289503e-06,
      "loss": 0.0419,
      "num_tokens": 63411725.0,
      "step": 832
    },
    {
      "epoch": 1.0386774797255147,
      "grad_norm": 0.11900831109108075,
      "learning_rate": 7.847238067319542e-06,
      "loss": 0.0369,
      "num_tokens": 63487166.0,
      "step": 833
    },
    {
      "epoch": 1.0399251403618215,
      "grad_norm": 0.13712182486003496,
      "learning_rate": 7.842065937097495e-06,
      "loss": 0.0388,
      "num_tokens": 63561195.0,
      "step": 834
    },
    {
      "epoch": 1.0411728009981285,
      "grad_norm": 0.1322764156887544,
      "learning_rate": 7.836889560001997e-06,
      "loss": 0.0401,
      "num_tokens": 63638227.0,
      "step": 835
    },
    {
      "epoch": 1.0424204616344355,
      "grad_norm": 0.12845001330725533,
      "learning_rate": 7.831708945419383e-06,
      "loss": 0.0382,
      "num_tokens": 63714554.0,
      "step": 836
    },
    {
      "epoch": 1.0436681222707425,
      "grad_norm": 0.12086453985035646,
      "learning_rate": 7.826524102743678e-06,
      "loss": 0.0385,
      "num_tokens": 63791759.0,
      "step": 837
    },
    {
      "epoch": 1.0449157829070492,
      "grad_norm": 0.11543493690448661,
      "learning_rate": 7.821335041376565e-06,
      "loss": 0.0351,
      "num_tokens": 63867593.0,
      "step": 838
    },
    {
      "epoch": 1.0461634435433562,
      "grad_norm": 0.11861579198072326,
      "learning_rate": 7.816141770727381e-06,
      "loss": 0.0379,
      "num_tokens": 63942779.0,
      "step": 839
    },
    {
      "epoch": 1.0474111041796632,
      "grad_norm": 0.13188356504968343,
      "learning_rate": 7.810944300213095e-06,
      "loss": 0.0384,
      "num_tokens": 64020656.0,
      "step": 840
    },
    {
      "epoch": 1.04865876481597,
      "grad_norm": 0.1323045158713511,
      "learning_rate": 7.805742639258297e-06,
      "loss": 0.0376,
      "num_tokens": 64095353.0,
      "step": 841
    },
    {
      "epoch": 1.049906425452277,
      "grad_norm": 0.13242016376152926,
      "learning_rate": 7.800536797295164e-06,
      "loss": 0.042,
      "num_tokens": 64172670.0,
      "step": 842
    },
    {
      "epoch": 1.051154086088584,
      "grad_norm": 0.12019785744720937,
      "learning_rate": 7.795326783763463e-06,
      "loss": 0.0411,
      "num_tokens": 64252219.0,
      "step": 843
    },
    {
      "epoch": 1.0524017467248907,
      "grad_norm": 0.12945649987325733,
      "learning_rate": 7.790112608110523e-06,
      "loss": 0.0383,
      "num_tokens": 64327950.0,
      "step": 844
    },
    {
      "epoch": 1.0536494073611977,
      "grad_norm": 0.11589408734195418,
      "learning_rate": 7.784894279791224e-06,
      "loss": 0.0373,
      "num_tokens": 64403741.0,
      "step": 845
    },
    {
      "epoch": 1.0548970679975047,
      "grad_norm": 0.11991035427603171,
      "learning_rate": 7.779671808267968e-06,
      "loss": 0.0371,
      "num_tokens": 64479652.0,
      "step": 846
    },
    {
      "epoch": 1.0561447286338117,
      "grad_norm": 0.13203629959562918,
      "learning_rate": 7.774445203010676e-06,
      "loss": 0.0386,
      "num_tokens": 64555318.0,
      "step": 847
    },
    {
      "epoch": 1.0573923892701185,
      "grad_norm": 0.129045226504299,
      "learning_rate": 7.769214473496766e-06,
      "loss": 0.0397,
      "num_tokens": 64630854.0,
      "step": 848
    },
    {
      "epoch": 1.0586400499064255,
      "grad_norm": 0.13424488516410796,
      "learning_rate": 7.763979629211127e-06,
      "loss": 0.0405,
      "num_tokens": 64706946.0,
      "step": 849
    },
    {
      "epoch": 1.0598877105427325,
      "grad_norm": 0.1315607090534157,
      "learning_rate": 7.758740679646115e-06,
      "loss": 0.0379,
      "num_tokens": 64782647.0,
      "step": 850
    },
    {
      "epoch": 1.0611353711790392,
      "grad_norm": 0.12879285629017997,
      "learning_rate": 7.753497634301532e-06,
      "loss": 0.038,
      "num_tokens": 64859089.0,
      "step": 851
    },
    {
      "epoch": 1.0623830318153462,
      "grad_norm": 0.15633026157590887,
      "learning_rate": 7.748250502684601e-06,
      "loss": 0.0398,
      "num_tokens": 64935733.0,
      "step": 852
    },
    {
      "epoch": 1.0636306924516532,
      "grad_norm": 0.13634460601034967,
      "learning_rate": 7.742999294309959e-06,
      "loss": 0.0361,
      "num_tokens": 65010556.0,
      "step": 853
    },
    {
      "epoch": 1.06487835308796,
      "grad_norm": 0.12412517833563698,
      "learning_rate": 7.737744018699634e-06,
      "loss": 0.0405,
      "num_tokens": 65087047.0,
      "step": 854
    },
    {
      "epoch": 1.066126013724267,
      "grad_norm": 0.12716959844997922,
      "learning_rate": 7.732484685383027e-06,
      "loss": 0.0374,
      "num_tokens": 65162777.0,
      "step": 855
    },
    {
      "epoch": 1.067373674360574,
      "grad_norm": 0.12952911734554354,
      "learning_rate": 7.7272213038969e-06,
      "loss": 0.0387,
      "num_tokens": 65239214.0,
      "step": 856
    },
    {
      "epoch": 1.068621334996881,
      "grad_norm": 0.126084293916441,
      "learning_rate": 7.72195388378536e-06,
      "loss": 0.0393,
      "num_tokens": 65315610.0,
      "step": 857
    },
    {
      "epoch": 1.0698689956331877,
      "grad_norm": 0.12342840379376283,
      "learning_rate": 7.716682434599823e-06,
      "loss": 0.0428,
      "num_tokens": 65393681.0,
      "step": 858
    },
    {
      "epoch": 1.0711166562694947,
      "grad_norm": 0.13963961410762832,
      "learning_rate": 7.711406965899026e-06,
      "loss": 0.0404,
      "num_tokens": 65469270.0,
      "step": 859
    },
    {
      "epoch": 1.0723643169058017,
      "grad_norm": 0.11725006905637131,
      "learning_rate": 7.706127487248984e-06,
      "loss": 0.0466,
      "num_tokens": 65544990.0,
      "step": 860
    },
    {
      "epoch": 1.0736119775421085,
      "grad_norm": 0.14170062304512607,
      "learning_rate": 7.70084400822299e-06,
      "loss": 0.0398,
      "num_tokens": 65620506.0,
      "step": 861
    },
    {
      "epoch": 1.0748596381784155,
      "grad_norm": 0.12578447988533376,
      "learning_rate": 7.695556538401588e-06,
      "loss": 0.0376,
      "num_tokens": 65696157.0,
      "step": 862
    },
    {
      "epoch": 1.0761072988147224,
      "grad_norm": 0.12367582814838458,
      "learning_rate": 7.690265087372559e-06,
      "loss": 0.0431,
      "num_tokens": 65773019.0,
      "step": 863
    },
    {
      "epoch": 1.0773549594510294,
      "grad_norm": 0.14315074651628906,
      "learning_rate": 7.684969664730903e-06,
      "loss": 0.0391,
      "num_tokens": 65848646.0,
      "step": 864
    },
    {
      "epoch": 1.0786026200873362,
      "grad_norm": 0.12376004136236546,
      "learning_rate": 7.679670280078823e-06,
      "loss": 0.0402,
      "num_tokens": 65925176.0,
      "step": 865
    },
    {
      "epoch": 1.0798502807236432,
      "grad_norm": 0.13019320712059,
      "learning_rate": 7.674366943025705e-06,
      "loss": 0.0407,
      "num_tokens": 66002394.0,
      "step": 866
    },
    {
      "epoch": 1.0810979413599502,
      "grad_norm": 0.11984237762577243,
      "learning_rate": 7.669059663188099e-06,
      "loss": 0.0381,
      "num_tokens": 66078590.0,
      "step": 867
    },
    {
      "epoch": 1.082345601996257,
      "grad_norm": 0.12615694715416223,
      "learning_rate": 7.66374845018971e-06,
      "loss": 0.0358,
      "num_tokens": 66152912.0,
      "step": 868
    },
    {
      "epoch": 1.083593262632564,
      "grad_norm": 0.12041922594151447,
      "learning_rate": 7.658433313661372e-06,
      "loss": 0.0393,
      "num_tokens": 66228873.0,
      "step": 869
    },
    {
      "epoch": 1.084840923268871,
      "grad_norm": 0.12121834014527919,
      "learning_rate": 7.653114263241034e-06,
      "loss": 0.0381,
      "num_tokens": 66305201.0,
      "step": 870
    },
    {
      "epoch": 1.0860885839051777,
      "grad_norm": 0.13007390524350204,
      "learning_rate": 7.647791308573744e-06,
      "loss": 0.0394,
      "num_tokens": 66381316.0,
      "step": 871
    },
    {
      "epoch": 1.0873362445414847,
      "grad_norm": 0.12809810180480036,
      "learning_rate": 7.642464459311623e-06,
      "loss": 0.0396,
      "num_tokens": 66457844.0,
      "step": 872
    },
    {
      "epoch": 1.0885839051777917,
      "grad_norm": 0.12234620308401925,
      "learning_rate": 7.637133725113864e-06,
      "loss": 0.0386,
      "num_tokens": 66533263.0,
      "step": 873
    },
    {
      "epoch": 1.0898315658140985,
      "grad_norm": 0.11097029253021012,
      "learning_rate": 7.631799115646697e-06,
      "loss": 0.0372,
      "num_tokens": 66609039.0,
      "step": 874
    },
    {
      "epoch": 1.0910792264504054,
      "grad_norm": 0.1387492283712563,
      "learning_rate": 7.6264606405833805e-06,
      "loss": 0.0382,
      "num_tokens": 66684998.0,
      "step": 875
    },
    {
      "epoch": 1.0923268870867124,
      "grad_norm": 0.12281557435913792,
      "learning_rate": 7.621118309604186e-06,
      "loss": 0.0376,
      "num_tokens": 66760201.0,
      "step": 876
    },
    {
      "epoch": 1.0935745477230194,
      "grad_norm": 0.12597782436592558,
      "learning_rate": 7.615772132396373e-06,
      "loss": 0.0387,
      "num_tokens": 66837068.0,
      "step": 877
    },
    {
      "epoch": 1.0948222083593262,
      "grad_norm": 0.12304719888375368,
      "learning_rate": 7.6104221186541745e-06,
      "loss": 0.04,
      "num_tokens": 66912815.0,
      "step": 878
    },
    {
      "epoch": 1.0960698689956332,
      "grad_norm": 0.13032677225921888,
      "learning_rate": 7.6050682780787865e-06,
      "loss": 0.0432,
      "num_tokens": 66989780.0,
      "step": 879
    },
    {
      "epoch": 1.0973175296319402,
      "grad_norm": 0.1236460004101698,
      "learning_rate": 7.599710620378337e-06,
      "loss": 0.0405,
      "num_tokens": 67066176.0,
      "step": 880
    },
    {
      "epoch": 1.098565190268247,
      "grad_norm": 0.11979239248552905,
      "learning_rate": 7.594349155267879e-06,
      "loss": 0.0383,
      "num_tokens": 67142617.0,
      "step": 881
    },
    {
      "epoch": 1.099812850904554,
      "grad_norm": 0.12602615420228466,
      "learning_rate": 7.588983892469372e-06,
      "loss": 0.0383,
      "num_tokens": 67218428.0,
      "step": 882
    },
    {
      "epoch": 1.101060511540861,
      "grad_norm": 0.12720326733722198,
      "learning_rate": 7.583614841711657e-06,
      "loss": 0.0391,
      "num_tokens": 67293819.0,
      "step": 883
    },
    {
      "epoch": 1.102308172177168,
      "grad_norm": 0.13572936415520973,
      "learning_rate": 7.5782420127304466e-06,
      "loss": 0.0414,
      "num_tokens": 67370173.0,
      "step": 884
    },
    {
      "epoch": 1.1035558328134747,
      "grad_norm": 0.11987891385152093,
      "learning_rate": 7.572865415268303e-06,
      "loss": 0.0392,
      "num_tokens": 67446246.0,
      "step": 885
    },
    {
      "epoch": 1.1048034934497817,
      "grad_norm": 0.131690397693751,
      "learning_rate": 7.567485059074623e-06,
      "loss": 0.0387,
      "num_tokens": 67522135.0,
      "step": 886
    },
    {
      "epoch": 1.1060511540860887,
      "grad_norm": 0.15768309161114522,
      "learning_rate": 7.5621009539056175e-06,
      "loss": 0.0425,
      "num_tokens": 67599922.0,
      "step": 887
    },
    {
      "epoch": 1.1072988147223954,
      "grad_norm": 0.14441359221837843,
      "learning_rate": 7.556713109524301e-06,
      "loss": 0.0405,
      "num_tokens": 67677580.0,
      "step": 888
    },
    {
      "epoch": 1.1085464753587024,
      "grad_norm": 0.1292843222465948,
      "learning_rate": 7.551321535700456e-06,
      "loss": 0.0402,
      "num_tokens": 67754390.0,
      "step": 889
    },
    {
      "epoch": 1.1097941359950094,
      "grad_norm": 0.11922840726185915,
      "learning_rate": 7.545926242210643e-06,
      "loss": 0.0383,
      "num_tokens": 67830609.0,
      "step": 890
    },
    {
      "epoch": 1.1110417966313162,
      "grad_norm": 0.11666209441411698,
      "learning_rate": 7.540527238838156e-06,
      "loss": 0.0373,
      "num_tokens": 67905462.0,
      "step": 891
    },
    {
      "epoch": 1.1122894572676232,
      "grad_norm": 0.12935191115335293,
      "learning_rate": 7.535124535373019e-06,
      "loss": 0.0386,
      "num_tokens": 67982653.0,
      "step": 892
    },
    {
      "epoch": 1.1135371179039302,
      "grad_norm": 0.12276881001813526,
      "learning_rate": 7.529718141611972e-06,
      "loss": 0.0373,
      "num_tokens": 68057328.0,
      "step": 893
    },
    {
      "epoch": 1.114784778540237,
      "grad_norm": 0.13184667165067263,
      "learning_rate": 7.5243080673584345e-06,
      "loss": 0.0367,
      "num_tokens": 68132425.0,
      "step": 894
    },
    {
      "epoch": 1.116032439176544,
      "grad_norm": 0.11687424678949991,
      "learning_rate": 7.51889432242251e-06,
      "loss": 0.0387,
      "num_tokens": 68209624.0,
      "step": 895
    },
    {
      "epoch": 1.117280099812851,
      "grad_norm": 0.12821591356114023,
      "learning_rate": 7.513476916620952e-06,
      "loss": 0.0396,
      "num_tokens": 68286396.0,
      "step": 896
    },
    {
      "epoch": 1.118527760449158,
      "grad_norm": 0.13397330617441067,
      "learning_rate": 7.508055859777157e-06,
      "loss": 0.0375,
      "num_tokens": 68362383.0,
      "step": 897
    },
    {
      "epoch": 1.1197754210854647,
      "grad_norm": 0.11767653777640133,
      "learning_rate": 7.502631161721139e-06,
      "loss": 0.0369,
      "num_tokens": 68438133.0,
      "step": 898
    },
    {
      "epoch": 1.1210230817217717,
      "grad_norm": 0.12018970675793803,
      "learning_rate": 7.497202832289514e-06,
      "loss": 0.0393,
      "num_tokens": 68514149.0,
      "step": 899
    },
    {
      "epoch": 1.1222707423580787,
      "grad_norm": 0.124291033559986,
      "learning_rate": 7.4917708813254865e-06,
      "loss": 0.0389,
      "num_tokens": 68590799.0,
      "step": 900
    },
    {
      "epoch": 1.1235184029943854,
      "grad_norm": 0.11807610234269636,
      "learning_rate": 7.4863353186788234e-06,
      "loss": 0.0374,
      "num_tokens": 68666423.0,
      "step": 901
    },
    {
      "epoch": 1.1247660636306924,
      "grad_norm": 0.13303725093015237,
      "learning_rate": 7.480896154205844e-06,
      "loss": 0.038,
      "num_tokens": 68741602.0,
      "step": 902
    },
    {
      "epoch": 1.1260137242669994,
      "grad_norm": 0.11757569211552092,
      "learning_rate": 7.475453397769396e-06,
      "loss": 0.0399,
      "num_tokens": 68818640.0,
      "step": 903
    },
    {
      "epoch": 1.1272613849033064,
      "grad_norm": 0.13038770004184272,
      "learning_rate": 7.470007059238842e-06,
      "loss": 0.0379,
      "num_tokens": 68894885.0,
      "step": 904
    },
    {
      "epoch": 1.1285090455396132,
      "grad_norm": 0.13109249236061676,
      "learning_rate": 7.464557148490041e-06,
      "loss": 0.0396,
      "num_tokens": 68970453.0,
      "step": 905
    },
    {
      "epoch": 1.1297567061759202,
      "grad_norm": 0.12634016847293342,
      "learning_rate": 7.459103675405328e-06,
      "loss": 0.0372,
      "num_tokens": 69047236.0,
      "step": 906
    },
    {
      "epoch": 1.1310043668122272,
      "grad_norm": 0.12299085861262797,
      "learning_rate": 7.4536466498735e-06,
      "loss": 0.0397,
      "num_tokens": 69123903.0,
      "step": 907
    },
    {
      "epoch": 1.132252027448534,
      "grad_norm": 0.1319786544964183,
      "learning_rate": 7.44818608178979e-06,
      "loss": 0.0407,
      "num_tokens": 69200269.0,
      "step": 908
    },
    {
      "epoch": 1.133499688084841,
      "grad_norm": 0.13334686706378163,
      "learning_rate": 7.442721981055862e-06,
      "loss": 0.039,
      "num_tokens": 69275902.0,
      "step": 909
    },
    {
      "epoch": 1.134747348721148,
      "grad_norm": 0.12151092426485671,
      "learning_rate": 7.43725435757978e-06,
      "loss": 0.0377,
      "num_tokens": 69351293.0,
      "step": 910
    },
    {
      "epoch": 1.1359950093574547,
      "grad_norm": 0.11348410697636156,
      "learning_rate": 7.431783221275997e-06,
      "loss": 0.0375,
      "num_tokens": 69428894.0,
      "step": 911
    },
    {
      "epoch": 1.1372426699937617,
      "grad_norm": 0.11800100940339057,
      "learning_rate": 7.426308582065339e-06,
      "loss": 0.0377,
      "num_tokens": 69505235.0,
      "step": 912
    },
    {
      "epoch": 1.1384903306300687,
      "grad_norm": 0.13201001269167364,
      "learning_rate": 7.4208304498749825e-06,
      "loss": 0.0372,
      "num_tokens": 69582920.0,
      "step": 913
    },
    {
      "epoch": 1.1397379912663754,
      "grad_norm": 0.12114896597052012,
      "learning_rate": 7.415348834638433e-06,
      "loss": 0.0402,
      "num_tokens": 69659327.0,
      "step": 914
    },
    {
      "epoch": 1.1409856519026824,
      "grad_norm": 0.1333493333338715,
      "learning_rate": 7.40986374629552e-06,
      "loss": 0.0395,
      "num_tokens": 69735223.0,
      "step": 915
    },
    {
      "epoch": 1.1422333125389894,
      "grad_norm": 0.12347541882666985,
      "learning_rate": 7.404375194792365e-06,
      "loss": 0.0361,
      "num_tokens": 69810589.0,
      "step": 916
    },
    {
      "epoch": 1.1434809731752964,
      "grad_norm": 0.13077255784117334,
      "learning_rate": 7.398883190081368e-06,
      "loss": 0.0396,
      "num_tokens": 69886142.0,
      "step": 917
    },
    {
      "epoch": 1.1447286338116032,
      "grad_norm": 0.12866167128639208,
      "learning_rate": 7.3933877421211986e-06,
      "loss": 0.042,
      "num_tokens": 69962815.0,
      "step": 918
    },
    {
      "epoch": 1.1459762944479102,
      "grad_norm": 0.1426644408260711,
      "learning_rate": 7.387888860876763e-06,
      "loss": 0.0371,
      "num_tokens": 70038450.0,
      "step": 919
    },
    {
      "epoch": 1.1472239550842172,
      "grad_norm": 0.1191136878964363,
      "learning_rate": 7.382386556319193e-06,
      "loss": 0.0367,
      "num_tokens": 70113830.0,
      "step": 920
    },
    {
      "epoch": 1.1484716157205241,
      "grad_norm": 0.11766488295941847,
      "learning_rate": 7.376880838425832e-06,
      "loss": 0.0401,
      "num_tokens": 70190878.0,
      "step": 921
    },
    {
      "epoch": 1.149719276356831,
      "grad_norm": 0.1299870363233401,
      "learning_rate": 7.3713717171802106e-06,
      "loss": 0.0376,
      "num_tokens": 70267685.0,
      "step": 922
    },
    {
      "epoch": 1.150966936993138,
      "grad_norm": 0.1258019855180102,
      "learning_rate": 7.3658592025720285e-06,
      "loss": 0.0379,
      "num_tokens": 70346328.0,
      "step": 923
    },
    {
      "epoch": 1.152214597629445,
      "grad_norm": 0.12005546000064814,
      "learning_rate": 7.360343304597144e-06,
      "loss": 0.0405,
      "num_tokens": 70421909.0,
      "step": 924
    },
    {
      "epoch": 1.1534622582657517,
      "grad_norm": 0.13917434227685876,
      "learning_rate": 7.354824033257546e-06,
      "loss": 0.0388,
      "num_tokens": 70498763.0,
      "step": 925
    },
    {
      "epoch": 1.1547099189020587,
      "grad_norm": 0.1135457646599094,
      "learning_rate": 7.349301398561342e-06,
      "loss": 0.0374,
      "num_tokens": 70574235.0,
      "step": 926
    },
    {
      "epoch": 1.1559575795383656,
      "grad_norm": 0.12251659437468533,
      "learning_rate": 7.3437754105227365e-06,
      "loss": 0.0376,
      "num_tokens": 70650021.0,
      "step": 927
    },
    {
      "epoch": 1.1572052401746724,
      "grad_norm": 0.12140798784037611,
      "learning_rate": 7.3382460791620165e-06,
      "loss": 0.0399,
      "num_tokens": 70726089.0,
      "step": 928
    },
    {
      "epoch": 1.1584529008109794,
      "grad_norm": 0.13722375296866418,
      "learning_rate": 7.332713414505534e-06,
      "loss": 0.0408,
      "num_tokens": 70802890.0,
      "step": 929
    },
    {
      "epoch": 1.1597005614472864,
      "grad_norm": 0.12538243166812518,
      "learning_rate": 7.32717742658568e-06,
      "loss": 0.0371,
      "num_tokens": 70878378.0,
      "step": 930
    },
    {
      "epoch": 1.1609482220835932,
      "grad_norm": 0.11340208178725925,
      "learning_rate": 7.321638125440872e-06,
      "loss": 0.0361,
      "num_tokens": 70953286.0,
      "step": 931
    },
    {
      "epoch": 1.1621958827199002,
      "grad_norm": 0.12042441254724807,
      "learning_rate": 7.316095521115541e-06,
      "loss": 0.0396,
      "num_tokens": 71029347.0,
      "step": 932
    },
    {
      "epoch": 1.1634435433562071,
      "grad_norm": 0.12180711623275445,
      "learning_rate": 7.310549623660101e-06,
      "loss": 0.0394,
      "num_tokens": 71105407.0,
      "step": 933
    },
    {
      "epoch": 1.164691203992514,
      "grad_norm": 0.1342300366297676,
      "learning_rate": 7.305000443130943e-06,
      "loss": 0.0379,
      "num_tokens": 71181708.0,
      "step": 934
    },
    {
      "epoch": 1.165938864628821,
      "grad_norm": 0.12903385987396218,
      "learning_rate": 7.299447989590406e-06,
      "loss": 0.0419,
      "num_tokens": 71258943.0,
      "step": 935
    },
    {
      "epoch": 1.167186525265128,
      "grad_norm": 0.14085079025904748,
      "learning_rate": 7.293892273106768e-06,
      "loss": 0.0398,
      "num_tokens": 71334191.0,
      "step": 936
    },
    {
      "epoch": 1.1684341859014349,
      "grad_norm": 0.1321285243624536,
      "learning_rate": 7.2883333037542205e-06,
      "loss": 0.0374,
      "num_tokens": 71409277.0,
      "step": 937
    },
    {
      "epoch": 1.1696818465377417,
      "grad_norm": 0.11949825200704543,
      "learning_rate": 7.282771091612858e-06,
      "loss": 0.0374,
      "num_tokens": 71485087.0,
      "step": 938
    },
    {
      "epoch": 1.1709295071740486,
      "grad_norm": 0.13062252929041882,
      "learning_rate": 7.27720564676865e-06,
      "loss": 0.0401,
      "num_tokens": 71561286.0,
      "step": 939
    },
    {
      "epoch": 1.1721771678103556,
      "grad_norm": 0.1477865360798684,
      "learning_rate": 7.271636979313432e-06,
      "loss": 0.0398,
      "num_tokens": 71637031.0,
      "step": 940
    },
    {
      "epoch": 1.1734248284466626,
      "grad_norm": 0.13512349224041548,
      "learning_rate": 7.266065099344881e-06,
      "loss": 0.0404,
      "num_tokens": 71713245.0,
      "step": 941
    },
    {
      "epoch": 1.1746724890829694,
      "grad_norm": 0.14893654503163767,
      "learning_rate": 7.260490016966497e-06,
      "loss": 0.0389,
      "num_tokens": 71789690.0,
      "step": 942
    },
    {
      "epoch": 1.1759201497192764,
      "grad_norm": 0.12893022128126744,
      "learning_rate": 7.2549117422875925e-06,
      "loss": 0.0375,
      "num_tokens": 71864989.0,
      "step": 943
    },
    {
      "epoch": 1.1771678103555834,
      "grad_norm": 0.12549664193940147,
      "learning_rate": 7.249330285423265e-06,
      "loss": 0.04,
      "num_tokens": 71940192.0,
      "step": 944
    },
    {
      "epoch": 1.1784154709918901,
      "grad_norm": 0.12274636302470389,
      "learning_rate": 7.243745656494382e-06,
      "loss": 0.0379,
      "num_tokens": 72016076.0,
      "step": 945
    },
    {
      "epoch": 1.1796631316281971,
      "grad_norm": 0.11710254534160251,
      "learning_rate": 7.238157865627562e-06,
      "loss": 0.0365,
      "num_tokens": 72092219.0,
      "step": 946
    },
    {
      "epoch": 1.1809107922645041,
      "grad_norm": 0.12861913048299303,
      "learning_rate": 7.2325669229551636e-06,
      "loss": 0.0394,
      "num_tokens": 72169088.0,
      "step": 947
    },
    {
      "epoch": 1.182158452900811,
      "grad_norm": 0.11952840728072552,
      "learning_rate": 7.226972838615251e-06,
      "loss": 0.0375,
      "num_tokens": 72245176.0,
      "step": 948
    },
    {
      "epoch": 1.1834061135371179,
      "grad_norm": 0.13750612692860223,
      "learning_rate": 7.221375622751593e-06,
      "loss": 0.0396,
      "num_tokens": 72322210.0,
      "step": 949
    },
    {
      "epoch": 1.1846537741734249,
      "grad_norm": 0.1352346077782367,
      "learning_rate": 7.215775285513633e-06,
      "loss": 0.0384,
      "num_tokens": 72399608.0,
      "step": 950
    },
    {
      "epoch": 1.1859014348097316,
      "grad_norm": 0.11887468445347979,
      "learning_rate": 7.210171837056474e-06,
      "loss": 0.0358,
      "num_tokens": 72476194.0,
      "step": 951
    },
    {
      "epoch": 1.1871490954460386,
      "grad_norm": 0.11588279896504751,
      "learning_rate": 7.2045652875408614e-06,
      "loss": 0.0377,
      "num_tokens": 72552070.0,
      "step": 952
    },
    {
      "epoch": 1.1883967560823456,
      "grad_norm": 0.12255586677955618,
      "learning_rate": 7.198955647133167e-06,
      "loss": 0.0374,
      "num_tokens": 72627986.0,
      "step": 953
    },
    {
      "epoch": 1.1896444167186526,
      "grad_norm": 0.12035533003159458,
      "learning_rate": 7.193342926005362e-06,
      "loss": 0.0363,
      "num_tokens": 72703693.0,
      "step": 954
    },
    {
      "epoch": 1.1908920773549594,
      "grad_norm": 0.1258968428683712,
      "learning_rate": 7.187727134335006e-06,
      "loss": 0.0388,
      "num_tokens": 72780942.0,
      "step": 955
    },
    {
      "epoch": 1.1921397379912664,
      "grad_norm": 0.12871987765476092,
      "learning_rate": 7.182108282305231e-06,
      "loss": 0.0396,
      "num_tokens": 72857465.0,
      "step": 956
    },
    {
      "epoch": 1.1933873986275734,
      "grad_norm": 0.13203954993001707,
      "learning_rate": 7.176486380104707e-06,
      "loss": 0.0406,
      "num_tokens": 72933547.0,
      "step": 957
    },
    {
      "epoch": 1.1946350592638801,
      "grad_norm": 0.12670518611183942,
      "learning_rate": 7.1708614379276485e-06,
      "loss": 0.0374,
      "num_tokens": 73009045.0,
      "step": 958
    },
    {
      "epoch": 1.1958827199001871,
      "grad_norm": 0.13557051840818,
      "learning_rate": 7.165233465973771e-06,
      "loss": 0.0384,
      "num_tokens": 73084864.0,
      "step": 959
    },
    {
      "epoch": 1.1971303805364941,
      "grad_norm": 0.12840733753866795,
      "learning_rate": 7.159602474448292e-06,
      "loss": 0.0387,
      "num_tokens": 73160525.0,
      "step": 960
    },
    {
      "epoch": 1.1983780411728011,
      "grad_norm": 0.11239106491031488,
      "learning_rate": 7.1539684735618995e-06,
      "loss": 0.0349,
      "num_tokens": 73235850.0,
      "step": 961
    },
    {
      "epoch": 1.1996257018091079,
      "grad_norm": 0.12567537645801974,
      "learning_rate": 7.148331473530741e-06,
      "loss": 0.0386,
      "num_tokens": 73310704.0,
      "step": 962
    },
    {
      "epoch": 1.2008733624454149,
      "grad_norm": 0.12585028625143363,
      "learning_rate": 7.142691484576399e-06,
      "loss": 0.0377,
      "num_tokens": 73386704.0,
      "step": 963
    },
    {
      "epoch": 1.2021210230817219,
      "grad_norm": 0.12328153422534764,
      "learning_rate": 7.137048516925882e-06,
      "loss": 0.0404,
      "num_tokens": 73463619.0,
      "step": 964
    },
    {
      "epoch": 1.2033686837180286,
      "grad_norm": 0.13021530210661525,
      "learning_rate": 7.131402580811593e-06,
      "loss": 0.0381,
      "num_tokens": 73538731.0,
      "step": 965
    },
    {
      "epoch": 1.2046163443543356,
      "grad_norm": 0.12202075613733623,
      "learning_rate": 7.125753686471322e-06,
      "loss": 0.0385,
      "num_tokens": 73615795.0,
      "step": 966
    },
    {
      "epoch": 1.2058640049906426,
      "grad_norm": 0.12961790504212412,
      "learning_rate": 7.120101844148222e-06,
      "loss": 0.039,
      "num_tokens": 73692228.0,
      "step": 967
    },
    {
      "epoch": 1.2071116656269494,
      "grad_norm": 0.13052470730079757,
      "learning_rate": 7.1144470640907906e-06,
      "loss": 0.0389,
      "num_tokens": 73768783.0,
      "step": 968
    },
    {
      "epoch": 1.2083593262632564,
      "grad_norm": 0.134664952991497,
      "learning_rate": 7.1087893565528545e-06,
      "loss": 0.038,
      "num_tokens": 73844511.0,
      "step": 969
    },
    {
      "epoch": 1.2096069868995634,
      "grad_norm": 0.11415812603318583,
      "learning_rate": 7.103128731793546e-06,
      "loss": 0.0378,
      "num_tokens": 73920806.0,
      "step": 970
    },
    {
      "epoch": 1.2108546475358701,
      "grad_norm": 0.12776307077120927,
      "learning_rate": 7.097465200077289e-06,
      "loss": 0.0385,
      "num_tokens": 73996188.0,
      "step": 971
    },
    {
      "epoch": 1.2121023081721771,
      "grad_norm": 0.13377749841884895,
      "learning_rate": 7.0917987716737795e-06,
      "loss": 0.0388,
      "num_tokens": 74073219.0,
      "step": 972
    },
    {
      "epoch": 1.2133499688084841,
      "grad_norm": 0.13091768787237276,
      "learning_rate": 7.086129456857963e-06,
      "loss": 0.0373,
      "num_tokens": 74149596.0,
      "step": 973
    },
    {
      "epoch": 1.214597629444791,
      "grad_norm": 0.12906183805216612,
      "learning_rate": 7.080457265910022e-06,
      "loss": 0.0408,
      "num_tokens": 74225555.0,
      "step": 974
    },
    {
      "epoch": 1.2158452900810979,
      "grad_norm": 0.1264473460077226,
      "learning_rate": 7.074782209115356e-06,
      "loss": 0.0377,
      "num_tokens": 74302435.0,
      "step": 975
    },
    {
      "epoch": 1.2170929507174049,
      "grad_norm": 0.12036445295841532,
      "learning_rate": 7.069104296764553e-06,
      "loss": 0.0387,
      "num_tokens": 74379211.0,
      "step": 976
    },
    {
      "epoch": 1.2183406113537119,
      "grad_norm": 0.1365334912475307,
      "learning_rate": 7.0634235391533874e-06,
      "loss": 0.0356,
      "num_tokens": 74454460.0,
      "step": 977
    },
    {
      "epoch": 1.2195882719900186,
      "grad_norm": 0.11930134399444615,
      "learning_rate": 7.05773994658279e-06,
      "loss": 0.0394,
      "num_tokens": 74531587.0,
      "step": 978
    },
    {
      "epoch": 1.2208359326263256,
      "grad_norm": 0.12394477513731392,
      "learning_rate": 7.052053529358831e-06,
      "loss": 0.0382,
      "num_tokens": 74607528.0,
      "step": 979
    },
    {
      "epoch": 1.2220835932626326,
      "grad_norm": 0.12312553920411257,
      "learning_rate": 7.046364297792703e-06,
      "loss": 0.0363,
      "num_tokens": 74683276.0,
      "step": 980
    },
    {
      "epoch": 1.2233312538989396,
      "grad_norm": 0.13785776039233755,
      "learning_rate": 7.040672262200705e-06,
      "loss": 0.0385,
      "num_tokens": 74759411.0,
      "step": 981
    },
    {
      "epoch": 1.2245789145352464,
      "grad_norm": 0.11472386239465589,
      "learning_rate": 7.0349774329042135e-06,
      "loss": 0.0368,
      "num_tokens": 74835216.0,
      "step": 982
    },
    {
      "epoch": 1.2258265751715534,
      "grad_norm": 0.12173051063783977,
      "learning_rate": 7.02927982022968e-06,
      "loss": 0.0396,
      "num_tokens": 74911974.0,
      "step": 983
    },
    {
      "epoch": 1.2270742358078603,
      "grad_norm": 0.12457778411901807,
      "learning_rate": 7.023579434508596e-06,
      "loss": 0.041,
      "num_tokens": 74988340.0,
      "step": 984
    },
    {
      "epoch": 1.2283218964441671,
      "grad_norm": 0.12731680919814084,
      "learning_rate": 7.017876286077484e-06,
      "loss": 0.0391,
      "num_tokens": 75064637.0,
      "step": 985
    },
    {
      "epoch": 1.229569557080474,
      "grad_norm": 0.12533235166622664,
      "learning_rate": 7.012170385277877e-06,
      "loss": 0.039,
      "num_tokens": 75140639.0,
      "step": 986
    },
    {
      "epoch": 1.230817217716781,
      "grad_norm": 0.11252297683062269,
      "learning_rate": 7.006461742456297e-06,
      "loss": 0.0377,
      "num_tokens": 75216537.0,
      "step": 987
    },
    {
      "epoch": 1.2320648783530879,
      "grad_norm": 0.11919903473585805,
      "learning_rate": 7.000750367964239e-06,
      "loss": 0.0403,
      "num_tokens": 75293221.0,
      "step": 988
    },
    {
      "epoch": 1.2333125389893949,
      "grad_norm": 0.11769788430742996,
      "learning_rate": 6.99503627215815e-06,
      "loss": 0.0372,
      "num_tokens": 75368699.0,
      "step": 989
    },
    {
      "epoch": 1.2345601996257018,
      "grad_norm": 0.1279368792464733,
      "learning_rate": 6.989319465399415e-06,
      "loss": 0.0387,
      "num_tokens": 75444397.0,
      "step": 990
    },
    {
      "epoch": 1.2358078602620086,
      "grad_norm": 0.12159070141319064,
      "learning_rate": 6.983599958054331e-06,
      "loss": 0.0376,
      "num_tokens": 75519906.0,
      "step": 991
    },
    {
      "epoch": 1.2370555208983156,
      "grad_norm": 0.12029451059502534,
      "learning_rate": 6.977877760494094e-06,
      "loss": 0.04,
      "num_tokens": 75595546.0,
      "step": 992
    },
    {
      "epoch": 1.2383031815346226,
      "grad_norm": 0.13069178124066883,
      "learning_rate": 6.972152883094778e-06,
      "loss": 0.0378,
      "num_tokens": 75672691.0,
      "step": 993
    },
    {
      "epoch": 1.2395508421709296,
      "grad_norm": 0.12074943091247309,
      "learning_rate": 6.966425336237317e-06,
      "loss": 0.0371,
      "num_tokens": 75748316.0,
      "step": 994
    },
    {
      "epoch": 1.2407985028072364,
      "grad_norm": 0.1289634381789005,
      "learning_rate": 6.960695130307484e-06,
      "loss": 0.041,
      "num_tokens": 75824693.0,
      "step": 995
    },
    {
      "epoch": 1.2420461634435433,
      "grad_norm": 0.1256483249863802,
      "learning_rate": 6.954962275695871e-06,
      "loss": 0.04,
      "num_tokens": 75901978.0,
      "step": 996
    },
    {
      "epoch": 1.2432938240798503,
      "grad_norm": 0.12999985473947567,
      "learning_rate": 6.9492267827978824e-06,
      "loss": 0.0375,
      "num_tokens": 75978741.0,
      "step": 997
    },
    {
      "epoch": 1.244541484716157,
      "grad_norm": 0.11847867937232391,
      "learning_rate": 6.943488662013697e-06,
      "loss": 0.0375,
      "num_tokens": 76054336.0,
      "step": 998
    },
    {
      "epoch": 1.245789145352464,
      "grad_norm": 0.12241451988551497,
      "learning_rate": 6.93774792374826e-06,
      "loss": 0.0388,
      "num_tokens": 76131145.0,
      "step": 999
    },
    {
      "epoch": 1.247036805988771,
      "grad_norm": 0.13576458392659807,
      "learning_rate": 6.93200457841127e-06,
      "loss": 0.0371,
      "num_tokens": 76207946.0,
      "step": 1000
    },
    {
      "epoch": 1.248284466625078,
      "grad_norm": 0.1357083738319804,
      "learning_rate": 6.9262586364171455e-06,
      "loss": 0.0407,
      "num_tokens": 76284678.0,
      "step": 1001
    },
    {
      "epoch": 1.2495321272613849,
      "grad_norm": 0.1417184857589529,
      "learning_rate": 6.920510108185016e-06,
      "loss": 0.0369,
      "num_tokens": 76361020.0,
      "step": 1002
    },
    {
      "epoch": 1.2507797878976918,
      "grad_norm": 0.12259900282643452,
      "learning_rate": 6.9147590041387e-06,
      "loss": 0.0371,
      "num_tokens": 76437197.0,
      "step": 1003
    },
    {
      "epoch": 1.2520274485339988,
      "grad_norm": 0.11833456048304121,
      "learning_rate": 6.909005334706688e-06,
      "loss": 0.0387,
      "num_tokens": 76513600.0,
      "step": 1004
    },
    {
      "epoch": 1.2532751091703056,
      "grad_norm": 0.1362755912509239,
      "learning_rate": 6.903249110322123e-06,
      "loss": 0.0461,
      "num_tokens": 76590862.0,
      "step": 1005
    },
    {
      "epoch": 1.2545227698066126,
      "grad_norm": 0.12371120963572597,
      "learning_rate": 6.897490341422779e-06,
      "loss": 0.036,
      "num_tokens": 76667344.0,
      "step": 1006
    },
    {
      "epoch": 1.2557704304429196,
      "grad_norm": 0.12643531689407858,
      "learning_rate": 6.8917290384510435e-06,
      "loss": 0.0391,
      "num_tokens": 76743875.0,
      "step": 1007
    },
    {
      "epoch": 1.2570180910792264,
      "grad_norm": 0.13473742006949435,
      "learning_rate": 6.885965211853902e-06,
      "loss": 0.0459,
      "num_tokens": 76819364.0,
      "step": 1008
    },
    {
      "epoch": 1.2582657517155333,
      "grad_norm": 0.13094378837230844,
      "learning_rate": 6.8801988720829134e-06,
      "loss": 0.0377,
      "num_tokens": 76894778.0,
      "step": 1009
    },
    {
      "epoch": 1.2595134123518403,
      "grad_norm": 0.13013560077380523,
      "learning_rate": 6.874430029594194e-06,
      "loss": 0.0382,
      "num_tokens": 76971000.0,
      "step": 1010
    },
    {
      "epoch": 1.260761072988147,
      "grad_norm": 0.11496471028200521,
      "learning_rate": 6.8686586948483995e-06,
      "loss": 0.0378,
      "num_tokens": 77046773.0,
      "step": 1011
    },
    {
      "epoch": 1.262008733624454,
      "grad_norm": 0.12396329610612375,
      "learning_rate": 6.862884878310705e-06,
      "loss": 0.0381,
      "num_tokens": 77122750.0,
      "step": 1012
    },
    {
      "epoch": 1.263256394260761,
      "grad_norm": 0.12570886085428049,
      "learning_rate": 6.8571085904507825e-06,
      "loss": 0.0368,
      "num_tokens": 77198494.0,
      "step": 1013
    },
    {
      "epoch": 1.264504054897068,
      "grad_norm": 0.11028117409017361,
      "learning_rate": 6.8513298417427895e-06,
      "loss": 0.0389,
      "num_tokens": 77273307.0,
      "step": 1014
    },
    {
      "epoch": 1.2657517155333748,
      "grad_norm": 0.1471893491163991,
      "learning_rate": 6.845548642665347e-06,
      "loss": 0.0388,
      "num_tokens": 77349388.0,
      "step": 1015
    },
    {
      "epoch": 1.2669993761696818,
      "grad_norm": 0.11262904086099527,
      "learning_rate": 6.839765003701511e-06,
      "loss": 0.0379,
      "num_tokens": 77426148.0,
      "step": 1016
    },
    {
      "epoch": 1.2682470368059888,
      "grad_norm": 0.13361856137063885,
      "learning_rate": 6.833978935338772e-06,
      "loss": 0.0379,
      "num_tokens": 77501759.0,
      "step": 1017
    },
    {
      "epoch": 1.2694946974422958,
      "grad_norm": 0.12279484859518144,
      "learning_rate": 6.828190448069016e-06,
      "loss": 0.0406,
      "num_tokens": 77579480.0,
      "step": 1018
    },
    {
      "epoch": 1.2707423580786026,
      "grad_norm": 0.12242840320310547,
      "learning_rate": 6.822399552388523e-06,
      "loss": 0.0359,
      "num_tokens": 77655675.0,
      "step": 1019
    },
    {
      "epoch": 1.2719900187149096,
      "grad_norm": 0.116893287227689,
      "learning_rate": 6.816606258797936e-06,
      "loss": 0.0381,
      "num_tokens": 77733976.0,
      "step": 1020
    },
    {
      "epoch": 1.2732376793512166,
      "grad_norm": 0.1222409289259211,
      "learning_rate": 6.810810577802249e-06,
      "loss": 0.0378,
      "num_tokens": 77809768.0,
      "step": 1021
    },
    {
      "epoch": 1.2744853399875233,
      "grad_norm": 0.12443386098402114,
      "learning_rate": 6.8050125199107835e-06,
      "loss": 0.0383,
      "num_tokens": 77886283.0,
      "step": 1022
    },
    {
      "epoch": 1.2757330006238303,
      "grad_norm": 0.12792841056094306,
      "learning_rate": 6.799212095637169e-06,
      "loss": 0.038,
      "num_tokens": 77962705.0,
      "step": 1023
    },
    {
      "epoch": 1.2769806612601373,
      "grad_norm": 0.12853700012331887,
      "learning_rate": 6.7934093154993285e-06,
      "loss": 0.0381,
      "num_tokens": 78038286.0,
      "step": 1024
    },
    {
      "epoch": 1.278228321896444,
      "grad_norm": 0.1204293670396625,
      "learning_rate": 6.787604190019456e-06,
      "loss": 0.0379,
      "num_tokens": 78113897.0,
      "step": 1025
    },
    {
      "epoch": 1.279475982532751,
      "grad_norm": 0.16278447289558357,
      "learning_rate": 6.781796729724001e-06,
      "loss": 0.039,
      "num_tokens": 78190823.0,
      "step": 1026
    },
    {
      "epoch": 1.280723643169058,
      "grad_norm": 0.14049403625476514,
      "learning_rate": 6.775986945143641e-06,
      "loss": 0.0365,
      "num_tokens": 78266929.0,
      "step": 1027
    },
    {
      "epoch": 1.2819713038053648,
      "grad_norm": 0.12034977499671139,
      "learning_rate": 6.770174846813273e-06,
      "loss": 0.0361,
      "num_tokens": 78343744.0,
      "step": 1028
    },
    {
      "epoch": 1.2832189644416718,
      "grad_norm": 0.11781958413703834,
      "learning_rate": 6.7643604452719894e-06,
      "loss": 0.0359,
      "num_tokens": 78418807.0,
      "step": 1029
    },
    {
      "epoch": 1.2844666250779788,
      "grad_norm": 0.12121542127811726,
      "learning_rate": 6.758543751063055e-06,
      "loss": 0.0382,
      "num_tokens": 78496287.0,
      "step": 1030
    },
    {
      "epoch": 1.2857142857142856,
      "grad_norm": 0.11962328271317485,
      "learning_rate": 6.752724774733899e-06,
      "loss": 0.0395,
      "num_tokens": 78571959.0,
      "step": 1031
    },
    {
      "epoch": 1.2869619463505926,
      "grad_norm": 0.11711191648751425,
      "learning_rate": 6.746903526836079e-06,
      "loss": 0.0368,
      "num_tokens": 78647917.0,
      "step": 1032
    },
    {
      "epoch": 1.2882096069868996,
      "grad_norm": 0.11174765287606012,
      "learning_rate": 6.741080017925279e-06,
      "loss": 0.038,
      "num_tokens": 78723022.0,
      "step": 1033
    },
    {
      "epoch": 1.2894572676232066,
      "grad_norm": 0.14213629513727627,
      "learning_rate": 6.735254258561281e-06,
      "loss": 0.041,
      "num_tokens": 78799598.0,
      "step": 1034
    },
    {
      "epoch": 1.2907049282595136,
      "grad_norm": 0.11404088303991229,
      "learning_rate": 6.729426259307948e-06,
      "loss": 0.035,
      "num_tokens": 78875225.0,
      "step": 1035
    },
    {
      "epoch": 1.2919525888958203,
      "grad_norm": 0.1259336273419687,
      "learning_rate": 6.723596030733204e-06,
      "loss": 0.0408,
      "num_tokens": 78953031.0,
      "step": 1036
    },
    {
      "epoch": 1.2932002495321273,
      "grad_norm": 0.15791825227165843,
      "learning_rate": 6.717763583409016e-06,
      "loss": 0.0614,
      "num_tokens": 79031801.0,
      "step": 1037
    },
    {
      "epoch": 1.2944479101684343,
      "grad_norm": 0.1304087409910658,
      "learning_rate": 6.711928927911373e-06,
      "loss": 0.0387,
      "num_tokens": 79107509.0,
      "step": 1038
    },
    {
      "epoch": 1.295695570804741,
      "grad_norm": 0.12600021318695712,
      "learning_rate": 6.7060920748202674e-06,
      "loss": 0.0346,
      "num_tokens": 79184034.0,
      "step": 1039
    },
    {
      "epoch": 1.296943231441048,
      "grad_norm": 0.12435271412670368,
      "learning_rate": 6.700253034719684e-06,
      "loss": 0.0381,
      "num_tokens": 79263280.0,
      "step": 1040
    },
    {
      "epoch": 1.298190892077355,
      "grad_norm": 0.11823691658700941,
      "learning_rate": 6.694411818197561e-06,
      "loss": 0.0418,
      "num_tokens": 79339787.0,
      "step": 1041
    },
    {
      "epoch": 1.2994385527136618,
      "grad_norm": 0.13713072747764696,
      "learning_rate": 6.688568435845792e-06,
      "loss": 0.0386,
      "num_tokens": 79414932.0,
      "step": 1042
    },
    {
      "epoch": 1.3006862133499688,
      "grad_norm": 0.12634187217101714,
      "learning_rate": 6.682722898260195e-06,
      "loss": 0.0377,
      "num_tokens": 79491173.0,
      "step": 1043
    },
    {
      "epoch": 1.3019338739862758,
      "grad_norm": 0.12086406055618043,
      "learning_rate": 6.676875216040498e-06,
      "loss": 0.0373,
      "num_tokens": 79567357.0,
      "step": 1044
    },
    {
      "epoch": 1.3031815346225826,
      "grad_norm": 0.11748638779428121,
      "learning_rate": 6.671025399790315e-06,
      "loss": 0.0349,
      "num_tokens": 79643546.0,
      "step": 1045
    },
    {
      "epoch": 1.3044291952588896,
      "grad_norm": 0.12900997225280608,
      "learning_rate": 6.66517346011713e-06,
      "loss": 0.0374,
      "num_tokens": 79719107.0,
      "step": 1046
    },
    {
      "epoch": 1.3056768558951966,
      "grad_norm": 0.1132473938030605,
      "learning_rate": 6.659319407632282e-06,
      "loss": 0.0391,
      "num_tokens": 79795324.0,
      "step": 1047
    },
    {
      "epoch": 1.3069245165315033,
      "grad_norm": 0.12056242211992067,
      "learning_rate": 6.653463252950933e-06,
      "loss": 0.0381,
      "num_tokens": 79872580.0,
      "step": 1048
    },
    {
      "epoch": 1.3081721771678103,
      "grad_norm": 0.12298594861515041,
      "learning_rate": 6.647605006692066e-06,
      "loss": 0.0372,
      "num_tokens": 79948484.0,
      "step": 1049
    },
    {
      "epoch": 1.3094198378041173,
      "grad_norm": 0.1224135552748089,
      "learning_rate": 6.641744679478448e-06,
      "loss": 0.0385,
      "num_tokens": 80024409.0,
      "step": 1050
    },
    {
      "epoch": 1.310667498440424,
      "grad_norm": 0.1238426241135653,
      "learning_rate": 6.635882281936625e-06,
      "loss": 0.0369,
      "num_tokens": 80100285.0,
      "step": 1051
    },
    {
      "epoch": 1.311915159076731,
      "grad_norm": 0.12678498517610146,
      "learning_rate": 6.630017824696898e-06,
      "loss": 0.0374,
      "num_tokens": 80175891.0,
      "step": 1052
    },
    {
      "epoch": 1.313162819713038,
      "grad_norm": 0.11416544137866638,
      "learning_rate": 6.624151318393298e-06,
      "loss": 0.0364,
      "num_tokens": 80250985.0,
      "step": 1053
    },
    {
      "epoch": 1.314410480349345,
      "grad_norm": 0.1186845318691531,
      "learning_rate": 6.618282773663576e-06,
      "loss": 0.0369,
      "num_tokens": 80326912.0,
      "step": 1054
    },
    {
      "epoch": 1.315658140985652,
      "grad_norm": 0.11768290749010996,
      "learning_rate": 6.612412201149175e-06,
      "loss": 0.0356,
      "num_tokens": 80402020.0,
      "step": 1055
    },
    {
      "epoch": 1.3169058016219588,
      "grad_norm": 0.11763439527023663,
      "learning_rate": 6.6065396114952195e-06,
      "loss": 0.0358,
      "num_tokens": 80477106.0,
      "step": 1056
    },
    {
      "epoch": 1.3181534622582658,
      "grad_norm": 0.10917918873769969,
      "learning_rate": 6.600665015350487e-06,
      "loss": 0.0362,
      "num_tokens": 80552797.0,
      "step": 1057
    },
    {
      "epoch": 1.3194011228945728,
      "grad_norm": 0.1297261226982308,
      "learning_rate": 6.594788423367399e-06,
      "loss": 0.0379,
      "num_tokens": 80630455.0,
      "step": 1058
    },
    {
      "epoch": 1.3206487835308796,
      "grad_norm": 0.1357107428559246,
      "learning_rate": 6.588909846201992e-06,
      "loss": 0.0359,
      "num_tokens": 80706916.0,
      "step": 1059
    },
    {
      "epoch": 1.3218964441671865,
      "grad_norm": 0.11167341542497014,
      "learning_rate": 6.583029294513902e-06,
      "loss": 0.0382,
      "num_tokens": 80783306.0,
      "step": 1060
    },
    {
      "epoch": 1.3231441048034935,
      "grad_norm": 0.126267003510051,
      "learning_rate": 6.577146778966347e-06,
      "loss": 0.0368,
      "num_tokens": 80858263.0,
      "step": 1061
    },
    {
      "epoch": 1.3243917654398003,
      "grad_norm": 0.11755835211629956,
      "learning_rate": 6.571262310226108e-06,
      "loss": 0.0363,
      "num_tokens": 80934605.0,
      "step": 1062
    },
    {
      "epoch": 1.3256394260761073,
      "grad_norm": 0.12868435000077522,
      "learning_rate": 6.565375898963503e-06,
      "loss": 0.0389,
      "num_tokens": 81010777.0,
      "step": 1063
    },
    {
      "epoch": 1.3268870867124143,
      "grad_norm": 0.13643865231728688,
      "learning_rate": 6.5594875558523755e-06,
      "loss": 0.0375,
      "num_tokens": 81086979.0,
      "step": 1064
    },
    {
      "epoch": 1.328134747348721,
      "grad_norm": 0.11784928737975484,
      "learning_rate": 6.553597291570071e-06,
      "loss": 0.0383,
      "num_tokens": 81162843.0,
      "step": 1065
    },
    {
      "epoch": 1.329382407985028,
      "grad_norm": 0.14182537453632305,
      "learning_rate": 6.547705116797422e-06,
      "loss": 0.0402,
      "num_tokens": 81239935.0,
      "step": 1066
    },
    {
      "epoch": 1.330630068621335,
      "grad_norm": 0.12586204808908688,
      "learning_rate": 6.5418110422187156e-06,
      "loss": 0.0383,
      "num_tokens": 81316699.0,
      "step": 1067
    },
    {
      "epoch": 1.3318777292576418,
      "grad_norm": 0.12092515001770596,
      "learning_rate": 6.535915078521697e-06,
      "loss": 0.0405,
      "num_tokens": 81392842.0,
      "step": 1068
    },
    {
      "epoch": 1.3331253898939488,
      "grad_norm": 0.12527079685312903,
      "learning_rate": 6.530017236397529e-06,
      "loss": 0.0385,
      "num_tokens": 81468914.0,
      "step": 1069
    },
    {
      "epoch": 1.3343730505302558,
      "grad_norm": 0.12405547111378336,
      "learning_rate": 6.52411752654078e-06,
      "loss": 0.0403,
      "num_tokens": 81546381.0,
      "step": 1070
    },
    {
      "epoch": 1.3356207111665626,
      "grad_norm": 0.12043411317614498,
      "learning_rate": 6.518215959649409e-06,
      "loss": 0.036,
      "num_tokens": 81621864.0,
      "step": 1071
    },
    {
      "epoch": 1.3368683718028695,
      "grad_norm": 0.11269431565422956,
      "learning_rate": 6.512312546424739e-06,
      "loss": 0.0367,
      "num_tokens": 81697399.0,
      "step": 1072
    },
    {
      "epoch": 1.3381160324391765,
      "grad_norm": 0.131008457319013,
      "learning_rate": 6.506407297571445e-06,
      "loss": 0.0387,
      "num_tokens": 81773914.0,
      "step": 1073
    },
    {
      "epoch": 1.3393636930754835,
      "grad_norm": 0.12914008887166922,
      "learning_rate": 6.500500223797526e-06,
      "loss": 0.0394,
      "num_tokens": 81851041.0,
      "step": 1074
    },
    {
      "epoch": 1.3406113537117905,
      "grad_norm": 0.11969733474999027,
      "learning_rate": 6.494591335814292e-06,
      "loss": 0.0373,
      "num_tokens": 81926575.0,
      "step": 1075
    },
    {
      "epoch": 1.3418590143480973,
      "grad_norm": 0.1256605085911434,
      "learning_rate": 6.488680644336344e-06,
      "loss": 0.0356,
      "num_tokens": 82001949.0,
      "step": 1076
    },
    {
      "epoch": 1.3431066749844043,
      "grad_norm": 0.12072878078758512,
      "learning_rate": 6.482768160081553e-06,
      "loss": 0.0372,
      "num_tokens": 82082965.0,
      "step": 1077
    },
    {
      "epoch": 1.3443543356207113,
      "grad_norm": 0.11564706297590938,
      "learning_rate": 6.4768538937710364e-06,
      "loss": 0.0364,
      "num_tokens": 82159572.0,
      "step": 1078
    },
    {
      "epoch": 1.345601996257018,
      "grad_norm": 0.11810783963495697,
      "learning_rate": 6.470937856129152e-06,
      "loss": 0.0362,
      "num_tokens": 82236392.0,
      "step": 1079
    },
    {
      "epoch": 1.346849656893325,
      "grad_norm": 0.12243552939696348,
      "learning_rate": 6.465020057883461e-06,
      "loss": 0.0402,
      "num_tokens": 82312762.0,
      "step": 1080
    },
    {
      "epoch": 1.348097317529632,
      "grad_norm": 0.12176519886783935,
      "learning_rate": 6.45910050976472e-06,
      "loss": 0.0356,
      "num_tokens": 82388841.0,
      "step": 1081
    },
    {
      "epoch": 1.3493449781659388,
      "grad_norm": 0.10978489412644651,
      "learning_rate": 6.45317922250686e-06,
      "loss": 0.0356,
      "num_tokens": 82464488.0,
      "step": 1082
    },
    {
      "epoch": 1.3505926388022458,
      "grad_norm": 0.11917674192716878,
      "learning_rate": 6.447256206846963e-06,
      "loss": 0.0385,
      "num_tokens": 82541359.0,
      "step": 1083
    },
    {
      "epoch": 1.3518402994385528,
      "grad_norm": 0.12641710504405018,
      "learning_rate": 6.44133147352525e-06,
      "loss": 0.0389,
      "num_tokens": 82620545.0,
      "step": 1084
    },
    {
      "epoch": 1.3530879600748595,
      "grad_norm": 0.1235769738581032,
      "learning_rate": 6.4354050332850505e-06,
      "loss": 0.0365,
      "num_tokens": 82696305.0,
      "step": 1085
    },
    {
      "epoch": 1.3543356207111665,
      "grad_norm": 0.11868272459465616,
      "learning_rate": 6.429476896872793e-06,
      "loss": 0.0366,
      "num_tokens": 82772823.0,
      "step": 1086
    },
    {
      "epoch": 1.3555832813474735,
      "grad_norm": 0.11716548574142212,
      "learning_rate": 6.4235470750379794e-06,
      "loss": 0.0378,
      "num_tokens": 82848581.0,
      "step": 1087
    },
    {
      "epoch": 1.3568309419837803,
      "grad_norm": 0.12977983514242444,
      "learning_rate": 6.4176155785331705e-06,
      "loss": 0.0373,
      "num_tokens": 82922803.0,
      "step": 1088
    },
    {
      "epoch": 1.3580786026200873,
      "grad_norm": 0.11983262533415129,
      "learning_rate": 6.411682418113961e-06,
      "loss": 0.0369,
      "num_tokens": 82999543.0,
      "step": 1089
    },
    {
      "epoch": 1.3593262632563943,
      "grad_norm": 0.10881217167861341,
      "learning_rate": 6.405747604538965e-06,
      "loss": 0.0391,
      "num_tokens": 83075716.0,
      "step": 1090
    },
    {
      "epoch": 1.3605739238927013,
      "grad_norm": 0.11312143319312479,
      "learning_rate": 6.399811148569794e-06,
      "loss": 0.037,
      "num_tokens": 83152235.0,
      "step": 1091
    },
    {
      "epoch": 1.361821584529008,
      "grad_norm": 0.12868490609288766,
      "learning_rate": 6.393873060971036e-06,
      "loss": 0.0382,
      "num_tokens": 83227490.0,
      "step": 1092
    },
    {
      "epoch": 1.363069245165315,
      "grad_norm": 0.11394818001578064,
      "learning_rate": 6.3879333525102375e-06,
      "loss": 0.0364,
      "num_tokens": 83304281.0,
      "step": 1093
    },
    {
      "epoch": 1.364316905801622,
      "grad_norm": 0.13139308054317986,
      "learning_rate": 6.381992033957889e-06,
      "loss": 0.0369,
      "num_tokens": 83381463.0,
      "step": 1094
    },
    {
      "epoch": 1.365564566437929,
      "grad_norm": 0.12142619982158942,
      "learning_rate": 6.376049116087393e-06,
      "loss": 0.0412,
      "num_tokens": 83458122.0,
      "step": 1095
    },
    {
      "epoch": 1.3668122270742358,
      "grad_norm": 0.12819557314527608,
      "learning_rate": 6.370104609675058e-06,
      "loss": 0.0375,
      "num_tokens": 83534927.0,
      "step": 1096
    },
    {
      "epoch": 1.3680598877105428,
      "grad_norm": 0.13688859870267972,
      "learning_rate": 6.364158525500069e-06,
      "loss": 0.0377,
      "num_tokens": 83611556.0,
      "step": 1097
    },
    {
      "epoch": 1.3693075483468498,
      "grad_norm": 0.1139664206178106,
      "learning_rate": 6.358210874344476e-06,
      "loss": 0.0377,
      "num_tokens": 83688500.0,
      "step": 1098
    },
    {
      "epoch": 1.3705552089831565,
      "grad_norm": 0.12878963288602277,
      "learning_rate": 6.352261666993167e-06,
      "loss": 0.0398,
      "num_tokens": 83764887.0,
      "step": 1099
    },
    {
      "epoch": 1.3718028696194635,
      "grad_norm": 0.13187044333532477,
      "learning_rate": 6.346310914233854e-06,
      "loss": 0.0402,
      "num_tokens": 83840573.0,
      "step": 1100
    },
    {
      "epoch": 1.3730505302557705,
      "grad_norm": 0.13218470782925437,
      "learning_rate": 6.340358626857049e-06,
      "loss": 0.0369,
      "num_tokens": 83915835.0,
      "step": 1101
    },
    {
      "epoch": 1.3742981908920773,
      "grad_norm": 0.123600988524384,
      "learning_rate": 6.334404815656049e-06,
      "loss": 0.037,
      "num_tokens": 83993107.0,
      "step": 1102
    },
    {
      "epoch": 1.3755458515283843,
      "grad_norm": 0.1141887638770041,
      "learning_rate": 6.328449491426914e-06,
      "loss": 0.0364,
      "num_tokens": 84069739.0,
      "step": 1103
    },
    {
      "epoch": 1.3767935121646913,
      "grad_norm": 0.11853940734782815,
      "learning_rate": 6.322492664968446e-06,
      "loss": 0.0376,
      "num_tokens": 84145150.0,
      "step": 1104
    },
    {
      "epoch": 1.378041172800998,
      "grad_norm": 0.12448334472309965,
      "learning_rate": 6.316534347082173e-06,
      "loss": 0.0369,
      "num_tokens": 84221521.0,
      "step": 1105
    },
    {
      "epoch": 1.379288833437305,
      "grad_norm": 0.11289826691946864,
      "learning_rate": 6.310574548572325e-06,
      "loss": 0.0339,
      "num_tokens": 84296673.0,
      "step": 1106
    },
    {
      "epoch": 1.380536494073612,
      "grad_norm": 0.11584642156593683,
      "learning_rate": 6.304613280245816e-06,
      "loss": 0.0368,
      "num_tokens": 84373020.0,
      "step": 1107
    },
    {
      "epoch": 1.3817841547099188,
      "grad_norm": 0.11898007042409253,
      "learning_rate": 6.298650552912233e-06,
      "loss": 0.0383,
      "num_tokens": 84449012.0,
      "step": 1108
    },
    {
      "epoch": 1.3830318153462258,
      "grad_norm": 0.13178915624086404,
      "learning_rate": 6.292686377383797e-06,
      "loss": 0.0379,
      "num_tokens": 84525080.0,
      "step": 1109
    },
    {
      "epoch": 1.3842794759825328,
      "grad_norm": 0.12485077614334157,
      "learning_rate": 6.286720764475365e-06,
      "loss": 0.039,
      "num_tokens": 84602186.0,
      "step": 1110
    },
    {
      "epoch": 1.3855271366188397,
      "grad_norm": 0.12594599877140716,
      "learning_rate": 6.280753725004395e-06,
      "loss": 0.04,
      "num_tokens": 84677075.0,
      "step": 1111
    },
    {
      "epoch": 1.3867747972551465,
      "grad_norm": 0.12855159432742083,
      "learning_rate": 6.274785269790932e-06,
      "loss": 0.0368,
      "num_tokens": 84752554.0,
      "step": 1112
    },
    {
      "epoch": 1.3880224578914535,
      "grad_norm": 0.11427665618225598,
      "learning_rate": 6.268815409657592e-06,
      "loss": 0.0373,
      "num_tokens": 84828980.0,
      "step": 1113
    },
    {
      "epoch": 1.3892701185277605,
      "grad_norm": 0.11937242460768584,
      "learning_rate": 6.262844155429533e-06,
      "loss": 0.0373,
      "num_tokens": 84905393.0,
      "step": 1114
    },
    {
      "epoch": 1.3905177791640675,
      "grad_norm": 0.12860043075057506,
      "learning_rate": 6.256871517934445e-06,
      "loss": 0.0374,
      "num_tokens": 84982169.0,
      "step": 1115
    },
    {
      "epoch": 1.3917654398003743,
      "grad_norm": 0.12212869867413878,
      "learning_rate": 6.2508975080025254e-06,
      "loss": 0.0388,
      "num_tokens": 85059133.0,
      "step": 1116
    },
    {
      "epoch": 1.3930131004366813,
      "grad_norm": 0.11678301135662582,
      "learning_rate": 6.24492213646646e-06,
      "loss": 0.0392,
      "num_tokens": 85135719.0,
      "step": 1117
    },
    {
      "epoch": 1.3942607610729882,
      "grad_norm": 0.12535257252564777,
      "learning_rate": 6.2389454141614024e-06,
      "loss": 0.0381,
      "num_tokens": 85211509.0,
      "step": 1118
    },
    {
      "epoch": 1.395508421709295,
      "grad_norm": 0.13291682129274268,
      "learning_rate": 6.232967351924959e-06,
      "loss": 0.0414,
      "num_tokens": 85287979.0,
      "step": 1119
    },
    {
      "epoch": 1.396756082345602,
      "grad_norm": 0.12295637500049969,
      "learning_rate": 6.226987960597161e-06,
      "loss": 0.0366,
      "num_tokens": 85364645.0,
      "step": 1120
    },
    {
      "epoch": 1.398003742981909,
      "grad_norm": 0.11736989037967782,
      "learning_rate": 6.22100725102045e-06,
      "loss": 0.0365,
      "num_tokens": 85440252.0,
      "step": 1121
    },
    {
      "epoch": 1.3992514036182158,
      "grad_norm": 0.11734689556837297,
      "learning_rate": 6.215025234039667e-06,
      "loss": 0.0381,
      "num_tokens": 85517224.0,
      "step": 1122
    },
    {
      "epoch": 1.4004990642545228,
      "grad_norm": 0.12117102979297299,
      "learning_rate": 6.209041920502012e-06,
      "loss": 0.041,
      "num_tokens": 85593781.0,
      "step": 1123
    },
    {
      "epoch": 1.4017467248908297,
      "grad_norm": 0.12125413473547068,
      "learning_rate": 6.203057321257041e-06,
      "loss": 0.0394,
      "num_tokens": 85669972.0,
      "step": 1124
    },
    {
      "epoch": 1.4029943855271365,
      "grad_norm": 0.12806121362449266,
      "learning_rate": 6.197071447156643e-06,
      "loss": 0.0429,
      "num_tokens": 85748190.0,
      "step": 1125
    },
    {
      "epoch": 1.4042420461634435,
      "grad_norm": 0.13610898293273987,
      "learning_rate": 6.191084309055018e-06,
      "loss": 0.0393,
      "num_tokens": 85824528.0,
      "step": 1126
    },
    {
      "epoch": 1.4054897067997505,
      "grad_norm": 0.1338585027234358,
      "learning_rate": 6.185095917808654e-06,
      "loss": 0.0385,
      "num_tokens": 85901267.0,
      "step": 1127
    },
    {
      "epoch": 1.4067373674360573,
      "grad_norm": 0.1224560732061581,
      "learning_rate": 6.179106284276315e-06,
      "loss": 0.0373,
      "num_tokens": 85976334.0,
      "step": 1128
    },
    {
      "epoch": 1.4079850280723643,
      "grad_norm": 0.12016867501292519,
      "learning_rate": 6.173115419319019e-06,
      "loss": 0.035,
      "num_tokens": 86050677.0,
      "step": 1129
    },
    {
      "epoch": 1.4092326887086712,
      "grad_norm": 0.1159217234754072,
      "learning_rate": 6.167123333800014e-06,
      "loss": 0.0372,
      "num_tokens": 86127558.0,
      "step": 1130
    },
    {
      "epoch": 1.4104803493449782,
      "grad_norm": 0.10442739115930703,
      "learning_rate": 6.161130038584762e-06,
      "loss": 0.0369,
      "num_tokens": 86204395.0,
      "step": 1131
    },
    {
      "epoch": 1.4117280099812852,
      "grad_norm": 0.11747181262627125,
      "learning_rate": 6.155135544540917e-06,
      "loss": 0.036,
      "num_tokens": 86280341.0,
      "step": 1132
    },
    {
      "epoch": 1.412975670617592,
      "grad_norm": 0.12634337046270444,
      "learning_rate": 6.1491398625383116e-06,
      "loss": 0.0379,
      "num_tokens": 86356632.0,
      "step": 1133
    },
    {
      "epoch": 1.414223331253899,
      "grad_norm": 0.12959775986090372,
      "learning_rate": 6.143143003448929e-06,
      "loss": 0.0399,
      "num_tokens": 86433774.0,
      "step": 1134
    },
    {
      "epoch": 1.415470991890206,
      "grad_norm": 0.11959740448330311,
      "learning_rate": 6.1371449781468835e-06,
      "loss": 0.0373,
      "num_tokens": 86509784.0,
      "step": 1135
    },
    {
      "epoch": 1.4167186525265127,
      "grad_norm": 0.1371476505333162,
      "learning_rate": 6.131145797508414e-06,
      "loss": 0.0395,
      "num_tokens": 86586524.0,
      "step": 1136
    },
    {
      "epoch": 1.4179663131628197,
      "grad_norm": 0.11223564174314306,
      "learning_rate": 6.125145472411845e-06,
      "loss": 0.0376,
      "num_tokens": 86663862.0,
      "step": 1137
    },
    {
      "epoch": 1.4192139737991267,
      "grad_norm": 0.12820345748805365,
      "learning_rate": 6.1191440137375775e-06,
      "loss": 0.0377,
      "num_tokens": 86740239.0,
      "step": 1138
    },
    {
      "epoch": 1.4204616344354335,
      "grad_norm": 0.11565453169043415,
      "learning_rate": 6.113141432368075e-06,
      "loss": 0.0377,
      "num_tokens": 86817067.0,
      "step": 1139
    },
    {
      "epoch": 1.4217092950717405,
      "grad_norm": 0.12044599920690999,
      "learning_rate": 6.107137739187827e-06,
      "loss": 0.0357,
      "num_tokens": 86892431.0,
      "step": 1140
    },
    {
      "epoch": 1.4229569557080475,
      "grad_norm": 0.13067772165236136,
      "learning_rate": 6.101132945083347e-06,
      "loss": 0.0395,
      "num_tokens": 86968048.0,
      "step": 1141
    },
    {
      "epoch": 1.4242046163443542,
      "grad_norm": 0.11788372463688408,
      "learning_rate": 6.095127060943141e-06,
      "loss": 0.0362,
      "num_tokens": 87042935.0,
      "step": 1142
    },
    {
      "epoch": 1.4254522769806612,
      "grad_norm": 0.11002089825894382,
      "learning_rate": 6.089120097657692e-06,
      "loss": 0.0383,
      "num_tokens": 87120151.0,
      "step": 1143
    },
    {
      "epoch": 1.4266999376169682,
      "grad_norm": 0.12686988054618584,
      "learning_rate": 6.083112066119439e-06,
      "loss": 0.0385,
      "num_tokens": 87196229.0,
      "step": 1144
    },
    {
      "epoch": 1.427947598253275,
      "grad_norm": 0.12173931969583418,
      "learning_rate": 6.077102977222763e-06,
      "loss": 0.0406,
      "num_tokens": 87273162.0,
      "step": 1145
    },
    {
      "epoch": 1.429195258889582,
      "grad_norm": 0.12795374944584967,
      "learning_rate": 6.0710928418639515e-06,
      "loss": 0.0384,
      "num_tokens": 87349259.0,
      "step": 1146
    },
    {
      "epoch": 1.430442919525889,
      "grad_norm": 0.14248994453706956,
      "learning_rate": 6.065081670941204e-06,
      "loss": 0.04,
      "num_tokens": 87426901.0,
      "step": 1147
    },
    {
      "epoch": 1.4316905801621957,
      "grad_norm": 0.11818189693034521,
      "learning_rate": 6.059069475354586e-06,
      "loss": 0.0371,
      "num_tokens": 87504412.0,
      "step": 1148
    },
    {
      "epoch": 1.4329382407985027,
      "grad_norm": 0.11293724815408349,
      "learning_rate": 6.0530562660060276e-06,
      "loss": 0.0365,
      "num_tokens": 87579968.0,
      "step": 1149
    },
    {
      "epoch": 1.4341859014348097,
      "grad_norm": 0.12150371365500466,
      "learning_rate": 6.0470420537992915e-06,
      "loss": 0.0375,
      "num_tokens": 87656442.0,
      "step": 1150
    },
    {
      "epoch": 1.4354335620711167,
      "grad_norm": 0.11127014521567379,
      "learning_rate": 6.041026849639966e-06,
      "loss": 0.0361,
      "num_tokens": 87732219.0,
      "step": 1151
    },
    {
      "epoch": 1.4366812227074237,
      "grad_norm": 0.12430543357003211,
      "learning_rate": 6.035010664435434e-06,
      "loss": 0.0414,
      "num_tokens": 87807872.0,
      "step": 1152
    },
    {
      "epoch": 1.4379288833437305,
      "grad_norm": 0.1215550972541805,
      "learning_rate": 6.0289935090948536e-06,
      "loss": 0.04,
      "num_tokens": 87885065.0,
      "step": 1153
    },
    {
      "epoch": 1.4391765439800375,
      "grad_norm": 0.12018062200786944,
      "learning_rate": 6.022975394529149e-06,
      "loss": 0.0369,
      "num_tokens": 87960894.0,
      "step": 1154
    },
    {
      "epoch": 1.4404242046163445,
      "grad_norm": 0.11641064053528057,
      "learning_rate": 6.016956331650984e-06,
      "loss": 0.0378,
      "num_tokens": 88037862.0,
      "step": 1155
    },
    {
      "epoch": 1.4416718652526512,
      "grad_norm": 0.12049199468836105,
      "learning_rate": 6.010936331374735e-06,
      "loss": 0.0362,
      "num_tokens": 88113631.0,
      "step": 1156
    },
    {
      "epoch": 1.4429195258889582,
      "grad_norm": 0.1117063446883456,
      "learning_rate": 6.00491540461648e-06,
      "loss": 0.0369,
      "num_tokens": 88189354.0,
      "step": 1157
    },
    {
      "epoch": 1.4441671865252652,
      "grad_norm": 0.1213557530389275,
      "learning_rate": 5.998893562293986e-06,
      "loss": 0.0373,
      "num_tokens": 88265382.0,
      "step": 1158
    },
    {
      "epoch": 1.445414847161572,
      "grad_norm": 0.12152558671097215,
      "learning_rate": 5.992870815326667e-06,
      "loss": 0.0367,
      "num_tokens": 88342173.0,
      "step": 1159
    },
    {
      "epoch": 1.446662507797879,
      "grad_norm": 0.11057929090406309,
      "learning_rate": 5.986847174635586e-06,
      "loss": 0.0354,
      "num_tokens": 88418805.0,
      "step": 1160
    },
    {
      "epoch": 1.447910168434186,
      "grad_norm": 0.12399316687481432,
      "learning_rate": 5.980822651143426e-06,
      "loss": 0.0387,
      "num_tokens": 88494308.0,
      "step": 1161
    },
    {
      "epoch": 1.4491578290704927,
      "grad_norm": 0.11572790861021126,
      "learning_rate": 5.9747972557744675e-06,
      "loss": 0.0381,
      "num_tokens": 88570633.0,
      "step": 1162
    },
    {
      "epoch": 1.4504054897067997,
      "grad_norm": 0.11539324409857993,
      "learning_rate": 5.968770999454572e-06,
      "loss": 0.0374,
      "num_tokens": 88647435.0,
      "step": 1163
    },
    {
      "epoch": 1.4516531503431067,
      "grad_norm": 0.12038386699531019,
      "learning_rate": 5.962743893111165e-06,
      "loss": 0.0395,
      "num_tokens": 88724582.0,
      "step": 1164
    },
    {
      "epoch": 1.4529008109794135,
      "grad_norm": 0.13615446233989564,
      "learning_rate": 5.956715947673212e-06,
      "loss": 0.0377,
      "num_tokens": 88800743.0,
      "step": 1165
    },
    {
      "epoch": 1.4541484716157205,
      "grad_norm": 0.11828852500542161,
      "learning_rate": 5.950687174071201e-06,
      "loss": 0.0344,
      "num_tokens": 88875189.0,
      "step": 1166
    },
    {
      "epoch": 1.4553961322520275,
      "grad_norm": 0.11198051066883659,
      "learning_rate": 5.944657583237119e-06,
      "loss": 0.0356,
      "num_tokens": 88950046.0,
      "step": 1167
    },
    {
      "epoch": 1.4566437928883345,
      "grad_norm": 0.12304932000403278,
      "learning_rate": 5.938627186104438e-06,
      "loss": 0.04,
      "num_tokens": 89026982.0,
      "step": 1168
    },
    {
      "epoch": 1.4578914535246412,
      "grad_norm": 0.1287086555877926,
      "learning_rate": 5.932595993608092e-06,
      "loss": 0.0399,
      "num_tokens": 89102664.0,
      "step": 1169
    },
    {
      "epoch": 1.4591391141609482,
      "grad_norm": 0.13685349466568447,
      "learning_rate": 5.926564016684453e-06,
      "loss": 0.0361,
      "num_tokens": 89178795.0,
      "step": 1170
    },
    {
      "epoch": 1.4603867747972552,
      "grad_norm": 0.11703659063026274,
      "learning_rate": 5.920531266271317e-06,
      "loss": 0.0386,
      "num_tokens": 89254845.0,
      "step": 1171
    },
    {
      "epoch": 1.4616344354335622,
      "grad_norm": 0.12135909606309495,
      "learning_rate": 5.9144977533078885e-06,
      "loss": 0.0382,
      "num_tokens": 89332936.0,
      "step": 1172
    },
    {
      "epoch": 1.462882096069869,
      "grad_norm": 0.12119899763340669,
      "learning_rate": 5.90846348873475e-06,
      "loss": 0.0384,
      "num_tokens": 89409906.0,
      "step": 1173
    },
    {
      "epoch": 1.464129756706176,
      "grad_norm": 0.1257634185270069,
      "learning_rate": 5.902428483493845e-06,
      "loss": 0.0382,
      "num_tokens": 89486609.0,
      "step": 1174
    },
    {
      "epoch": 1.465377417342483,
      "grad_norm": 0.1258432085205612,
      "learning_rate": 5.89639274852846e-06,
      "loss": 0.0362,
      "num_tokens": 89562321.0,
      "step": 1175
    },
    {
      "epoch": 1.4666250779787897,
      "grad_norm": 0.12013512152698796,
      "learning_rate": 5.890356294783213e-06,
      "loss": 0.038,
      "num_tokens": 89638226.0,
      "step": 1176
    },
    {
      "epoch": 1.4678727386150967,
      "grad_norm": 0.136427771320821,
      "learning_rate": 5.8843191332040125e-06,
      "loss": 0.04,
      "num_tokens": 89714165.0,
      "step": 1177
    },
    {
      "epoch": 1.4691203992514037,
      "grad_norm": 0.12180732554236498,
      "learning_rate": 5.878281274738061e-06,
      "loss": 0.0365,
      "num_tokens": 89790575.0,
      "step": 1178
    },
    {
      "epoch": 1.4703680598877105,
      "grad_norm": 0.12196206277349746,
      "learning_rate": 5.872242730333822e-06,
      "loss": 0.0366,
      "num_tokens": 89866156.0,
      "step": 1179
    },
    {
      "epoch": 1.4716157205240175,
      "grad_norm": 0.11523527520323638,
      "learning_rate": 5.866203510940998e-06,
      "loss": 0.039,
      "num_tokens": 89941867.0,
      "step": 1180
    },
    {
      "epoch": 1.4728633811603244,
      "grad_norm": 0.1296767966657419,
      "learning_rate": 5.860163627510521e-06,
      "loss": 0.0395,
      "num_tokens": 90019668.0,
      "step": 1181
    },
    {
      "epoch": 1.4741110417966312,
      "grad_norm": 0.11213362843515545,
      "learning_rate": 5.854123090994524e-06,
      "loss": 0.0377,
      "num_tokens": 90096135.0,
      "step": 1182
    },
    {
      "epoch": 1.4753587024329382,
      "grad_norm": 0.10840648305544608,
      "learning_rate": 5.848081912346329e-06,
      "loss": 0.0362,
      "num_tokens": 90172349.0,
      "step": 1183
    },
    {
      "epoch": 1.4766063630692452,
      "grad_norm": 0.11369092296626487,
      "learning_rate": 5.842040102520416e-06,
      "loss": 0.0375,
      "num_tokens": 90249853.0,
      "step": 1184
    },
    {
      "epoch": 1.477854023705552,
      "grad_norm": 0.12170948408983744,
      "learning_rate": 5.8359976724724146e-06,
      "loss": 0.0381,
      "num_tokens": 90327724.0,
      "step": 1185
    },
    {
      "epoch": 1.479101684341859,
      "grad_norm": 0.11993171807152257,
      "learning_rate": 5.829954633159073e-06,
      "loss": 0.0364,
      "num_tokens": 90404285.0,
      "step": 1186
    },
    {
      "epoch": 1.480349344978166,
      "grad_norm": 0.11597109699676006,
      "learning_rate": 5.823910995538251e-06,
      "loss": 0.0371,
      "num_tokens": 90480540.0,
      "step": 1187
    },
    {
      "epoch": 1.481597005614473,
      "grad_norm": 0.11978029510757789,
      "learning_rate": 5.8178667705688895e-06,
      "loss": 0.0367,
      "num_tokens": 90557248.0,
      "step": 1188
    },
    {
      "epoch": 1.4828446662507797,
      "grad_norm": 0.121547607943497,
      "learning_rate": 5.811821969210995e-06,
      "loss": 0.0365,
      "num_tokens": 90631870.0,
      "step": 1189
    },
    {
      "epoch": 1.4840923268870867,
      "grad_norm": 0.11514573974937457,
      "learning_rate": 5.8057766024256205e-06,
      "loss": 0.0366,
      "num_tokens": 90707243.0,
      "step": 1190
    },
    {
      "epoch": 1.4853399875233937,
      "grad_norm": 0.1186525151878151,
      "learning_rate": 5.799730681174842e-06,
      "loss": 0.0361,
      "num_tokens": 90784104.0,
      "step": 1191
    },
    {
      "epoch": 1.4865876481597007,
      "grad_norm": 0.11477823661113125,
      "learning_rate": 5.793684216421744e-06,
      "loss": 0.0387,
      "num_tokens": 90859903.0,
      "step": 1192
    },
    {
      "epoch": 1.4878353087960074,
      "grad_norm": 0.12385914646775886,
      "learning_rate": 5.787637219130392e-06,
      "loss": 0.0384,
      "num_tokens": 90934859.0,
      "step": 1193
    },
    {
      "epoch": 1.4890829694323144,
      "grad_norm": 0.12832223803870393,
      "learning_rate": 5.781589700265823e-06,
      "loss": 0.0369,
      "num_tokens": 91011413.0,
      "step": 1194
    },
    {
      "epoch": 1.4903306300686214,
      "grad_norm": 0.11256583038803755,
      "learning_rate": 5.7755416707940135e-06,
      "loss": 0.035,
      "num_tokens": 91086803.0,
      "step": 1195
    },
    {
      "epoch": 1.4915782907049282,
      "grad_norm": 0.12243730113925984,
      "learning_rate": 5.76949314168187e-06,
      "loss": 0.0374,
      "num_tokens": 91161954.0,
      "step": 1196
    },
    {
      "epoch": 1.4928259513412352,
      "grad_norm": 0.12768023862843755,
      "learning_rate": 5.763444123897206e-06,
      "loss": 0.0377,
      "num_tokens": 91239788.0,
      "step": 1197
    },
    {
      "epoch": 1.4940736119775422,
      "grad_norm": 0.11012832453778412,
      "learning_rate": 5.757394628408716e-06,
      "loss": 0.0361,
      "num_tokens": 91314747.0,
      "step": 1198
    },
    {
      "epoch": 1.495321272613849,
      "grad_norm": 0.12227103519562003,
      "learning_rate": 5.7513446661859664e-06,
      "loss": 0.0353,
      "num_tokens": 91390165.0,
      "step": 1199
    },
    {
      "epoch": 1.496568933250156,
      "grad_norm": 0.13124194561569905,
      "learning_rate": 5.7452942481993655e-06,
      "loss": 0.0388,
      "num_tokens": 91467820.0,
      "step": 1200
    },
    {
      "epoch": 1.497816593886463,
      "grad_norm": 0.12306926635997256,
      "learning_rate": 5.739243385420151e-06,
      "loss": 0.0344,
      "num_tokens": 91542554.0,
      "step": 1201
    },
    {
      "epoch": 1.4990642545227697,
      "grad_norm": 0.10980086429236924,
      "learning_rate": 5.7331920888203655e-06,
      "loss": 0.043,
      "num_tokens": 91618265.0,
      "step": 1202
    },
    {
      "epoch": 1.5003119151590767,
      "grad_norm": 0.12750053322537247,
      "learning_rate": 5.727140369372838e-06,
      "loss": 0.0374,
      "num_tokens": 91693957.0,
      "step": 1203
    },
    {
      "epoch": 1.5015595757953837,
      "grad_norm": 0.1350691229685045,
      "learning_rate": 5.721088238051168e-06,
      "loss": 0.0423,
      "num_tokens": 91770974.0,
      "step": 1204
    },
    {
      "epoch": 1.5028072364316905,
      "grad_norm": 0.12111928534382646,
      "learning_rate": 5.715035705829696e-06,
      "loss": 0.0378,
      "num_tokens": 91848134.0,
      "step": 1205
    },
    {
      "epoch": 1.5040548970679977,
      "grad_norm": 0.12794780455854038,
      "learning_rate": 5.708982783683492e-06,
      "loss": 0.0357,
      "num_tokens": 91924118.0,
      "step": 1206
    },
    {
      "epoch": 1.5053025577043044,
      "grad_norm": 0.1084695976853833,
      "learning_rate": 5.7029294825883365e-06,
      "loss": 0.0346,
      "num_tokens": 91999247.0,
      "step": 1207
    },
    {
      "epoch": 1.5065502183406112,
      "grad_norm": 0.11562042009012471,
      "learning_rate": 5.696875813520691e-06,
      "loss": 0.0428,
      "num_tokens": 92074627.0,
      "step": 1208
    },
    {
      "epoch": 1.5077978789769184,
      "grad_norm": 0.12053188830493768,
      "learning_rate": 5.69082178745769e-06,
      "loss": 0.0364,
      "num_tokens": 92151320.0,
      "step": 1209
    },
    {
      "epoch": 1.5090455396132252,
      "grad_norm": 0.11805305700246772,
      "learning_rate": 5.68476741537711e-06,
      "loss": 0.0372,
      "num_tokens": 92226893.0,
      "step": 1210
    },
    {
      "epoch": 1.5102932002495322,
      "grad_norm": 0.11833239189442073,
      "learning_rate": 5.678712708257358e-06,
      "loss": 0.0353,
      "num_tokens": 92305191.0,
      "step": 1211
    },
    {
      "epoch": 1.5115408608858392,
      "grad_norm": 0.11372795928925503,
      "learning_rate": 5.672657677077449e-06,
      "loss": 0.0366,
      "num_tokens": 92380981.0,
      "step": 1212
    },
    {
      "epoch": 1.512788521522146,
      "grad_norm": 0.12034696590824727,
      "learning_rate": 5.666602332816985e-06,
      "loss": 0.0355,
      "num_tokens": 92455797.0,
      "step": 1213
    },
    {
      "epoch": 1.514036182158453,
      "grad_norm": 0.12788533264540333,
      "learning_rate": 5.6605466864561344e-06,
      "loss": 0.0365,
      "num_tokens": 92532365.0,
      "step": 1214
    },
    {
      "epoch": 1.51528384279476,
      "grad_norm": 0.11721602859972623,
      "learning_rate": 5.654490748975615e-06,
      "loss": 0.0382,
      "num_tokens": 92608889.0,
      "step": 1215
    },
    {
      "epoch": 1.5165315034310667,
      "grad_norm": 0.11978374195139166,
      "learning_rate": 5.648434531356671e-06,
      "loss": 0.0374,
      "num_tokens": 92685321.0,
      "step": 1216
    },
    {
      "epoch": 1.5177791640673737,
      "grad_norm": 0.11893827548661735,
      "learning_rate": 5.642378044581057e-06,
      "loss": 0.0365,
      "num_tokens": 92762518.0,
      "step": 1217
    },
    {
      "epoch": 1.5190268247036807,
      "grad_norm": 0.1160725816936642,
      "learning_rate": 5.636321299631015e-06,
      "loss": 0.037,
      "num_tokens": 92837516.0,
      "step": 1218
    },
    {
      "epoch": 1.5202744853399874,
      "grad_norm": 0.12435223239608878,
      "learning_rate": 5.630264307489251e-06,
      "loss": 0.0368,
      "num_tokens": 92913538.0,
      "step": 1219
    },
    {
      "epoch": 1.5215221459762944,
      "grad_norm": 0.1152960182186084,
      "learning_rate": 5.624207079138922e-06,
      "loss": 0.0352,
      "num_tokens": 92989240.0,
      "step": 1220
    },
    {
      "epoch": 1.5227698066126014,
      "grad_norm": 0.12002871315414164,
      "learning_rate": 5.6181496255636195e-06,
      "loss": 0.0369,
      "num_tokens": 93065615.0,
      "step": 1221
    },
    {
      "epoch": 1.5240174672489082,
      "grad_norm": 0.11869764826176993,
      "learning_rate": 5.612091957747333e-06,
      "loss": 0.0359,
      "num_tokens": 93141553.0,
      "step": 1222
    },
    {
      "epoch": 1.5252651278852152,
      "grad_norm": 0.11533368448603282,
      "learning_rate": 5.606034086674447e-06,
      "loss": 0.0377,
      "num_tokens": 93217211.0,
      "step": 1223
    },
    {
      "epoch": 1.5265127885215222,
      "grad_norm": 0.12493028735607448,
      "learning_rate": 5.5999760233297115e-06,
      "loss": 0.0371,
      "num_tokens": 93292801.0,
      "step": 1224
    },
    {
      "epoch": 1.527760449157829,
      "grad_norm": 0.12755055829863948,
      "learning_rate": 5.593917778698227e-06,
      "loss": 0.0397,
      "num_tokens": 93369348.0,
      "step": 1225
    },
    {
      "epoch": 1.5290081097941361,
      "grad_norm": 0.1197732646050575,
      "learning_rate": 5.5878593637654226e-06,
      "loss": 0.0381,
      "num_tokens": 93445929.0,
      "step": 1226
    },
    {
      "epoch": 1.530255770430443,
      "grad_norm": 0.11713177274161533,
      "learning_rate": 5.581800789517036e-06,
      "loss": 0.0368,
      "num_tokens": 93523595.0,
      "step": 1227
    },
    {
      "epoch": 1.5315034310667497,
      "grad_norm": 0.11452856294775239,
      "learning_rate": 5.5757420669390925e-06,
      "loss": 0.0367,
      "num_tokens": 93599007.0,
      "step": 1228
    },
    {
      "epoch": 1.532751091703057,
      "grad_norm": 0.11955913139155055,
      "learning_rate": 5.5696832070178885e-06,
      "loss": 0.0384,
      "num_tokens": 93674617.0,
      "step": 1229
    },
    {
      "epoch": 1.5339987523393637,
      "grad_norm": 0.11445496167989216,
      "learning_rate": 5.563624220739969e-06,
      "loss": 0.0375,
      "num_tokens": 93750264.0,
      "step": 1230
    },
    {
      "epoch": 1.5352464129756707,
      "grad_norm": 0.1134416374479405,
      "learning_rate": 5.557565119092106e-06,
      "loss": 0.0366,
      "num_tokens": 93825604.0,
      "step": 1231
    },
    {
      "epoch": 1.5364940736119777,
      "grad_norm": 0.11119872151750049,
      "learning_rate": 5.551505913061281e-06,
      "loss": 0.0351,
      "num_tokens": 93900858.0,
      "step": 1232
    },
    {
      "epoch": 1.5377417342482844,
      "grad_norm": 0.12334463348947557,
      "learning_rate": 5.54544661363467e-06,
      "loss": 0.0389,
      "num_tokens": 93978548.0,
      "step": 1233
    },
    {
      "epoch": 1.5389893948845914,
      "grad_norm": 0.10909156980739217,
      "learning_rate": 5.53938723179961e-06,
      "loss": 0.0368,
      "num_tokens": 94054145.0,
      "step": 1234
    },
    {
      "epoch": 1.5402370555208984,
      "grad_norm": 0.12443155582203641,
      "learning_rate": 5.533327778543593e-06,
      "loss": 0.0388,
      "num_tokens": 94129620.0,
      "step": 1235
    },
    {
      "epoch": 1.5414847161572052,
      "grad_norm": 0.11000189874354667,
      "learning_rate": 5.527268264854241e-06,
      "loss": 0.0367,
      "num_tokens": 94207087.0,
      "step": 1236
    },
    {
      "epoch": 1.5427323767935122,
      "grad_norm": 0.13284023214221294,
      "learning_rate": 5.521208701719284e-06,
      "loss": 0.0394,
      "num_tokens": 94283829.0,
      "step": 1237
    },
    {
      "epoch": 1.5439800374298192,
      "grad_norm": 0.11877957335168572,
      "learning_rate": 5.515149100126539e-06,
      "loss": 0.0384,
      "num_tokens": 94360479.0,
      "step": 1238
    },
    {
      "epoch": 1.545227698066126,
      "grad_norm": 0.11498644957104806,
      "learning_rate": 5.509089471063897e-06,
      "loss": 0.039,
      "num_tokens": 94439077.0,
      "step": 1239
    },
    {
      "epoch": 1.546475358702433,
      "grad_norm": 0.1257089848903344,
      "learning_rate": 5.503029825519296e-06,
      "loss": 0.0374,
      "num_tokens": 94515106.0,
      "step": 1240
    },
    {
      "epoch": 1.54772301933874,
      "grad_norm": 0.11451851398012343,
      "learning_rate": 5.496970174480706e-06,
      "loss": 0.0384,
      "num_tokens": 94591702.0,
      "step": 1241
    },
    {
      "epoch": 1.5489706799750467,
      "grad_norm": 0.14041696423500313,
      "learning_rate": 5.4909105289361055e-06,
      "loss": 0.056,
      "num_tokens": 94668695.0,
      "step": 1242
    },
    {
      "epoch": 1.5502183406113537,
      "grad_norm": 0.12498370183180903,
      "learning_rate": 5.4848508998734626e-06,
      "loss": 0.039,
      "num_tokens": 94744775.0,
      "step": 1243
    },
    {
      "epoch": 1.5514660012476607,
      "grad_norm": 0.1315344410628175,
      "learning_rate": 5.478791298280719e-06,
      "loss": 0.0362,
      "num_tokens": 94819820.0,
      "step": 1244
    },
    {
      "epoch": 1.5527136618839674,
      "grad_norm": 0.11033302554173556,
      "learning_rate": 5.47273173514576e-06,
      "loss": 0.0354,
      "num_tokens": 94895512.0,
      "step": 1245
    },
    {
      "epoch": 1.5539613225202746,
      "grad_norm": 0.1110224016665635,
      "learning_rate": 5.466672221456408e-06,
      "loss": 0.0376,
      "num_tokens": 94971206.0,
      "step": 1246
    },
    {
      "epoch": 1.5552089831565814,
      "grad_norm": 0.11225773695191615,
      "learning_rate": 5.4606127682003915e-06,
      "loss": 0.0422,
      "num_tokens": 95047311.0,
      "step": 1247
    },
    {
      "epoch": 1.5564566437928882,
      "grad_norm": 0.11231998616938674,
      "learning_rate": 5.454553386365333e-06,
      "loss": 0.0352,
      "num_tokens": 95123071.0,
      "step": 1248
    },
    {
      "epoch": 1.5577043044291954,
      "grad_norm": 0.11468464677858389,
      "learning_rate": 5.44849408693872e-06,
      "loss": 0.0381,
      "num_tokens": 95200046.0,
      "step": 1249
    },
    {
      "epoch": 1.5589519650655022,
      "grad_norm": 0.11582658442663075,
      "learning_rate": 5.4424348809078974e-06,
      "loss": 0.0365,
      "num_tokens": 95275311.0,
      "step": 1250
    },
    {
      "epoch": 1.5601996257018091,
      "grad_norm": 0.11669328477123285,
      "learning_rate": 5.436375779260034e-06,
      "loss": 0.0348,
      "num_tokens": 95351246.0,
      "step": 1251
    },
    {
      "epoch": 1.5614472863381161,
      "grad_norm": 0.11537314709151678,
      "learning_rate": 5.430316792982112e-06,
      "loss": 0.0373,
      "num_tokens": 95426887.0,
      "step": 1252
    },
    {
      "epoch": 1.562694946974423,
      "grad_norm": 0.11989449996947185,
      "learning_rate": 5.424257933060908e-06,
      "loss": 0.0346,
      "num_tokens": 95502541.0,
      "step": 1253
    },
    {
      "epoch": 1.56394260761073,
      "grad_norm": 0.1142438417076298,
      "learning_rate": 5.418199210482965e-06,
      "loss": 0.0397,
      "num_tokens": 95580052.0,
      "step": 1254
    },
    {
      "epoch": 1.5651902682470369,
      "grad_norm": 0.11351778634936657,
      "learning_rate": 5.412140636234579e-06,
      "loss": 0.037,
      "num_tokens": 95655811.0,
      "step": 1255
    },
    {
      "epoch": 1.5664379288833437,
      "grad_norm": 0.11591053947282466,
      "learning_rate": 5.4060822213017745e-06,
      "loss": 0.0374,
      "num_tokens": 95732008.0,
      "step": 1256
    },
    {
      "epoch": 1.5676855895196506,
      "grad_norm": 0.11163580600728568,
      "learning_rate": 5.400023976670291e-06,
      "loss": 0.0388,
      "num_tokens": 95808735.0,
      "step": 1257
    },
    {
      "epoch": 1.5689332501559576,
      "grad_norm": 0.11912649417015345,
      "learning_rate": 5.393965913325555e-06,
      "loss": 0.0376,
      "num_tokens": 95884978.0,
      "step": 1258
    },
    {
      "epoch": 1.5701809107922644,
      "grad_norm": 0.12010176790938462,
      "learning_rate": 5.387908042252667e-06,
      "loss": 0.0364,
      "num_tokens": 95962331.0,
      "step": 1259
    },
    {
      "epoch": 1.5714285714285714,
      "grad_norm": 0.11404213718160137,
      "learning_rate": 5.381850374436383e-06,
      "loss": 0.0375,
      "num_tokens": 96039168.0,
      "step": 1260
    },
    {
      "epoch": 1.5726762320648784,
      "grad_norm": 0.11763900222873848,
      "learning_rate": 5.3757929208610784e-06,
      "loss": 0.0382,
      "num_tokens": 96115308.0,
      "step": 1261
    },
    {
      "epoch": 1.5739238927011852,
      "grad_norm": 0.1099813965721135,
      "learning_rate": 5.3697356925107514e-06,
      "loss": 0.0359,
      "num_tokens": 96191451.0,
      "step": 1262
    },
    {
      "epoch": 1.5751715533374921,
      "grad_norm": 0.1092899473174701,
      "learning_rate": 5.363678700368987e-06,
      "loss": 0.0351,
      "num_tokens": 96267849.0,
      "step": 1263
    },
    {
      "epoch": 1.5764192139737991,
      "grad_norm": 0.11219756592113182,
      "learning_rate": 5.3576219554189445e-06,
      "loss": 0.0392,
      "num_tokens": 96343518.0,
      "step": 1264
    },
    {
      "epoch": 1.577666874610106,
      "grad_norm": 0.11513865500406484,
      "learning_rate": 5.35156546864333e-06,
      "loss": 0.0359,
      "num_tokens": 96419685.0,
      "step": 1265
    },
    {
      "epoch": 1.5789145352464131,
      "grad_norm": 0.11755138786676107,
      "learning_rate": 5.345509251024387e-06,
      "loss": 0.0365,
      "num_tokens": 96494987.0,
      "step": 1266
    },
    {
      "epoch": 1.5801621958827199,
      "grad_norm": 0.10723917441034322,
      "learning_rate": 5.339453313543868e-06,
      "loss": 0.037,
      "num_tokens": 96571590.0,
      "step": 1267
    },
    {
      "epoch": 1.5814098565190269,
      "grad_norm": 0.11831131657005481,
      "learning_rate": 5.3333976671830165e-06,
      "loss": 0.0376,
      "num_tokens": 96648687.0,
      "step": 1268
    },
    {
      "epoch": 1.5826575171553339,
      "grad_norm": 0.12060744557345109,
      "learning_rate": 5.327342322922553e-06,
      "loss": 0.0377,
      "num_tokens": 96726743.0,
      "step": 1269
    },
    {
      "epoch": 1.5839051777916406,
      "grad_norm": 0.10568069351830238,
      "learning_rate": 5.321287291742645e-06,
      "loss": 0.034,
      "num_tokens": 96801933.0,
      "step": 1270
    },
    {
      "epoch": 1.5851528384279476,
      "grad_norm": 0.1138618386162909,
      "learning_rate": 5.315232584622893e-06,
      "loss": 0.0354,
      "num_tokens": 96876827.0,
      "step": 1271
    },
    {
      "epoch": 1.5864004990642546,
      "grad_norm": 0.11948278608459084,
      "learning_rate": 5.309178212542313e-06,
      "loss": 0.0349,
      "num_tokens": 96951983.0,
      "step": 1272
    },
    {
      "epoch": 1.5876481597005614,
      "grad_norm": 0.11306709736045806,
      "learning_rate": 5.303124186479309e-06,
      "loss": 0.0355,
      "num_tokens": 97027382.0,
      "step": 1273
    },
    {
      "epoch": 1.5888958203368684,
      "grad_norm": 0.11536595148537067,
      "learning_rate": 5.297070517411664e-06,
      "loss": 0.0354,
      "num_tokens": 97102461.0,
      "step": 1274
    },
    {
      "epoch": 1.5901434809731754,
      "grad_norm": 0.11243412386645973,
      "learning_rate": 5.2910172163165096e-06,
      "loss": 0.0375,
      "num_tokens": 97179120.0,
      "step": 1275
    },
    {
      "epoch": 1.5913911416094821,
      "grad_norm": 0.12089055030621573,
      "learning_rate": 5.284964294170306e-06,
      "loss": 0.0367,
      "num_tokens": 97255298.0,
      "step": 1276
    },
    {
      "epoch": 1.5926388022457891,
      "grad_norm": 0.12214280046521081,
      "learning_rate": 5.278911761948834e-06,
      "loss": 0.037,
      "num_tokens": 97331060.0,
      "step": 1277
    },
    {
      "epoch": 1.5938864628820961,
      "grad_norm": 0.1062549969079459,
      "learning_rate": 5.272859630627164e-06,
      "loss": 0.0352,
      "num_tokens": 97406110.0,
      "step": 1278
    },
    {
      "epoch": 1.595134123518403,
      "grad_norm": 0.12246112884589401,
      "learning_rate": 5.266807911179638e-06,
      "loss": 0.039,
      "num_tokens": 97483449.0,
      "step": 1279
    },
    {
      "epoch": 1.5963817841547099,
      "grad_norm": 0.11103361608657275,
      "learning_rate": 5.260756614579851e-06,
      "loss": 0.0372,
      "num_tokens": 97560311.0,
      "step": 1280
    },
    {
      "epoch": 1.5976294447910169,
      "grad_norm": 0.135106005974476,
      "learning_rate": 5.254705751800636e-06,
      "loss": 0.0391,
      "num_tokens": 97638070.0,
      "step": 1281
    },
    {
      "epoch": 1.5988771054273236,
      "grad_norm": 0.11795350414372122,
      "learning_rate": 5.248655333814036e-06,
      "loss": 0.0367,
      "num_tokens": 97714118.0,
      "step": 1282
    },
    {
      "epoch": 1.6001247660636309,
      "grad_norm": 0.1277071179872957,
      "learning_rate": 5.242605371591286e-06,
      "loss": 0.0378,
      "num_tokens": 97790110.0,
      "step": 1283
    },
    {
      "epoch": 1.6013724266999376,
      "grad_norm": 0.10832036457269213,
      "learning_rate": 5.236555876102797e-06,
      "loss": 0.0368,
      "num_tokens": 97865186.0,
      "step": 1284
    },
    {
      "epoch": 1.6026200873362444,
      "grad_norm": 0.12298238402683899,
      "learning_rate": 5.2305068583181314e-06,
      "loss": 0.0359,
      "num_tokens": 97941587.0,
      "step": 1285
    },
    {
      "epoch": 1.6038677479725516,
      "grad_norm": 0.11205375865529897,
      "learning_rate": 5.2244583292059896e-06,
      "loss": 0.0351,
      "num_tokens": 98017073.0,
      "step": 1286
    },
    {
      "epoch": 1.6051154086088584,
      "grad_norm": 0.11398489887533124,
      "learning_rate": 5.218410299734181e-06,
      "loss": 0.0361,
      "num_tokens": 98093084.0,
      "step": 1287
    },
    {
      "epoch": 1.6063630692451654,
      "grad_norm": 0.11331925773487742,
      "learning_rate": 5.2123627808696084e-06,
      "loss": 0.0378,
      "num_tokens": 98169702.0,
      "step": 1288
    },
    {
      "epoch": 1.6076107298814724,
      "grad_norm": 0.12754453783748512,
      "learning_rate": 5.206315783578258e-06,
      "loss": 0.0379,
      "num_tokens": 98246352.0,
      "step": 1289
    },
    {
      "epoch": 1.6088583905177791,
      "grad_norm": 0.11695300113526581,
      "learning_rate": 5.20026931882516e-06,
      "loss": 0.0351,
      "num_tokens": 98321908.0,
      "step": 1290
    },
    {
      "epoch": 1.6101060511540861,
      "grad_norm": 0.10619866825878661,
      "learning_rate": 5.194223397574381e-06,
      "loss": 0.034,
      "num_tokens": 98398421.0,
      "step": 1291
    },
    {
      "epoch": 1.611353711790393,
      "grad_norm": 0.11601433765217581,
      "learning_rate": 5.188178030789008e-06,
      "loss": 0.0375,
      "num_tokens": 98475418.0,
      "step": 1292
    },
    {
      "epoch": 1.6126013724266999,
      "grad_norm": 0.11176870281043169,
      "learning_rate": 5.1821332294311136e-06,
      "loss": 0.035,
      "num_tokens": 98551489.0,
      "step": 1293
    },
    {
      "epoch": 1.6138490330630069,
      "grad_norm": 0.12198142407040487,
      "learning_rate": 5.176089004461752e-06,
      "loss": 0.0376,
      "num_tokens": 98628378.0,
      "step": 1294
    },
    {
      "epoch": 1.6150966936993139,
      "grad_norm": 0.12752726575125428,
      "learning_rate": 5.170045366840929e-06,
      "loss": 0.0384,
      "num_tokens": 98704929.0,
      "step": 1295
    },
    {
      "epoch": 1.6163443543356206,
      "grad_norm": 0.12085444832378747,
      "learning_rate": 5.164002327527588e-06,
      "loss": 0.04,
      "num_tokens": 98782920.0,
      "step": 1296
    },
    {
      "epoch": 1.6175920149719276,
      "grad_norm": 0.13632096954307515,
      "learning_rate": 5.157959897479587e-06,
      "loss": 0.0373,
      "num_tokens": 98858537.0,
      "step": 1297
    },
    {
      "epoch": 1.6188396756082346,
      "grad_norm": 0.12306314338751427,
      "learning_rate": 5.151918087653672e-06,
      "loss": 0.0361,
      "num_tokens": 98934085.0,
      "step": 1298
    },
    {
      "epoch": 1.6200873362445414,
      "grad_norm": 0.1403645125183729,
      "learning_rate": 5.145876909005477e-06,
      "loss": 0.034,
      "num_tokens": 99009665.0,
      "step": 1299
    },
    {
      "epoch": 1.6213349968808484,
      "grad_norm": 0.1144609737580006,
      "learning_rate": 5.139836372489481e-06,
      "loss": 0.0341,
      "num_tokens": 99084393.0,
      "step": 1300
    },
    {
      "epoch": 1.6225826575171554,
      "grad_norm": 0.1073101633798548,
      "learning_rate": 5.133796489059005e-06,
      "loss": 0.0414,
      "num_tokens": 99163283.0,
      "step": 1301
    },
    {
      "epoch": 1.6238303181534621,
      "grad_norm": 0.12155196636230708,
      "learning_rate": 5.1277572696661806e-06,
      "loss": 0.0359,
      "num_tokens": 99240403.0,
      "step": 1302
    },
    {
      "epoch": 1.6250779787897693,
      "grad_norm": 0.11193137861888704,
      "learning_rate": 5.12171872526194e-06,
      "loss": 0.0391,
      "num_tokens": 99316331.0,
      "step": 1303
    },
    {
      "epoch": 1.626325639426076,
      "grad_norm": 0.1297846009862795,
      "learning_rate": 5.115680866795989e-06,
      "loss": 0.0386,
      "num_tokens": 99392413.0,
      "step": 1304
    },
    {
      "epoch": 1.6275733000623829,
      "grad_norm": 0.12104786113120562,
      "learning_rate": 5.109643705216789e-06,
      "loss": 0.0387,
      "num_tokens": 99468998.0,
      "step": 1305
    },
    {
      "epoch": 1.62882096069869,
      "grad_norm": 0.1259339680092441,
      "learning_rate": 5.103607251471541e-06,
      "loss": 0.0361,
      "num_tokens": 99544604.0,
      "step": 1306
    },
    {
      "epoch": 1.6300686213349969,
      "grad_norm": 0.11907192744895645,
      "learning_rate": 5.097571516506158e-06,
      "loss": 0.038,
      "num_tokens": 99621509.0,
      "step": 1307
    },
    {
      "epoch": 1.6313162819713038,
      "grad_norm": 0.12076333993471534,
      "learning_rate": 5.091536511265253e-06,
      "loss": 0.0382,
      "num_tokens": 99698692.0,
      "step": 1308
    },
    {
      "epoch": 1.6325639426076108,
      "grad_norm": 0.11339455289910984,
      "learning_rate": 5.085502246692111e-06,
      "loss": 0.0376,
      "num_tokens": 99775293.0,
      "step": 1309
    },
    {
      "epoch": 1.6338116032439176,
      "grad_norm": 0.1264608232290441,
      "learning_rate": 5.079468733728684e-06,
      "loss": 0.0364,
      "num_tokens": 99852057.0,
      "step": 1310
    },
    {
      "epoch": 1.6350592638802246,
      "grad_norm": 0.1091582073216211,
      "learning_rate": 5.07343598331555e-06,
      "loss": 0.0362,
      "num_tokens": 99928654.0,
      "step": 1311
    },
    {
      "epoch": 1.6363069245165316,
      "grad_norm": 0.11123927985290005,
      "learning_rate": 5.0674040063919114e-06,
      "loss": 0.0349,
      "num_tokens": 100003903.0,
      "step": 1312
    },
    {
      "epoch": 1.6375545851528384,
      "grad_norm": 0.11986584833501389,
      "learning_rate": 5.0613728138955644e-06,
      "loss": 0.0355,
      "num_tokens": 100079727.0,
      "step": 1313
    },
    {
      "epoch": 1.6388022457891454,
      "grad_norm": 0.11450610998058318,
      "learning_rate": 5.055342416762883e-06,
      "loss": 0.0378,
      "num_tokens": 100155922.0,
      "step": 1314
    },
    {
      "epoch": 1.6400499064254523,
      "grad_norm": 0.11831806865766152,
      "learning_rate": 5.0493128259288025e-06,
      "loss": 0.0371,
      "num_tokens": 100231834.0,
      "step": 1315
    },
    {
      "epoch": 1.641297567061759,
      "grad_norm": 0.12149075339785302,
      "learning_rate": 5.043284052326789e-06,
      "loss": 0.0368,
      "num_tokens": 100307612.0,
      "step": 1316
    },
    {
      "epoch": 1.642545227698066,
      "grad_norm": 0.12082559817872005,
      "learning_rate": 5.037256106888837e-06,
      "loss": 0.0365,
      "num_tokens": 100386062.0,
      "step": 1317
    },
    {
      "epoch": 1.643792888334373,
      "grad_norm": 0.12561599057345044,
      "learning_rate": 5.03122900054543e-06,
      "loss": 0.0355,
      "num_tokens": 100462184.0,
      "step": 1318
    },
    {
      "epoch": 1.6450405489706799,
      "grad_norm": 0.11883116105714127,
      "learning_rate": 5.025202744225535e-06,
      "loss": 0.0346,
      "num_tokens": 100538383.0,
      "step": 1319
    },
    {
      "epoch": 1.6462882096069869,
      "grad_norm": 0.12249245203298004,
      "learning_rate": 5.019177348856576e-06,
      "loss": 0.0371,
      "num_tokens": 100616079.0,
      "step": 1320
    },
    {
      "epoch": 1.6475358702432938,
      "grad_norm": 0.11058943111150786,
      "learning_rate": 5.013152825364416e-06,
      "loss": 0.0359,
      "num_tokens": 100692007.0,
      "step": 1321
    },
    {
      "epoch": 1.6487835308796006,
      "grad_norm": 0.12415625273211772,
      "learning_rate": 5.007129184673335e-06,
      "loss": 0.0371,
      "num_tokens": 100768530.0,
      "step": 1322
    },
    {
      "epoch": 1.6500311915159078,
      "grad_norm": 0.1105120072809638,
      "learning_rate": 5.001106437706016e-06,
      "loss": 0.035,
      "num_tokens": 100843567.0,
      "step": 1323
    },
    {
      "epoch": 1.6512788521522146,
      "grad_norm": 0.11777793444144743,
      "learning_rate": 4.99508459538352e-06,
      "loss": 0.038,
      "num_tokens": 100919975.0,
      "step": 1324
    },
    {
      "epoch": 1.6525265127885214,
      "grad_norm": 0.11792502762787642,
      "learning_rate": 4.989063668625267e-06,
      "loss": 0.0387,
      "num_tokens": 100996420.0,
      "step": 1325
    },
    {
      "epoch": 1.6537741734248286,
      "grad_norm": 0.12674786529957854,
      "learning_rate": 4.983043668349018e-06,
      "loss": 0.0366,
      "num_tokens": 101072591.0,
      "step": 1326
    },
    {
      "epoch": 1.6550218340611353,
      "grad_norm": 0.11470804032572889,
      "learning_rate": 4.977024605470851e-06,
      "loss": 0.0399,
      "num_tokens": 101149731.0,
      "step": 1327
    },
    {
      "epoch": 1.6562694946974423,
      "grad_norm": 0.1231867095999115,
      "learning_rate": 4.971006490905148e-06,
      "loss": 0.0391,
      "num_tokens": 101226108.0,
      "step": 1328
    },
    {
      "epoch": 1.6575171553337493,
      "grad_norm": 0.11568699365185363,
      "learning_rate": 4.964989335564571e-06,
      "loss": 0.0364,
      "num_tokens": 101301980.0,
      "step": 1329
    },
    {
      "epoch": 1.658764815970056,
      "grad_norm": 0.11488559020695345,
      "learning_rate": 4.958973150360034e-06,
      "loss": 0.038,
      "num_tokens": 101377961.0,
      "step": 1330
    },
    {
      "epoch": 1.660012476606363,
      "grad_norm": 0.10528800902274923,
      "learning_rate": 4.952957946200709e-06,
      "loss": 0.0334,
      "num_tokens": 101452942.0,
      "step": 1331
    },
    {
      "epoch": 1.66126013724267,
      "grad_norm": 0.11814463514870395,
      "learning_rate": 4.946943733993974e-06,
      "loss": 0.0369,
      "num_tokens": 101528775.0,
      "step": 1332
    },
    {
      "epoch": 1.6625077978789768,
      "grad_norm": 0.11188542272715878,
      "learning_rate": 4.940930524645414e-06,
      "loss": 0.0398,
      "num_tokens": 101605154.0,
      "step": 1333
    },
    {
      "epoch": 1.6637554585152838,
      "grad_norm": 0.13041645138750732,
      "learning_rate": 4.934918329058798e-06,
      "loss": 0.0383,
      "num_tokens": 101682190.0,
      "step": 1334
    },
    {
      "epoch": 1.6650031191515908,
      "grad_norm": 0.11426497483003493,
      "learning_rate": 4.928907158136049e-06,
      "loss": 0.0365,
      "num_tokens": 101757739.0,
      "step": 1335
    },
    {
      "epoch": 1.6662507797878976,
      "grad_norm": 0.11568107261393834,
      "learning_rate": 4.922897022777241e-06,
      "loss": 0.0363,
      "num_tokens": 101834357.0,
      "step": 1336
    },
    {
      "epoch": 1.6674984404242046,
      "grad_norm": 0.11574082195970792,
      "learning_rate": 4.916887933880562e-06,
      "loss": 0.0355,
      "num_tokens": 101909908.0,
      "step": 1337
    },
    {
      "epoch": 1.6687461010605116,
      "grad_norm": 0.1172235586101275,
      "learning_rate": 4.910879902342309e-06,
      "loss": 0.0364,
      "num_tokens": 101987781.0,
      "step": 1338
    },
    {
      "epoch": 1.6699937616968183,
      "grad_norm": 0.13798170432817686,
      "learning_rate": 4.904872939056859e-06,
      "loss": 0.0378,
      "num_tokens": 102063911.0,
      "step": 1339
    },
    {
      "epoch": 1.6712414223331253,
      "grad_norm": 0.1212513945751985,
      "learning_rate": 4.898867054916655e-06,
      "loss": 0.0384,
      "num_tokens": 102141336.0,
      "step": 1340
    },
    {
      "epoch": 1.6724890829694323,
      "grad_norm": 0.12928789262202314,
      "learning_rate": 4.892862260812174e-06,
      "loss": 0.0345,
      "num_tokens": 102218148.0,
      "step": 1341
    },
    {
      "epoch": 1.673736743605739,
      "grad_norm": 0.11821647712469623,
      "learning_rate": 4.886858567631927e-06,
      "loss": 0.039,
      "num_tokens": 102294670.0,
      "step": 1342
    },
    {
      "epoch": 1.6749844042420463,
      "grad_norm": 0.1192083959471349,
      "learning_rate": 4.880855986262424e-06,
      "loss": 0.0353,
      "num_tokens": 102371793.0,
      "step": 1343
    },
    {
      "epoch": 1.676232064878353,
      "grad_norm": 0.11043237689952129,
      "learning_rate": 4.874854527588159e-06,
      "loss": 0.0357,
      "num_tokens": 102446940.0,
      "step": 1344
    },
    {
      "epoch": 1.6774797255146598,
      "grad_norm": 0.12091130350927373,
      "learning_rate": 4.868854202491587e-06,
      "loss": 0.0365,
      "num_tokens": 102522517.0,
      "step": 1345
    },
    {
      "epoch": 1.678727386150967,
      "grad_norm": 0.11717551247947147,
      "learning_rate": 4.862855021853117e-06,
      "loss": 0.0364,
      "num_tokens": 102599001.0,
      "step": 1346
    },
    {
      "epoch": 1.6799750467872738,
      "grad_norm": 0.11055524190241156,
      "learning_rate": 4.856856996551074e-06,
      "loss": 0.0373,
      "num_tokens": 102675717.0,
      "step": 1347
    },
    {
      "epoch": 1.6812227074235808,
      "grad_norm": 0.12424488352828751,
      "learning_rate": 4.850860137461691e-06,
      "loss": 0.0373,
      "num_tokens": 102751593.0,
      "step": 1348
    },
    {
      "epoch": 1.6824703680598878,
      "grad_norm": 0.11204955443556182,
      "learning_rate": 4.844864455459085e-06,
      "loss": 0.0362,
      "num_tokens": 102827243.0,
      "step": 1349
    },
    {
      "epoch": 1.6837180286961946,
      "grad_norm": 0.11865902950722064,
      "learning_rate": 4.83886996141524e-06,
      "loss": 0.0359,
      "num_tokens": 102902365.0,
      "step": 1350
    },
    {
      "epoch": 1.6849656893325016,
      "grad_norm": 0.12365165052998159,
      "learning_rate": 4.8328766661999885e-06,
      "loss": 0.0367,
      "num_tokens": 102977634.0,
      "step": 1351
    },
    {
      "epoch": 1.6862133499688086,
      "grad_norm": 0.11769618162902643,
      "learning_rate": 4.826884580680981e-06,
      "loss": 0.036,
      "num_tokens": 103053963.0,
      "step": 1352
    },
    {
      "epoch": 1.6874610106051153,
      "grad_norm": 0.11005279715923194,
      "learning_rate": 4.8208937157236855e-06,
      "loss": 0.0363,
      "num_tokens": 103129617.0,
      "step": 1353
    },
    {
      "epoch": 1.6887086712414223,
      "grad_norm": 0.12190292059359654,
      "learning_rate": 4.814904082191349e-06,
      "loss": 0.0376,
      "num_tokens": 103205831.0,
      "step": 1354
    },
    {
      "epoch": 1.6899563318777293,
      "grad_norm": 0.12379736841695456,
      "learning_rate": 4.8089156909449845e-06,
      "loss": 0.0364,
      "num_tokens": 103281361.0,
      "step": 1355
    },
    {
      "epoch": 1.691203992514036,
      "grad_norm": 0.12172911071756917,
      "learning_rate": 4.802928552843358e-06,
      "loss": 0.0479,
      "num_tokens": 103359453.0,
      "step": 1356
    },
    {
      "epoch": 1.692451653150343,
      "grad_norm": 0.11458592687454641,
      "learning_rate": 4.79694267874296e-06,
      "loss": 0.0404,
      "num_tokens": 103436739.0,
      "step": 1357
    },
    {
      "epoch": 1.69369931378665,
      "grad_norm": 0.12664265435736952,
      "learning_rate": 4.790958079497991e-06,
      "loss": 0.0399,
      "num_tokens": 103513765.0,
      "step": 1358
    },
    {
      "epoch": 1.6949469744229568,
      "grad_norm": 0.1231061961741426,
      "learning_rate": 4.784974765960335e-06,
      "loss": 0.0378,
      "num_tokens": 103589063.0,
      "step": 1359
    },
    {
      "epoch": 1.696194635059264,
      "grad_norm": 0.11331377001939717,
      "learning_rate": 4.77899274897955e-06,
      "loss": 0.0359,
      "num_tokens": 103665303.0,
      "step": 1360
    },
    {
      "epoch": 1.6974422956955708,
      "grad_norm": 0.116217295144265,
      "learning_rate": 4.773012039402841e-06,
      "loss": 0.0351,
      "num_tokens": 103741004.0,
      "step": 1361
    },
    {
      "epoch": 1.6986899563318776,
      "grad_norm": 0.1176520133320377,
      "learning_rate": 4.767032648075043e-06,
      "loss": 0.0383,
      "num_tokens": 103817035.0,
      "step": 1362
    },
    {
      "epoch": 1.6999376169681848,
      "grad_norm": 0.12185429747688767,
      "learning_rate": 4.761054585838599e-06,
      "loss": 0.0372,
      "num_tokens": 103893856.0,
      "step": 1363
    },
    {
      "epoch": 1.7011852776044916,
      "grad_norm": 0.12816254491207327,
      "learning_rate": 4.755077863533541e-06,
      "loss": 0.0351,
      "num_tokens": 103970137.0,
      "step": 1364
    },
    {
      "epoch": 1.7024329382407986,
      "grad_norm": 0.11787000881995827,
      "learning_rate": 4.749102491997476e-06,
      "loss": 0.037,
      "num_tokens": 104046396.0,
      "step": 1365
    },
    {
      "epoch": 1.7036805988771055,
      "grad_norm": 0.10708489924117709,
      "learning_rate": 4.743128482065555e-06,
      "loss": 0.0351,
      "num_tokens": 104122836.0,
      "step": 1366
    },
    {
      "epoch": 1.7049282595134123,
      "grad_norm": 0.12224915882356628,
      "learning_rate": 4.737155844570468e-06,
      "loss": 0.0387,
      "num_tokens": 104200752.0,
      "step": 1367
    },
    {
      "epoch": 1.7061759201497193,
      "grad_norm": 0.11927797892198116,
      "learning_rate": 4.7311845903424104e-06,
      "loss": 0.0379,
      "num_tokens": 104276463.0,
      "step": 1368
    },
    {
      "epoch": 1.7074235807860263,
      "grad_norm": 0.11933248557644462,
      "learning_rate": 4.725214730209069e-06,
      "loss": 0.0376,
      "num_tokens": 104354190.0,
      "step": 1369
    },
    {
      "epoch": 1.708671241422333,
      "grad_norm": 0.11891032616900842,
      "learning_rate": 4.719246274995607e-06,
      "loss": 0.0339,
      "num_tokens": 104429464.0,
      "step": 1370
    },
    {
      "epoch": 1.70991890205864,
      "grad_norm": 0.11202623027798712,
      "learning_rate": 4.713279235524637e-06,
      "loss": 0.0377,
      "num_tokens": 104505684.0,
      "step": 1371
    },
    {
      "epoch": 1.711166562694947,
      "grad_norm": 0.11951117716097824,
      "learning_rate": 4.707313622616205e-06,
      "loss": 0.0354,
      "num_tokens": 104581004.0,
      "step": 1372
    },
    {
      "epoch": 1.7124142233312538,
      "grad_norm": 0.11819752380067314,
      "learning_rate": 4.701349447087769e-06,
      "loss": 0.0384,
      "num_tokens": 104658361.0,
      "step": 1373
    },
    {
      "epoch": 1.7136618839675608,
      "grad_norm": 0.11366388384713198,
      "learning_rate": 4.695386719754184e-06,
      "loss": 0.037,
      "num_tokens": 104734938.0,
      "step": 1374
    },
    {
      "epoch": 1.7149095446038678,
      "grad_norm": 0.12740314308973466,
      "learning_rate": 4.689425451427677e-06,
      "loss": 0.0356,
      "num_tokens": 104810696.0,
      "step": 1375
    },
    {
      "epoch": 1.7161572052401746,
      "grad_norm": 0.11222467035026419,
      "learning_rate": 4.683465652917828e-06,
      "loss": 0.0362,
      "num_tokens": 104886709.0,
      "step": 1376
    },
    {
      "epoch": 1.7174048658764816,
      "grad_norm": 0.11092206647603566,
      "learning_rate": 4.677507335031555e-06,
      "loss": 0.0364,
      "num_tokens": 104962238.0,
      "step": 1377
    },
    {
      "epoch": 1.7186525265127885,
      "grad_norm": 0.1160369465140294,
      "learning_rate": 4.671550508573087e-06,
      "loss": 0.0359,
      "num_tokens": 105037646.0,
      "step": 1378
    },
    {
      "epoch": 1.7199001871490953,
      "grad_norm": 0.11495558179367521,
      "learning_rate": 4.6655951843439514e-06,
      "loss": 0.0375,
      "num_tokens": 105113774.0,
      "step": 1379
    },
    {
      "epoch": 1.7211478477854025,
      "grad_norm": 0.12435135754767197,
      "learning_rate": 4.659641373142953e-06,
      "loss": 0.0386,
      "num_tokens": 105190747.0,
      "step": 1380
    },
    {
      "epoch": 1.7223955084217093,
      "grad_norm": 0.11693604602399771,
      "learning_rate": 4.653689085766147e-06,
      "loss": 0.0341,
      "num_tokens": 105267912.0,
      "step": 1381
    },
    {
      "epoch": 1.723643169058016,
      "grad_norm": 0.10558552346143842,
      "learning_rate": 4.6477383330068335e-06,
      "loss": 0.0367,
      "num_tokens": 105343487.0,
      "step": 1382
    },
    {
      "epoch": 1.7248908296943233,
      "grad_norm": 0.12228098738763672,
      "learning_rate": 4.641789125655526e-06,
      "loss": 0.0377,
      "num_tokens": 105420088.0,
      "step": 1383
    },
    {
      "epoch": 1.72613849033063,
      "grad_norm": 0.11328598660616651,
      "learning_rate": 4.6358414744999324e-06,
      "loss": 0.038,
      "num_tokens": 105496352.0,
      "step": 1384
    },
    {
      "epoch": 1.727386150966937,
      "grad_norm": 0.12964544697403055,
      "learning_rate": 4.6298953903249455e-06,
      "loss": 0.0394,
      "num_tokens": 105573063.0,
      "step": 1385
    },
    {
      "epoch": 1.728633811603244,
      "grad_norm": 0.1238931831498432,
      "learning_rate": 4.623950883912609e-06,
      "loss": 0.0367,
      "num_tokens": 105649475.0,
      "step": 1386
    },
    {
      "epoch": 1.7298814722395508,
      "grad_norm": 0.13042704795491192,
      "learning_rate": 4.618007966042114e-06,
      "loss": 0.0353,
      "num_tokens": 105725225.0,
      "step": 1387
    },
    {
      "epoch": 1.7311291328758578,
      "grad_norm": 0.11540798304998945,
      "learning_rate": 4.612066647489762e-06,
      "loss": 0.0346,
      "num_tokens": 105799845.0,
      "step": 1388
    },
    {
      "epoch": 1.7323767935121648,
      "grad_norm": 0.10019808616223308,
      "learning_rate": 4.606126939028965e-06,
      "loss": 0.0352,
      "num_tokens": 105875906.0,
      "step": 1389
    },
    {
      "epoch": 1.7336244541484715,
      "grad_norm": 0.11294492281764848,
      "learning_rate": 4.600188851430206e-06,
      "loss": 0.0339,
      "num_tokens": 105951201.0,
      "step": 1390
    },
    {
      "epoch": 1.7348721147847785,
      "grad_norm": 0.11156801308172176,
      "learning_rate": 4.594252395461036e-06,
      "loss": 0.0361,
      "num_tokens": 106027130.0,
      "step": 1391
    },
    {
      "epoch": 1.7361197754210855,
      "grad_norm": 0.11665884377063734,
      "learning_rate": 4.588317581886041e-06,
      "loss": 0.0362,
      "num_tokens": 106103346.0,
      "step": 1392
    },
    {
      "epoch": 1.7373674360573923,
      "grad_norm": 0.11454295291582704,
      "learning_rate": 4.5823844214668326e-06,
      "loss": 0.0381,
      "num_tokens": 106179489.0,
      "step": 1393
    },
    {
      "epoch": 1.7386150966936993,
      "grad_norm": 0.11318625067239595,
      "learning_rate": 4.576452924962024e-06,
      "loss": 0.0356,
      "num_tokens": 106255533.0,
      "step": 1394
    },
    {
      "epoch": 1.7398627573300063,
      "grad_norm": 0.11952096031253687,
      "learning_rate": 4.570523103127209e-06,
      "loss": 0.0369,
      "num_tokens": 106330931.0,
      "step": 1395
    },
    {
      "epoch": 1.741110417966313,
      "grad_norm": 0.12371072052697842,
      "learning_rate": 4.564594966714952e-06,
      "loss": 0.0345,
      "num_tokens": 106406216.0,
      "step": 1396
    },
    {
      "epoch": 1.74235807860262,
      "grad_norm": 0.10756413875277834,
      "learning_rate": 4.558668526474751e-06,
      "loss": 0.0369,
      "num_tokens": 106482048.0,
      "step": 1397
    },
    {
      "epoch": 1.743605739238927,
      "grad_norm": 0.10973946525613622,
      "learning_rate": 4.552743793153037e-06,
      "loss": 0.0374,
      "num_tokens": 106558458.0,
      "step": 1398
    },
    {
      "epoch": 1.7448533998752338,
      "grad_norm": 0.11554088759005555,
      "learning_rate": 4.5468207774931414e-06,
      "loss": 0.0371,
      "num_tokens": 106634391.0,
      "step": 1399
    },
    {
      "epoch": 1.746101060511541,
      "grad_norm": 0.11857009926153536,
      "learning_rate": 4.540899490235282e-06,
      "loss": 0.0369,
      "num_tokens": 106711587.0,
      "step": 1400
    },
    {
      "epoch": 1.7473487211478478,
      "grad_norm": 0.11113088782412625,
      "learning_rate": 4.534979942116542e-06,
      "loss": 0.0363,
      "num_tokens": 106787790.0,
      "step": 1401
    },
    {
      "epoch": 1.7485963817841546,
      "grad_norm": 0.11521710213865818,
      "learning_rate": 4.529062143870849e-06,
      "loss": 0.0363,
      "num_tokens": 106863918.0,
      "step": 1402
    },
    {
      "epoch": 1.7498440424204618,
      "grad_norm": 0.12355721482825922,
      "learning_rate": 4.5231461062289624e-06,
      "loss": 0.0352,
      "num_tokens": 106940044.0,
      "step": 1403
    },
    {
      "epoch": 1.7510917030567685,
      "grad_norm": 0.11027027150618844,
      "learning_rate": 4.5172318399184485e-06,
      "loss": 0.0371,
      "num_tokens": 107016503.0,
      "step": 1404
    },
    {
      "epoch": 1.7523393636930755,
      "grad_norm": 0.11557790009746331,
      "learning_rate": 4.511319355663657e-06,
      "loss": 0.0373,
      "num_tokens": 107093464.0,
      "step": 1405
    },
    {
      "epoch": 1.7535870243293825,
      "grad_norm": 0.10997539686939024,
      "learning_rate": 4.50540866418571e-06,
      "loss": 0.033,
      "num_tokens": 107169232.0,
      "step": 1406
    },
    {
      "epoch": 1.7548346849656893,
      "grad_norm": 0.10508302585453105,
      "learning_rate": 4.499499776202476e-06,
      "loss": 0.0365,
      "num_tokens": 107246281.0,
      "step": 1407
    },
    {
      "epoch": 1.7560823456019963,
      "grad_norm": 0.11986514390120263,
      "learning_rate": 4.493592702428558e-06,
      "loss": 0.0355,
      "num_tokens": 107322121.0,
      "step": 1408
    },
    {
      "epoch": 1.7573300062383033,
      "grad_norm": 0.11382666929205047,
      "learning_rate": 4.487687453575261e-06,
      "loss": 0.0349,
      "num_tokens": 107397732.0,
      "step": 1409
    },
    {
      "epoch": 1.75857766687461,
      "grad_norm": 0.11452741596501294,
      "learning_rate": 4.481784040350593e-06,
      "loss": 0.0377,
      "num_tokens": 107474105.0,
      "step": 1410
    },
    {
      "epoch": 1.759825327510917,
      "grad_norm": 0.11978232669192639,
      "learning_rate": 4.475882473459221e-06,
      "loss": 0.0346,
      "num_tokens": 107549721.0,
      "step": 1411
    },
    {
      "epoch": 1.761072988147224,
      "grad_norm": 0.10810211082067261,
      "learning_rate": 4.469982763602473e-06,
      "loss": 0.0377,
      "num_tokens": 107626252.0,
      "step": 1412
    },
    {
      "epoch": 1.7623206487835308,
      "grad_norm": 0.12430884790032933,
      "learning_rate": 4.464084921478303e-06,
      "loss": 0.0362,
      "num_tokens": 107701970.0,
      "step": 1413
    },
    {
      "epoch": 1.7635683094198378,
      "grad_norm": 0.1080229919974577,
      "learning_rate": 4.458188957781285e-06,
      "loss": 0.0363,
      "num_tokens": 107778230.0,
      "step": 1414
    },
    {
      "epoch": 1.7648159700561448,
      "grad_norm": 0.12546433579895047,
      "learning_rate": 4.452294883202581e-06,
      "loss": 0.0354,
      "num_tokens": 107854033.0,
      "step": 1415
    },
    {
      "epoch": 1.7660636306924515,
      "grad_norm": 0.10862233462213136,
      "learning_rate": 4.44640270842993e-06,
      "loss": 0.0344,
      "num_tokens": 107928602.0,
      "step": 1416
    },
    {
      "epoch": 1.7673112913287585,
      "grad_norm": 0.11234450273923224,
      "learning_rate": 4.440512444147626e-06,
      "loss": 0.0356,
      "num_tokens": 108005283.0,
      "step": 1417
    },
    {
      "epoch": 1.7685589519650655,
      "grad_norm": 0.12290779064818019,
      "learning_rate": 4.434624101036498e-06,
      "loss": 0.0408,
      "num_tokens": 108082409.0,
      "step": 1418
    },
    {
      "epoch": 1.7698066126013723,
      "grad_norm": 0.12019256254394751,
      "learning_rate": 4.4287376897738945e-06,
      "loss": 0.038,
      "num_tokens": 108159111.0,
      "step": 1419
    },
    {
      "epoch": 1.7710542732376795,
      "grad_norm": 0.12938504724279448,
      "learning_rate": 4.4228532210336535e-06,
      "loss": 0.0368,
      "num_tokens": 108236277.0,
      "step": 1420
    },
    {
      "epoch": 1.7723019338739863,
      "grad_norm": 0.12063659002771691,
      "learning_rate": 4.4169707054861e-06,
      "loss": 0.0356,
      "num_tokens": 108313163.0,
      "step": 1421
    },
    {
      "epoch": 1.773549594510293,
      "grad_norm": 0.12467033356761577,
      "learning_rate": 4.411090153798011e-06,
      "loss": 0.0357,
      "num_tokens": 108391348.0,
      "step": 1422
    },
    {
      "epoch": 1.7747972551466002,
      "grad_norm": 0.11455930343881551,
      "learning_rate": 4.405211576632602e-06,
      "loss": 0.0348,
      "num_tokens": 108467824.0,
      "step": 1423
    },
    {
      "epoch": 1.776044915782907,
      "grad_norm": 0.1069971790671395,
      "learning_rate": 4.3993349846495136e-06,
      "loss": 0.0341,
      "num_tokens": 108544120.0,
      "step": 1424
    },
    {
      "epoch": 1.777292576419214,
      "grad_norm": 0.11837827088623372,
      "learning_rate": 4.393460388504784e-06,
      "loss": 0.0355,
      "num_tokens": 108620246.0,
      "step": 1425
    },
    {
      "epoch": 1.778540237055521,
      "grad_norm": 0.11802718964613297,
      "learning_rate": 4.387587798850826e-06,
      "loss": 0.0358,
      "num_tokens": 108696646.0,
      "step": 1426
    },
    {
      "epoch": 1.7797878976918278,
      "grad_norm": 0.10473684840636525,
      "learning_rate": 4.381717226336426e-06,
      "loss": 0.0382,
      "num_tokens": 108773998.0,
      "step": 1427
    },
    {
      "epoch": 1.7810355583281348,
      "grad_norm": 0.1156664764311051,
      "learning_rate": 4.375848681606704e-06,
      "loss": 0.0349,
      "num_tokens": 108850108.0,
      "step": 1428
    },
    {
      "epoch": 1.7822832189644418,
      "grad_norm": 0.1186189185581627,
      "learning_rate": 4.369982175303104e-06,
      "loss": 0.0369,
      "num_tokens": 108927238.0,
      "step": 1429
    },
    {
      "epoch": 1.7835308796007485,
      "grad_norm": 0.12029945296887255,
      "learning_rate": 4.364117718063375e-06,
      "loss": 0.0362,
      "num_tokens": 109003852.0,
      "step": 1430
    },
    {
      "epoch": 1.7847785402370555,
      "grad_norm": 0.11467347487239687,
      "learning_rate": 4.358255320521553e-06,
      "loss": 0.0381,
      "num_tokens": 109080943.0,
      "step": 1431
    },
    {
      "epoch": 1.7860262008733625,
      "grad_norm": 0.11219671293656201,
      "learning_rate": 4.352394993307935e-06,
      "loss": 0.0353,
      "num_tokens": 109156308.0,
      "step": 1432
    },
    {
      "epoch": 1.7872738615096693,
      "grad_norm": 0.1195758432084249,
      "learning_rate": 4.346536747049068e-06,
      "loss": 0.0383,
      "num_tokens": 109233773.0,
      "step": 1433
    },
    {
      "epoch": 1.7885215221459763,
      "grad_norm": 0.12257041651086298,
      "learning_rate": 4.340680592367721e-06,
      "loss": 0.0367,
      "num_tokens": 109309017.0,
      "step": 1434
    },
    {
      "epoch": 1.7897691827822833,
      "grad_norm": 0.11644283699770981,
      "learning_rate": 4.33482653988287e-06,
      "loss": 0.0365,
      "num_tokens": 109386304.0,
      "step": 1435
    },
    {
      "epoch": 1.79101684341859,
      "grad_norm": 0.1159891995707703,
      "learning_rate": 4.328974600209687e-06,
      "loss": 0.0364,
      "num_tokens": 109462078.0,
      "step": 1436
    },
    {
      "epoch": 1.7922645040548972,
      "grad_norm": 0.11271455439524168,
      "learning_rate": 4.3231247839595045e-06,
      "loss": 0.0356,
      "num_tokens": 109538386.0,
      "step": 1437
    },
    {
      "epoch": 1.793512164691204,
      "grad_norm": 0.11081718708525366,
      "learning_rate": 4.317277101739806e-06,
      "loss": 0.0352,
      "num_tokens": 109614385.0,
      "step": 1438
    },
    {
      "epoch": 1.7947598253275108,
      "grad_norm": 0.1125273822950325,
      "learning_rate": 4.3114315641542105e-06,
      "loss": 0.0356,
      "num_tokens": 109690309.0,
      "step": 1439
    },
    {
      "epoch": 1.796007485963818,
      "grad_norm": 0.10944966687820053,
      "learning_rate": 4.305588181802441e-06,
      "loss": 0.0348,
      "num_tokens": 109766780.0,
      "step": 1440
    },
    {
      "epoch": 1.7972551466001248,
      "grad_norm": 0.11460545262390623,
      "learning_rate": 4.2997469652803185e-06,
      "loss": 0.0345,
      "num_tokens": 109843255.0,
      "step": 1441
    },
    {
      "epoch": 1.7985028072364317,
      "grad_norm": 0.11010943084688467,
      "learning_rate": 4.293907925179733e-06,
      "loss": 0.0365,
      "num_tokens": 109920554.0,
      "step": 1442
    },
    {
      "epoch": 1.7997504678727387,
      "grad_norm": 0.12295684254699428,
      "learning_rate": 4.28807107208863e-06,
      "loss": 0.0392,
      "num_tokens": 109997732.0,
      "step": 1443
    },
    {
      "epoch": 1.8009981285090455,
      "grad_norm": 0.13807569493569224,
      "learning_rate": 4.282236416590986e-06,
      "loss": 0.037,
      "num_tokens": 110074362.0,
      "step": 1444
    },
    {
      "epoch": 1.8022457891453525,
      "grad_norm": 0.11589812396647654,
      "learning_rate": 4.276403969266797e-06,
      "loss": 0.0334,
      "num_tokens": 110149499.0,
      "step": 1445
    },
    {
      "epoch": 1.8034934497816595,
      "grad_norm": 0.11874331032886756,
      "learning_rate": 4.270573740692053e-06,
      "loss": 0.0364,
      "num_tokens": 110225383.0,
      "step": 1446
    },
    {
      "epoch": 1.8047411104179663,
      "grad_norm": 0.10769893454787335,
      "learning_rate": 4.2647457414387205e-06,
      "loss": 0.0346,
      "num_tokens": 110300178.0,
      "step": 1447
    },
    {
      "epoch": 1.8059887710542732,
      "grad_norm": 0.10952676895213141,
      "learning_rate": 4.2589199820747226e-06,
      "loss": 0.0341,
      "num_tokens": 110378682.0,
      "step": 1448
    },
    {
      "epoch": 1.8072364316905802,
      "grad_norm": 0.11986450037494077,
      "learning_rate": 4.253096473163923e-06,
      "loss": 0.0357,
      "num_tokens": 110453993.0,
      "step": 1449
    },
    {
      "epoch": 1.808484092326887,
      "grad_norm": 0.10575750292359529,
      "learning_rate": 4.247275225266103e-06,
      "loss": 0.0368,
      "num_tokens": 110530495.0,
      "step": 1450
    },
    {
      "epoch": 1.809731752963194,
      "grad_norm": 0.1211718172176368,
      "learning_rate": 4.241456248936946e-06,
      "loss": 0.0374,
      "num_tokens": 110606906.0,
      "step": 1451
    },
    {
      "epoch": 1.810979413599501,
      "grad_norm": 0.11575168655236709,
      "learning_rate": 4.23563955472801e-06,
      "loss": 0.0357,
      "num_tokens": 110683315.0,
      "step": 1452
    },
    {
      "epoch": 1.8122270742358078,
      "grad_norm": 0.11205750328573585,
      "learning_rate": 4.229825153186727e-06,
      "loss": 0.0338,
      "num_tokens": 110758384.0,
      "step": 1453
    },
    {
      "epoch": 1.8134747348721147,
      "grad_norm": 0.1109122728392621,
      "learning_rate": 4.22401305485636e-06,
      "loss": 0.0362,
      "num_tokens": 110834195.0,
      "step": 1454
    },
    {
      "epoch": 1.8147223955084217,
      "grad_norm": 0.11438862268810039,
      "learning_rate": 4.218203270276e-06,
      "loss": 0.0355,
      "num_tokens": 110910964.0,
      "step": 1455
    },
    {
      "epoch": 1.8159700561447285,
      "grad_norm": 0.11552343707809745,
      "learning_rate": 4.2123958099805466e-06,
      "loss": 0.0364,
      "num_tokens": 110987489.0,
      "step": 1456
    },
    {
      "epoch": 1.8172177167810357,
      "grad_norm": 0.1128780835528746,
      "learning_rate": 4.206590684500675e-06,
      "loss": 0.0352,
      "num_tokens": 111063323.0,
      "step": 1457
    },
    {
      "epoch": 1.8184653774173425,
      "grad_norm": 0.11606306376773205,
      "learning_rate": 4.200787904362833e-06,
      "loss": 0.0365,
      "num_tokens": 111139548.0,
      "step": 1458
    },
    {
      "epoch": 1.8197130380536493,
      "grad_norm": 0.10291445670113526,
      "learning_rate": 4.194987480089218e-06,
      "loss": 0.0351,
      "num_tokens": 111216126.0,
      "step": 1459
    },
    {
      "epoch": 1.8209606986899565,
      "grad_norm": 0.11644649365731295,
      "learning_rate": 4.189189422197751e-06,
      "loss": 0.0348,
      "num_tokens": 111291904.0,
      "step": 1460
    },
    {
      "epoch": 1.8222083593262632,
      "grad_norm": 0.11302317074819644,
      "learning_rate": 4.183393741202065e-06,
      "loss": 0.0362,
      "num_tokens": 111367687.0,
      "step": 1461
    },
    {
      "epoch": 1.8234560199625702,
      "grad_norm": 0.11555775768411664,
      "learning_rate": 4.177600447611478e-06,
      "loss": 0.0358,
      "num_tokens": 111443817.0,
      "step": 1462
    },
    {
      "epoch": 1.8247036805988772,
      "grad_norm": 0.11640861821564545,
      "learning_rate": 4.171809551930985e-06,
      "loss": 0.0389,
      "num_tokens": 111521320.0,
      "step": 1463
    },
    {
      "epoch": 1.825951341235184,
      "grad_norm": 0.11354843479910538,
      "learning_rate": 4.166021064661231e-06,
      "loss": 0.038,
      "num_tokens": 111599182.0,
      "step": 1464
    },
    {
      "epoch": 1.827199001871491,
      "grad_norm": 0.11962115255796753,
      "learning_rate": 4.160234996298491e-06,
      "loss": 0.035,
      "num_tokens": 111675551.0,
      "step": 1465
    },
    {
      "epoch": 1.828446662507798,
      "grad_norm": 0.11860626927504106,
      "learning_rate": 4.154451357334654e-06,
      "loss": 0.0361,
      "num_tokens": 111754629.0,
      "step": 1466
    },
    {
      "epoch": 1.8296943231441047,
      "grad_norm": 0.1087452520721984,
      "learning_rate": 4.148670158257211e-06,
      "loss": 0.0359,
      "num_tokens": 111829688.0,
      "step": 1467
    },
    {
      "epoch": 1.8309419837804117,
      "grad_norm": 0.12262396185253335,
      "learning_rate": 4.142891409549219e-06,
      "loss": 0.036,
      "num_tokens": 111906321.0,
      "step": 1468
    },
    {
      "epoch": 1.8321896444167187,
      "grad_norm": 0.11301164282196123,
      "learning_rate": 4.137115121689297e-06,
      "loss": 0.035,
      "num_tokens": 111981952.0,
      "step": 1469
    },
    {
      "epoch": 1.8334373050530255,
      "grad_norm": 0.12102968567645797,
      "learning_rate": 4.131341305151603e-06,
      "loss": 0.0381,
      "num_tokens": 112058744.0,
      "step": 1470
    },
    {
      "epoch": 1.8346849656893325,
      "grad_norm": 0.11651089328857946,
      "learning_rate": 4.1255699704058085e-06,
      "loss": 0.0365,
      "num_tokens": 112135004.0,
      "step": 1471
    },
    {
      "epoch": 1.8359326263256395,
      "grad_norm": 0.12048112446389629,
      "learning_rate": 4.119801127917089e-06,
      "loss": 0.0358,
      "num_tokens": 112211639.0,
      "step": 1472
    },
    {
      "epoch": 1.8371802869619462,
      "grad_norm": 0.11324293941061209,
      "learning_rate": 4.114034788146101e-06,
      "loss": 0.0351,
      "num_tokens": 112287073.0,
      "step": 1473
    },
    {
      "epoch": 1.8384279475982532,
      "grad_norm": 0.11065510997356777,
      "learning_rate": 4.108270961548957e-06,
      "loss": 0.035,
      "num_tokens": 112362219.0,
      "step": 1474
    },
    {
      "epoch": 1.8396756082345602,
      "grad_norm": 0.1235587604718455,
      "learning_rate": 4.102509658577223e-06,
      "loss": 0.0358,
      "num_tokens": 112438186.0,
      "step": 1475
    },
    {
      "epoch": 1.840923268870867,
      "grad_norm": 0.10529808856557056,
      "learning_rate": 4.096750889677878e-06,
      "loss": 0.0332,
      "num_tokens": 112513405.0,
      "step": 1476
    },
    {
      "epoch": 1.8421709295071742,
      "grad_norm": 0.11603734080740498,
      "learning_rate": 4.090994665293313e-06,
      "loss": 0.0345,
      "num_tokens": 112588629.0,
      "step": 1477
    },
    {
      "epoch": 1.843418590143481,
      "grad_norm": 0.10787689083647872,
      "learning_rate": 4.085240995861301e-06,
      "loss": 0.0357,
      "num_tokens": 112664529.0,
      "step": 1478
    },
    {
      "epoch": 1.8446662507797877,
      "grad_norm": 0.11127451349079631,
      "learning_rate": 4.079489891814986e-06,
      "loss": 0.0355,
      "num_tokens": 112741228.0,
      "step": 1479
    },
    {
      "epoch": 1.845913911416095,
      "grad_norm": 0.13166792903179028,
      "learning_rate": 4.073741363582856e-06,
      "loss": 0.0371,
      "num_tokens": 112817011.0,
      "step": 1480
    },
    {
      "epoch": 1.8471615720524017,
      "grad_norm": 0.11657324844692983,
      "learning_rate": 4.06799542158873e-06,
      "loss": 0.0355,
      "num_tokens": 112892718.0,
      "step": 1481
    },
    {
      "epoch": 1.8484092326887087,
      "grad_norm": 0.10109047613936198,
      "learning_rate": 4.062252076251739e-06,
      "loss": 0.0349,
      "num_tokens": 112970058.0,
      "step": 1482
    },
    {
      "epoch": 1.8496568933250157,
      "grad_norm": 0.1095850182691673,
      "learning_rate": 4.056511337986304e-06,
      "loss": 0.0354,
      "num_tokens": 113046254.0,
      "step": 1483
    },
    {
      "epoch": 1.8509045539613225,
      "grad_norm": 0.1141555383648403,
      "learning_rate": 4.05077321720212e-06,
      "loss": 0.0377,
      "num_tokens": 113122580.0,
      "step": 1484
    },
    {
      "epoch": 1.8521522145976295,
      "grad_norm": 0.11965767255633106,
      "learning_rate": 4.045037724304129e-06,
      "loss": 0.0373,
      "num_tokens": 113199298.0,
      "step": 1485
    },
    {
      "epoch": 1.8533998752339365,
      "grad_norm": 0.12531984936041318,
      "learning_rate": 4.039304869692518e-06,
      "loss": 0.0374,
      "num_tokens": 113275703.0,
      "step": 1486
    },
    {
      "epoch": 1.8546475358702432,
      "grad_norm": 0.10134551113093508,
      "learning_rate": 4.033574663762685e-06,
      "loss": 0.0355,
      "num_tokens": 113351254.0,
      "step": 1487
    },
    {
      "epoch": 1.8558951965065502,
      "grad_norm": 0.11773826372613343,
      "learning_rate": 4.0278471169052224e-06,
      "loss": 0.0372,
      "num_tokens": 113427252.0,
      "step": 1488
    },
    {
      "epoch": 1.8571428571428572,
      "grad_norm": 0.11661724552812294,
      "learning_rate": 4.022122239505906e-06,
      "loss": 0.0358,
      "num_tokens": 113503323.0,
      "step": 1489
    },
    {
      "epoch": 1.858390517779164,
      "grad_norm": 0.11749716033346444,
      "learning_rate": 4.0164000419456715e-06,
      "loss": 0.0375,
      "num_tokens": 113580371.0,
      "step": 1490
    },
    {
      "epoch": 1.859638178415471,
      "grad_norm": 0.1305695968942797,
      "learning_rate": 4.010680534600587e-06,
      "loss": 0.0356,
      "num_tokens": 113656947.0,
      "step": 1491
    },
    {
      "epoch": 1.860885839051778,
      "grad_norm": 0.11359160111796697,
      "learning_rate": 4.004963727841852e-06,
      "loss": 0.0341,
      "num_tokens": 113732938.0,
      "step": 1492
    },
    {
      "epoch": 1.8621334996880847,
      "grad_norm": 0.10258761359426599,
      "learning_rate": 3.9992496320357645e-06,
      "loss": 0.0336,
      "num_tokens": 113809002.0,
      "step": 1493
    },
    {
      "epoch": 1.8633811603243917,
      "grad_norm": 0.11804430357862333,
      "learning_rate": 3.993538257543706e-06,
      "loss": 0.0328,
      "num_tokens": 113883796.0,
      "step": 1494
    },
    {
      "epoch": 1.8646288209606987,
      "grad_norm": 0.12483243738508598,
      "learning_rate": 3.987829614722124e-06,
      "loss": 0.0371,
      "num_tokens": 113960768.0,
      "step": 1495
    },
    {
      "epoch": 1.8658764815970055,
      "grad_norm": 0.1150151462214644,
      "learning_rate": 3.982123713922517e-06,
      "loss": 0.0358,
      "num_tokens": 114036180.0,
      "step": 1496
    },
    {
      "epoch": 1.8671241422333127,
      "grad_norm": 0.11869044286652397,
      "learning_rate": 3.976420565491404e-06,
      "loss": 0.0369,
      "num_tokens": 114111980.0,
      "step": 1497
    },
    {
      "epoch": 1.8683718028696195,
      "grad_norm": 0.11008881374766814,
      "learning_rate": 3.970720179770322e-06,
      "loss": 0.0346,
      "num_tokens": 114187724.0,
      "step": 1498
    },
    {
      "epoch": 1.8696194635059262,
      "grad_norm": 0.11080992842795516,
      "learning_rate": 3.965022567095788e-06,
      "loss": 0.0363,
      "num_tokens": 114263240.0,
      "step": 1499
    },
    {
      "epoch": 1.8708671241422334,
      "grad_norm": 0.1092058974188812,
      "learning_rate": 3.959327737799298e-06,
      "loss": 0.0353,
      "num_tokens": 114338209.0,
      "step": 1500
    },
    {
      "epoch": 1.8721147847785402,
      "grad_norm": 0.11386637728893095,
      "learning_rate": 3.953635702207299e-06,
      "loss": 0.0368,
      "num_tokens": 114415547.0,
      "step": 1501
    },
    {
      "epoch": 1.8733624454148472,
      "grad_norm": 0.1155459395703279,
      "learning_rate": 3.947946470641169e-06,
      "loss": 0.0366,
      "num_tokens": 114492678.0,
      "step": 1502
    },
    {
      "epoch": 1.8746101060511542,
      "grad_norm": 0.12225214682735673,
      "learning_rate": 3.9422600534172105e-06,
      "loss": 0.0373,
      "num_tokens": 114569035.0,
      "step": 1503
    },
    {
      "epoch": 1.875857766687461,
      "grad_norm": 0.11191387297145151,
      "learning_rate": 3.936576460846614e-06,
      "loss": 0.0345,
      "num_tokens": 114644776.0,
      "step": 1504
    },
    {
      "epoch": 1.877105427323768,
      "grad_norm": 0.1281184082430015,
      "learning_rate": 3.930895703235448e-06,
      "loss": 0.0358,
      "num_tokens": 114720908.0,
      "step": 1505
    },
    {
      "epoch": 1.878353087960075,
      "grad_norm": 0.11720996450584123,
      "learning_rate": 3.925217790884646e-06,
      "loss": 0.0334,
      "num_tokens": 114796173.0,
      "step": 1506
    },
    {
      "epoch": 1.8796007485963817,
      "grad_norm": 0.10440882158568637,
      "learning_rate": 3.919542734089978e-06,
      "loss": 0.0359,
      "num_tokens": 114872400.0,
      "step": 1507
    },
    {
      "epoch": 1.8808484092326887,
      "grad_norm": 0.11942205769199313,
      "learning_rate": 3.913870543142038e-06,
      "loss": 0.036,
      "num_tokens": 114948412.0,
      "step": 1508
    },
    {
      "epoch": 1.8820960698689957,
      "grad_norm": 0.11278966937999528,
      "learning_rate": 3.908201228326222e-06,
      "loss": 0.0371,
      "num_tokens": 115025101.0,
      "step": 1509
    },
    {
      "epoch": 1.8833437305053025,
      "grad_norm": 0.12097590188866718,
      "learning_rate": 3.902534799922713e-06,
      "loss": 0.034,
      "num_tokens": 115100026.0,
      "step": 1510
    },
    {
      "epoch": 1.8845913911416095,
      "grad_norm": 0.11125819036903851,
      "learning_rate": 3.896871268206456e-06,
      "loss": 0.0333,
      "num_tokens": 115175198.0,
      "step": 1511
    },
    {
      "epoch": 1.8858390517779164,
      "grad_norm": 0.13253062734929877,
      "learning_rate": 3.8912106434471486e-06,
      "loss": 0.0609,
      "num_tokens": 115251378.0,
      "step": 1512
    },
    {
      "epoch": 1.8870867124142232,
      "grad_norm": 0.12632121435115587,
      "learning_rate": 3.885552935909212e-06,
      "loss": 0.0362,
      "num_tokens": 115327638.0,
      "step": 1513
    },
    {
      "epoch": 1.8883343730505302,
      "grad_norm": 0.11211297136521399,
      "learning_rate": 3.879898155851779e-06,
      "loss": 0.0346,
      "num_tokens": 115404601.0,
      "step": 1514
    },
    {
      "epoch": 1.8895820336868372,
      "grad_norm": 0.10494844882789488,
      "learning_rate": 3.874246313528679e-06,
      "loss": 0.0341,
      "num_tokens": 115479999.0,
      "step": 1515
    },
    {
      "epoch": 1.890829694323144,
      "grad_norm": 0.10745141921792603,
      "learning_rate": 3.868597419188409e-06,
      "loss": 0.0352,
      "num_tokens": 115555643.0,
      "step": 1516
    },
    {
      "epoch": 1.8920773549594512,
      "grad_norm": 0.11732863353127684,
      "learning_rate": 3.862951483074119e-06,
      "loss": 0.0374,
      "num_tokens": 115632535.0,
      "step": 1517
    },
    {
      "epoch": 1.893325015595758,
      "grad_norm": 0.11654268076982084,
      "learning_rate": 3.857308515423601e-06,
      "loss": 0.0367,
      "num_tokens": 115710345.0,
      "step": 1518
    },
    {
      "epoch": 1.8945726762320647,
      "grad_norm": 0.11351548860109464,
      "learning_rate": 3.851668526469261e-06,
      "loss": 0.0351,
      "num_tokens": 115786951.0,
      "step": 1519
    },
    {
      "epoch": 1.895820336868372,
      "grad_norm": 0.11602516099241723,
      "learning_rate": 3.846031526438102e-06,
      "loss": 0.0398,
      "num_tokens": 115864384.0,
      "step": 1520
    },
    {
      "epoch": 1.8970679975046787,
      "grad_norm": 0.11306702153588803,
      "learning_rate": 3.84039752555171e-06,
      "loss": 0.0333,
      "num_tokens": 115940486.0,
      "step": 1521
    },
    {
      "epoch": 1.8983156581409857,
      "grad_norm": 0.1087232323419847,
      "learning_rate": 3.834766534026231e-06,
      "loss": 0.0342,
      "num_tokens": 116015978.0,
      "step": 1522
    },
    {
      "epoch": 1.8995633187772927,
      "grad_norm": 0.1108836772324257,
      "learning_rate": 3.829138562072353e-06,
      "loss": 0.0352,
      "num_tokens": 116091783.0,
      "step": 1523
    },
    {
      "epoch": 1.9008109794135994,
      "grad_norm": 0.12237650658630188,
      "learning_rate": 3.823513619895293e-06,
      "loss": 0.036,
      "num_tokens": 116167365.0,
      "step": 1524
    },
    {
      "epoch": 1.9020586400499064,
      "grad_norm": 0.10179348152240558,
      "learning_rate": 3.81789171769477e-06,
      "loss": 0.0349,
      "num_tokens": 116242859.0,
      "step": 1525
    },
    {
      "epoch": 1.9033063006862134,
      "grad_norm": 0.13829513509173896,
      "learning_rate": 3.812272865664994e-06,
      "loss": 0.0352,
      "num_tokens": 116319663.0,
      "step": 1526
    },
    {
      "epoch": 1.9045539613225202,
      "grad_norm": 0.1074739931014396,
      "learning_rate": 3.8066570739946394e-06,
      "loss": 0.0353,
      "num_tokens": 116396800.0,
      "step": 1527
    },
    {
      "epoch": 1.9058016219588272,
      "grad_norm": 0.1104254235716302,
      "learning_rate": 3.801044352866834e-06,
      "loss": 0.0338,
      "num_tokens": 116473030.0,
      "step": 1528
    },
    {
      "epoch": 1.9070492825951342,
      "grad_norm": 0.12040467420139594,
      "learning_rate": 3.7954347124591395e-06,
      "loss": 0.0381,
      "num_tokens": 116551011.0,
      "step": 1529
    },
    {
      "epoch": 1.908296943231441,
      "grad_norm": 0.12410608661446856,
      "learning_rate": 3.7898281629435286e-06,
      "loss": 0.0367,
      "num_tokens": 116627045.0,
      "step": 1530
    },
    {
      "epoch": 1.909544603867748,
      "grad_norm": 0.11448003184524622,
      "learning_rate": 3.7842247144863686e-06,
      "loss": 0.0329,
      "num_tokens": 116702568.0,
      "step": 1531
    },
    {
      "epoch": 1.910792264504055,
      "grad_norm": 0.1101418263894307,
      "learning_rate": 3.778624377248409e-06,
      "loss": 0.036,
      "num_tokens": 116778196.0,
      "step": 1532
    },
    {
      "epoch": 1.9120399251403617,
      "grad_norm": 0.1181310983123205,
      "learning_rate": 3.77302716138475e-06,
      "loss": 0.0351,
      "num_tokens": 116854139.0,
      "step": 1533
    },
    {
      "epoch": 1.913287585776669,
      "grad_norm": 0.12725788230803384,
      "learning_rate": 3.7674330770448374e-06,
      "loss": 0.0381,
      "num_tokens": 116930576.0,
      "step": 1534
    },
    {
      "epoch": 1.9145352464129757,
      "grad_norm": 0.1263442950413413,
      "learning_rate": 3.7618421343724386e-06,
      "loss": 0.0365,
      "num_tokens": 117010379.0,
      "step": 1535
    },
    {
      "epoch": 1.9157829070492824,
      "grad_norm": 0.11884255667209397,
      "learning_rate": 3.756254343505621e-06,
      "loss": 0.035,
      "num_tokens": 117086064.0,
      "step": 1536
    },
    {
      "epoch": 1.9170305676855897,
      "grad_norm": 0.11112128984019812,
      "learning_rate": 3.7506697145767367e-06,
      "loss": 0.0362,
      "num_tokens": 117162657.0,
      "step": 1537
    },
    {
      "epoch": 1.9182782283218964,
      "grad_norm": 0.11780736604644747,
      "learning_rate": 3.745088257712408e-06,
      "loss": 0.0373,
      "num_tokens": 117239941.0,
      "step": 1538
    },
    {
      "epoch": 1.9195258889582034,
      "grad_norm": 0.10960815283331832,
      "learning_rate": 3.7395099830335034e-06,
      "loss": 0.0341,
      "num_tokens": 117314955.0,
      "step": 1539
    },
    {
      "epoch": 1.9207735495945104,
      "grad_norm": 0.10987391122252545,
      "learning_rate": 3.7339349006551193e-06,
      "loss": 0.0362,
      "num_tokens": 117390856.0,
      "step": 1540
    },
    {
      "epoch": 1.9220212102308172,
      "grad_norm": 0.12286203958441236,
      "learning_rate": 3.7283630206865696e-06,
      "loss": 0.0348,
      "num_tokens": 117466317.0,
      "step": 1541
    },
    {
      "epoch": 1.9232688708671242,
      "grad_norm": 0.1136853088390901,
      "learning_rate": 3.7227943532313504e-06,
      "loss": 0.0371,
      "num_tokens": 117544324.0,
      "step": 1542
    },
    {
      "epoch": 1.9245165315034312,
      "grad_norm": 0.11982734473388944,
      "learning_rate": 3.7172289083871436e-06,
      "loss": 0.0362,
      "num_tokens": 117620317.0,
      "step": 1543
    },
    {
      "epoch": 1.925764192139738,
      "grad_norm": 0.12175325190701178,
      "learning_rate": 3.7116666962457813e-06,
      "loss": 0.0369,
      "num_tokens": 117696600.0,
      "step": 1544
    },
    {
      "epoch": 1.927011852776045,
      "grad_norm": 0.11092763932924518,
      "learning_rate": 3.7061077268932333e-06,
      "loss": 0.0351,
      "num_tokens": 117772431.0,
      "step": 1545
    },
    {
      "epoch": 1.928259513412352,
      "grad_norm": 0.11266769207847174,
      "learning_rate": 3.700552010409596e-06,
      "loss": 0.0377,
      "num_tokens": 117847727.0,
      "step": 1546
    },
    {
      "epoch": 1.9295071740486587,
      "grad_norm": 0.12131889331136142,
      "learning_rate": 3.694999556869059e-06,
      "loss": 0.0353,
      "num_tokens": 117923710.0,
      "step": 1547
    },
    {
      "epoch": 1.9307548346849657,
      "grad_norm": 0.10375682393956039,
      "learning_rate": 3.6894503763399003e-06,
      "loss": 0.0354,
      "num_tokens": 118000196.0,
      "step": 1548
    },
    {
      "epoch": 1.9320024953212727,
      "grad_norm": 0.12150821898167924,
      "learning_rate": 3.683904478884461e-06,
      "loss": 0.0368,
      "num_tokens": 118076950.0,
      "step": 1549
    },
    {
      "epoch": 1.9332501559575794,
      "grad_norm": 0.10745802148743575,
      "learning_rate": 3.67836187455913e-06,
      "loss": 0.0362,
      "num_tokens": 118152566.0,
      "step": 1550
    },
    {
      "epoch": 1.9344978165938864,
      "grad_norm": 0.11641734699356358,
      "learning_rate": 3.672822573414323e-06,
      "loss": 0.035,
      "num_tokens": 118228975.0,
      "step": 1551
    },
    {
      "epoch": 1.9357454772301934,
      "grad_norm": 0.113905776443254,
      "learning_rate": 3.6672865854944673e-06,
      "loss": 0.0361,
      "num_tokens": 118306295.0,
      "step": 1552
    },
    {
      "epoch": 1.9369931378665002,
      "grad_norm": 0.11805635474708707,
      "learning_rate": 3.6617539208379836e-06,
      "loss": 0.0344,
      "num_tokens": 118382040.0,
      "step": 1553
    },
    {
      "epoch": 1.9382407985028074,
      "grad_norm": 0.11668783333522152,
      "learning_rate": 3.656224589477264e-06,
      "loss": 0.0353,
      "num_tokens": 118457858.0,
      "step": 1554
    },
    {
      "epoch": 1.9394884591391142,
      "grad_norm": 0.11690683661248272,
      "learning_rate": 3.65069860143866e-06,
      "loss": 0.0351,
      "num_tokens": 118534064.0,
      "step": 1555
    },
    {
      "epoch": 1.940736119775421,
      "grad_norm": 0.12479882763354747,
      "learning_rate": 3.645175966742456e-06,
      "loss": 0.0361,
      "num_tokens": 118609978.0,
      "step": 1556
    },
    {
      "epoch": 1.9419837804117281,
      "grad_norm": 0.10499026791561165,
      "learning_rate": 3.639656695402858e-06,
      "loss": 0.035,
      "num_tokens": 118686812.0,
      "step": 1557
    },
    {
      "epoch": 1.943231441048035,
      "grad_norm": 0.11043958990536244,
      "learning_rate": 3.634140797427974e-06,
      "loss": 0.0355,
      "num_tokens": 118762829.0,
      "step": 1558
    },
    {
      "epoch": 1.944479101684342,
      "grad_norm": 0.12081921654678775,
      "learning_rate": 3.6286282828197904e-06,
      "loss": 0.0361,
      "num_tokens": 118838419.0,
      "step": 1559
    },
    {
      "epoch": 1.945726762320649,
      "grad_norm": 0.11487103811024217,
      "learning_rate": 3.623119161574169e-06,
      "loss": 0.0356,
      "num_tokens": 118914155.0,
      "step": 1560
    },
    {
      "epoch": 1.9469744229569557,
      "grad_norm": 0.11360686688440211,
      "learning_rate": 3.6176134436808074e-06,
      "loss": 0.0372,
      "num_tokens": 118990465.0,
      "step": 1561
    },
    {
      "epoch": 1.9482220835932627,
      "grad_norm": 0.11155062075729752,
      "learning_rate": 3.612111139123239e-06,
      "loss": 0.0332,
      "num_tokens": 119066805.0,
      "step": 1562
    },
    {
      "epoch": 1.9494697442295696,
      "grad_norm": 0.1127360747157812,
      "learning_rate": 3.6066122578788033e-06,
      "loss": 0.0354,
      "num_tokens": 119143624.0,
      "step": 1563
    },
    {
      "epoch": 1.9507174048658764,
      "grad_norm": 0.11141572186918379,
      "learning_rate": 3.6011168099186322e-06,
      "loss": 0.0361,
      "num_tokens": 119219639.0,
      "step": 1564
    },
    {
      "epoch": 1.9519650655021834,
      "grad_norm": 0.10663689560547111,
      "learning_rate": 3.5956248052076383e-06,
      "loss": 0.0346,
      "num_tokens": 119296081.0,
      "step": 1565
    },
    {
      "epoch": 1.9532127261384904,
      "grad_norm": 0.11391871771412312,
      "learning_rate": 3.5901362537044826e-06,
      "loss": 0.0349,
      "num_tokens": 119371763.0,
      "step": 1566
    },
    {
      "epoch": 1.9544603867747972,
      "grad_norm": 0.11909019651801186,
      "learning_rate": 3.584651165361568e-06,
      "loss": 0.0374,
      "num_tokens": 119448704.0,
      "step": 1567
    },
    {
      "epoch": 1.9557080474111042,
      "grad_norm": 0.10989187981100837,
      "learning_rate": 3.579169550125019e-06,
      "loss": 0.0373,
      "num_tokens": 119526959.0,
      "step": 1568
    },
    {
      "epoch": 1.9569557080474111,
      "grad_norm": 0.12551303966185134,
      "learning_rate": 3.5736914179346626e-06,
      "loss": 0.0377,
      "num_tokens": 119603721.0,
      "step": 1569
    },
    {
      "epoch": 1.958203368683718,
      "grad_norm": 0.11488802357223445,
      "learning_rate": 3.5682167787240053e-06,
      "loss": 0.0397,
      "num_tokens": 119682215.0,
      "step": 1570
    },
    {
      "epoch": 1.959451029320025,
      "grad_norm": 0.13645450743495965,
      "learning_rate": 3.5627456424202223e-06,
      "loss": 0.0348,
      "num_tokens": 119758988.0,
      "step": 1571
    },
    {
      "epoch": 1.960698689956332,
      "grad_norm": 0.11011490733351555,
      "learning_rate": 3.55727801894414e-06,
      "loss": 0.0338,
      "num_tokens": 119835006.0,
      "step": 1572
    },
    {
      "epoch": 1.9619463505926387,
      "grad_norm": 0.10610282995025075,
      "learning_rate": 3.5518139182102106e-06,
      "loss": 0.0351,
      "num_tokens": 119910802.0,
      "step": 1573
    },
    {
      "epoch": 1.9631940112289459,
      "grad_norm": 0.11487742308324417,
      "learning_rate": 3.5463533501265e-06,
      "loss": 0.0352,
      "num_tokens": 119986133.0,
      "step": 1574
    },
    {
      "epoch": 1.9644416718652526,
      "grad_norm": 0.11242988977931746,
      "learning_rate": 3.5408963245946714e-06,
      "loss": 0.0367,
      "num_tokens": 120062879.0,
      "step": 1575
    },
    {
      "epoch": 1.9656893325015594,
      "grad_norm": 0.12786058469254943,
      "learning_rate": 3.53544285150996e-06,
      "loss": 0.0342,
      "num_tokens": 120139678.0,
      "step": 1576
    },
    {
      "epoch": 1.9669369931378666,
      "grad_norm": 0.10314080876800046,
      "learning_rate": 3.529992940761159e-06,
      "loss": 0.0339,
      "num_tokens": 120215749.0,
      "step": 1577
    },
    {
      "epoch": 1.9681846537741734,
      "grad_norm": 0.10840591571518475,
      "learning_rate": 3.524546602230606e-06,
      "loss": 0.0368,
      "num_tokens": 120291566.0,
      "step": 1578
    },
    {
      "epoch": 1.9694323144104804,
      "grad_norm": 0.12653971547288437,
      "learning_rate": 3.5191038457941596e-06,
      "loss": 0.0354,
      "num_tokens": 120368332.0,
      "step": 1579
    },
    {
      "epoch": 1.9706799750467874,
      "grad_norm": 0.10154295626268736,
      "learning_rate": 3.5136646813211784e-06,
      "loss": 0.0326,
      "num_tokens": 120443551.0,
      "step": 1580
    },
    {
      "epoch": 1.9719276356830941,
      "grad_norm": 0.10391823409071556,
      "learning_rate": 3.5082291186745145e-06,
      "loss": 0.0359,
      "num_tokens": 120520401.0,
      "step": 1581
    },
    {
      "epoch": 1.9731752963194011,
      "grad_norm": 0.12143861281656622,
      "learning_rate": 3.5027971677104867e-06,
      "loss": 0.0364,
      "num_tokens": 120596577.0,
      "step": 1582
    },
    {
      "epoch": 1.9744229569557081,
      "grad_norm": 0.11710635362818253,
      "learning_rate": 3.497368838278862e-06,
      "loss": 0.0365,
      "num_tokens": 120672610.0,
      "step": 1583
    },
    {
      "epoch": 1.975670617592015,
      "grad_norm": 0.10505989387647319,
      "learning_rate": 3.491944140222845e-06,
      "loss": 0.0341,
      "num_tokens": 120748072.0,
      "step": 1584
    },
    {
      "epoch": 1.976918278228322,
      "grad_norm": 0.11325553204443413,
      "learning_rate": 3.486523083379051e-06,
      "loss": 0.035,
      "num_tokens": 120824607.0,
      "step": 1585
    },
    {
      "epoch": 1.9781659388646289,
      "grad_norm": 0.10616594120542275,
      "learning_rate": 3.481105677577493e-06,
      "loss": 0.0337,
      "num_tokens": 120899699.0,
      "step": 1586
    },
    {
      "epoch": 1.9794135995009356,
      "grad_norm": 0.11658220768049024,
      "learning_rate": 3.475691932641569e-06,
      "loss": 0.0356,
      "num_tokens": 120976512.0,
      "step": 1587
    },
    {
      "epoch": 1.9806612601372426,
      "grad_norm": 0.11335818137173843,
      "learning_rate": 3.4702818583880305e-06,
      "loss": 0.0331,
      "num_tokens": 121051822.0,
      "step": 1588
    },
    {
      "epoch": 1.9819089207735496,
      "grad_norm": 0.10858050015524784,
      "learning_rate": 3.46487546462698e-06,
      "loss": 0.0346,
      "num_tokens": 121127103.0,
      "step": 1589
    },
    {
      "epoch": 1.9831565814098564,
      "grad_norm": 0.12886983678658803,
      "learning_rate": 3.4594727611618462e-06,
      "loss": 0.0411,
      "num_tokens": 121204889.0,
      "step": 1590
    },
    {
      "epoch": 1.9844042420461634,
      "grad_norm": 0.1124176219267256,
      "learning_rate": 3.454073757789359e-06,
      "loss": 0.0365,
      "num_tokens": 121280596.0,
      "step": 1591
    },
    {
      "epoch": 1.9856519026824704,
      "grad_norm": 0.11711254550831095,
      "learning_rate": 3.4486784642995442e-06,
      "loss": 0.0355,
      "num_tokens": 121358235.0,
      "step": 1592
    },
    {
      "epoch": 1.9868995633187772,
      "grad_norm": 0.12398515386115859,
      "learning_rate": 3.4432868904757024e-06,
      "loss": 0.035,
      "num_tokens": 121434785.0,
      "step": 1593
    },
    {
      "epoch": 1.9881472239550844,
      "grad_norm": 0.11418604832158563,
      "learning_rate": 3.437899046094384e-06,
      "loss": 0.0348,
      "num_tokens": 121511577.0,
      "step": 1594
    },
    {
      "epoch": 1.9893948845913911,
      "grad_norm": 0.112301925129123,
      "learning_rate": 3.432514940925378e-06,
      "loss": 0.036,
      "num_tokens": 121588869.0,
      "step": 1595
    },
    {
      "epoch": 1.990642545227698,
      "grad_norm": 0.10984815919551645,
      "learning_rate": 3.4271345847316974e-06,
      "loss": 0.0378,
      "num_tokens": 121667192.0,
      "step": 1596
    },
    {
      "epoch": 1.9918902058640051,
      "grad_norm": 0.1126552804474371,
      "learning_rate": 3.421757987269554e-06,
      "loss": 0.0369,
      "num_tokens": 121743520.0,
      "step": 1597
    },
    {
      "epoch": 1.9931378665003119,
      "grad_norm": 0.11902040360206023,
      "learning_rate": 3.416385158288343e-06,
      "loss": 0.0358,
      "num_tokens": 121820019.0,
      "step": 1598
    },
    {
      "epoch": 1.9943855271366189,
      "grad_norm": 0.11822457506008535,
      "learning_rate": 3.411016107530628e-06,
      "loss": 0.0338,
      "num_tokens": 121896092.0,
      "step": 1599
    },
    {
      "epoch": 1.9956331877729259,
      "grad_norm": 0.11654657768961584,
      "learning_rate": 3.405650844732122e-06,
      "loss": 0.036,
      "num_tokens": 121973290.0,
      "step": 1600
    },
    {
      "epoch": 1.9968808484092326,
      "grad_norm": 0.1073126712837225,
      "learning_rate": 3.400289379621664e-06,
      "loss": 0.0353,
      "num_tokens": 122048846.0,
      "step": 1601
    },
    {
      "epoch": 1.9981285090455396,
      "grad_norm": 0.11588726397968585,
      "learning_rate": 3.394931721921214e-06,
      "loss": 0.036,
      "num_tokens": 122124025.0,
      "step": 1602
    },
    {
      "epoch": 1.9993761696818466,
      "grad_norm": 0.11056330436185885,
      "learning_rate": 3.3895778813458256e-06,
      "loss": 0.0354,
      "num_tokens": 122200456.0,
      "step": 1603
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.16899944282914875,
      "learning_rate": 3.3842278676036293e-06,
      "loss": 0.0342,
      "num_tokens": 122238397.0,
      "step": 1604
    },
    {
      "epoch": 2.0012476606363068,
      "grad_norm": 0.09094457519669298,
      "learning_rate": 3.3788816903958145e-06,
      "loss": 0.0293,
      "num_tokens": 122313704.0,
      "step": 1605
    },
    {
      "epoch": 2.002495321272614,
      "grad_norm": 0.10470187666679959,
      "learning_rate": 3.37353935941662e-06,
      "loss": 0.031,
      "num_tokens": 122389665.0,
      "step": 1606
    },
    {
      "epoch": 2.0037429819089208,
      "grad_norm": 0.10261607757231991,
      "learning_rate": 3.3682008843533055e-06,
      "loss": 0.0306,
      "num_tokens": 122465436.0,
      "step": 1607
    },
    {
      "epoch": 2.0049906425452275,
      "grad_norm": 0.10768005430067183,
      "learning_rate": 3.3628662748861374e-06,
      "loss": 0.0301,
      "num_tokens": 122542091.0,
      "step": 1608
    },
    {
      "epoch": 2.0062383031815347,
      "grad_norm": 0.11091901445964379,
      "learning_rate": 3.357535540688379e-06,
      "loss": 0.0294,
      "num_tokens": 122617623.0,
      "step": 1609
    },
    {
      "epoch": 2.0074859638178415,
      "grad_norm": 0.10028034432839329,
      "learning_rate": 3.3522086914262585e-06,
      "loss": 0.031,
      "num_tokens": 122694522.0,
      "step": 1610
    },
    {
      "epoch": 2.0087336244541483,
      "grad_norm": 0.10536087513490514,
      "learning_rate": 3.3468857367589665e-06,
      "loss": 0.0314,
      "num_tokens": 122771954.0,
      "step": 1611
    },
    {
      "epoch": 2.0099812850904555,
      "grad_norm": 0.10430826658048525,
      "learning_rate": 3.3415666863386298e-06,
      "loss": 0.0297,
      "num_tokens": 122848328.0,
      "step": 1612
    },
    {
      "epoch": 2.0112289457267623,
      "grad_norm": 0.10453848093823005,
      "learning_rate": 3.3362515498102934e-06,
      "loss": 0.0303,
      "num_tokens": 122924877.0,
      "step": 1613
    },
    {
      "epoch": 2.0124766063630695,
      "grad_norm": 0.10028839319505682,
      "learning_rate": 3.330940336811903e-06,
      "loss": 0.029,
      "num_tokens": 123000873.0,
      "step": 1614
    },
    {
      "epoch": 2.0137242669993762,
      "grad_norm": 0.11743564991525901,
      "learning_rate": 3.325633056974298e-06,
      "loss": 0.0295,
      "num_tokens": 123076071.0,
      "step": 1615
    },
    {
      "epoch": 2.014971927635683,
      "grad_norm": 0.11146806831799687,
      "learning_rate": 3.3203297199211794e-06,
      "loss": 0.0289,
      "num_tokens": 123151982.0,
      "step": 1616
    },
    {
      "epoch": 2.01621958827199,
      "grad_norm": 0.12382288674623765,
      "learning_rate": 3.315030335269096e-06,
      "loss": 0.0293,
      "num_tokens": 123228726.0,
      "step": 1617
    },
    {
      "epoch": 2.017467248908297,
      "grad_norm": 0.11950129176714563,
      "learning_rate": 3.309734912627441e-06,
      "loss": 0.0277,
      "num_tokens": 123304605.0,
      "step": 1618
    },
    {
      "epoch": 2.0187149095446038,
      "grad_norm": 0.13106824723008773,
      "learning_rate": 3.304443461598413e-06,
      "loss": 0.0295,
      "num_tokens": 123382565.0,
      "step": 1619
    },
    {
      "epoch": 2.019962570180911,
      "grad_norm": 0.1441982452291627,
      "learning_rate": 3.299155991777011e-06,
      "loss": 0.0307,
      "num_tokens": 123457876.0,
      "step": 1620
    },
    {
      "epoch": 2.0212102308172177,
      "grad_norm": 0.1325658173846908,
      "learning_rate": 3.2938725127510185e-06,
      "loss": 0.0292,
      "num_tokens": 123533788.0,
      "step": 1621
    },
    {
      "epoch": 2.0224578914535245,
      "grad_norm": 0.1297831271928566,
      "learning_rate": 3.2885930341009774e-06,
      "loss": 0.0293,
      "num_tokens": 123609961.0,
      "step": 1622
    },
    {
      "epoch": 2.0237055520898317,
      "grad_norm": 0.121658116064928,
      "learning_rate": 3.2833175654001787e-06,
      "loss": 0.0283,
      "num_tokens": 123687877.0,
      "step": 1623
    },
    {
      "epoch": 2.0249532127261385,
      "grad_norm": 0.1180258767424536,
      "learning_rate": 3.278046116214642e-06,
      "loss": 0.0283,
      "num_tokens": 123764590.0,
      "step": 1624
    },
    {
      "epoch": 2.0262008733624453,
      "grad_norm": 0.11629528263227036,
      "learning_rate": 3.272778696103099e-06,
      "loss": 0.0282,
      "num_tokens": 123841467.0,
      "step": 1625
    },
    {
      "epoch": 2.0274485339987525,
      "grad_norm": 0.11430394671374194,
      "learning_rate": 3.2675153146169736e-06,
      "loss": 0.0284,
      "num_tokens": 123917776.0,
      "step": 1626
    },
    {
      "epoch": 2.0286961946350592,
      "grad_norm": 0.11713709735142017,
      "learning_rate": 3.2622559813003684e-06,
      "loss": 0.0286,
      "num_tokens": 123993601.0,
      "step": 1627
    },
    {
      "epoch": 2.029943855271366,
      "grad_norm": 0.11804813522003368,
      "learning_rate": 3.2570007056900437e-06,
      "loss": 0.0287,
      "num_tokens": 124069540.0,
      "step": 1628
    },
    {
      "epoch": 2.031191515907673,
      "grad_norm": 0.13641435448063324,
      "learning_rate": 3.2517494973154008e-06,
      "loss": 0.0311,
      "num_tokens": 124146287.0,
      "step": 1629
    },
    {
      "epoch": 2.03243917654398,
      "grad_norm": 0.11999347396308309,
      "learning_rate": 3.2465023656984707e-06,
      "loss": 0.0284,
      "num_tokens": 124222142.0,
      "step": 1630
    },
    {
      "epoch": 2.0336868371802868,
      "grad_norm": 0.11183330241969212,
      "learning_rate": 3.2412593203538857e-06,
      "loss": 0.0293,
      "num_tokens": 124298900.0,
      "step": 1631
    },
    {
      "epoch": 2.034934497816594,
      "grad_norm": 0.11671051353373899,
      "learning_rate": 3.236020370788876e-06,
      "loss": 0.0302,
      "num_tokens": 124374841.0,
      "step": 1632
    },
    {
      "epoch": 2.0361821584529007,
      "grad_norm": 0.11468680865831452,
      "learning_rate": 3.230785526503236e-06,
      "loss": 0.0288,
      "num_tokens": 124451278.0,
      "step": 1633
    },
    {
      "epoch": 2.037429819089208,
      "grad_norm": 0.12259831247354039,
      "learning_rate": 3.225554796989325e-06,
      "loss": 0.0307,
      "num_tokens": 124528188.0,
      "step": 1634
    },
    {
      "epoch": 2.0386774797255147,
      "grad_norm": 0.126955831725854,
      "learning_rate": 3.2203281917320328e-06,
      "loss": 0.0309,
      "num_tokens": 124605124.0,
      "step": 1635
    },
    {
      "epoch": 2.0399251403618215,
      "grad_norm": 0.1266009278644561,
      "learning_rate": 3.2151057202087783e-06,
      "loss": 0.0297,
      "num_tokens": 124682843.0,
      "step": 1636
    },
    {
      "epoch": 2.0411728009981287,
      "grad_norm": 0.11629716217576579,
      "learning_rate": 3.209887391889479e-06,
      "loss": 0.0289,
      "num_tokens": 124761248.0,
      "step": 1637
    },
    {
      "epoch": 2.0424204616344355,
      "grad_norm": 0.11719507314291479,
      "learning_rate": 3.204673216236539e-06,
      "loss": 0.03,
      "num_tokens": 124836567.0,
      "step": 1638
    },
    {
      "epoch": 2.0436681222707422,
      "grad_norm": 0.13293749631217391,
      "learning_rate": 3.199463202704838e-06,
      "loss": 0.0301,
      "num_tokens": 124914356.0,
      "step": 1639
    },
    {
      "epoch": 2.0449157829070495,
      "grad_norm": 0.13159382652993765,
      "learning_rate": 3.194257360741706e-06,
      "loss": 0.0302,
      "num_tokens": 124992108.0,
      "step": 1640
    },
    {
      "epoch": 2.046163443543356,
      "grad_norm": 0.10907795483596473,
      "learning_rate": 3.189055699786906e-06,
      "loss": 0.0287,
      "num_tokens": 125068440.0,
      "step": 1641
    },
    {
      "epoch": 2.047411104179663,
      "grad_norm": 0.1323478428940742,
      "learning_rate": 3.1838582292726206e-06,
      "loss": 0.031,
      "num_tokens": 125145344.0,
      "step": 1642
    },
    {
      "epoch": 2.04865876481597,
      "grad_norm": 0.11288069563704346,
      "learning_rate": 3.1786649586234373e-06,
      "loss": 0.0282,
      "num_tokens": 125220797.0,
      "step": 1643
    },
    {
      "epoch": 2.049906425452277,
      "grad_norm": 0.1305477821978057,
      "learning_rate": 3.173475897256325e-06,
      "loss": 0.0297,
      "num_tokens": 125296316.0,
      "step": 1644
    },
    {
      "epoch": 2.0511540860885837,
      "grad_norm": 0.11482519476065899,
      "learning_rate": 3.1682910545806167e-06,
      "loss": 0.0286,
      "num_tokens": 125373059.0,
      "step": 1645
    },
    {
      "epoch": 2.052401746724891,
      "grad_norm": 0.11295210479276795,
      "learning_rate": 3.1631104399980053e-06,
      "loss": 0.0281,
      "num_tokens": 125449203.0,
      "step": 1646
    },
    {
      "epoch": 2.0536494073611977,
      "grad_norm": 0.12032389651169483,
      "learning_rate": 3.157934062902508e-06,
      "loss": 0.0298,
      "num_tokens": 125525599.0,
      "step": 1647
    },
    {
      "epoch": 2.0548970679975045,
      "grad_norm": 0.11902065154734578,
      "learning_rate": 3.1527619326804594e-06,
      "loss": 0.0286,
      "num_tokens": 125601496.0,
      "step": 1648
    },
    {
      "epoch": 2.0561447286338117,
      "grad_norm": 0.11410725158268568,
      "learning_rate": 3.147594058710498e-06,
      "loss": 0.0286,
      "num_tokens": 125677905.0,
      "step": 1649
    },
    {
      "epoch": 2.0573923892701185,
      "grad_norm": 0.11816837329563465,
      "learning_rate": 3.14243045036354e-06,
      "loss": 0.0299,
      "num_tokens": 125756319.0,
      "step": 1650
    },
    {
      "epoch": 2.0586400499064252,
      "grad_norm": 0.11995125205915291,
      "learning_rate": 3.1372711170027666e-06,
      "loss": 0.0285,
      "num_tokens": 125832029.0,
      "step": 1651
    },
    {
      "epoch": 2.0598877105427325,
      "grad_norm": 0.12010979336540205,
      "learning_rate": 3.13211606798361e-06,
      "loss": 0.0294,
      "num_tokens": 125908359.0,
      "step": 1652
    },
    {
      "epoch": 2.061135371179039,
      "grad_norm": 0.11991952735264713,
      "learning_rate": 3.1269653126537344e-06,
      "loss": 0.029,
      "num_tokens": 125985504.0,
      "step": 1653
    },
    {
      "epoch": 2.0623830318153464,
      "grad_norm": 0.12245491314302823,
      "learning_rate": 3.121818860353011e-06,
      "loss": 0.0285,
      "num_tokens": 126061654.0,
      "step": 1654
    },
    {
      "epoch": 2.063630692451653,
      "grad_norm": 0.11426788535740982,
      "learning_rate": 3.116676720413519e-06,
      "loss": 0.0285,
      "num_tokens": 126138177.0,
      "step": 1655
    },
    {
      "epoch": 2.06487835308796,
      "grad_norm": 0.12048808823445821,
      "learning_rate": 3.11153890215951e-06,
      "loss": 0.0294,
      "num_tokens": 126214439.0,
      "step": 1656
    },
    {
      "epoch": 2.066126013724267,
      "grad_norm": 0.11638863421566648,
      "learning_rate": 3.1064054149073984e-06,
      "loss": 0.0289,
      "num_tokens": 126289984.0,
      "step": 1657
    },
    {
      "epoch": 2.067373674360574,
      "grad_norm": 0.13366366869107843,
      "learning_rate": 3.1012762679657525e-06,
      "loss": 0.0297,
      "num_tokens": 126365459.0,
      "step": 1658
    },
    {
      "epoch": 2.0686213349968807,
      "grad_norm": 0.11864627470251912,
      "learning_rate": 3.0961514706352654e-06,
      "loss": 0.0297,
      "num_tokens": 126442037.0,
      "step": 1659
    },
    {
      "epoch": 2.069868995633188,
      "grad_norm": 0.1204172393930324,
      "learning_rate": 3.09103103220874e-06,
      "loss": 0.0285,
      "num_tokens": 126517872.0,
      "step": 1660
    },
    {
      "epoch": 2.0711166562694947,
      "grad_norm": 0.11684206507561638,
      "learning_rate": 3.085914961971082e-06,
      "loss": 0.0285,
      "num_tokens": 126593424.0,
      "step": 1661
    },
    {
      "epoch": 2.0723643169058015,
      "grad_norm": 0.11878319279046977,
      "learning_rate": 3.080803269199275e-06,
      "loss": 0.0291,
      "num_tokens": 126668395.0,
      "step": 1662
    },
    {
      "epoch": 2.0736119775421087,
      "grad_norm": 0.13003391776490136,
      "learning_rate": 3.0756959631623583e-06,
      "loss": 0.0295,
      "num_tokens": 126744950.0,
      "step": 1663
    },
    {
      "epoch": 2.0748596381784155,
      "grad_norm": 0.11180545254543592,
      "learning_rate": 3.0705930531214255e-06,
      "loss": 0.0298,
      "num_tokens": 126820804.0,
      "step": 1664
    },
    {
      "epoch": 2.0761072988147222,
      "grad_norm": 0.12429937470575604,
      "learning_rate": 3.065494548329594e-06,
      "loss": 0.0293,
      "num_tokens": 126898763.0,
      "step": 1665
    },
    {
      "epoch": 2.0773549594510294,
      "grad_norm": 0.11939233163914764,
      "learning_rate": 3.060400458031991e-06,
      "loss": 0.0299,
      "num_tokens": 126975067.0,
      "step": 1666
    },
    {
      "epoch": 2.078602620087336,
      "grad_norm": 0.11859731452377882,
      "learning_rate": 3.055310791465744e-06,
      "loss": 0.0293,
      "num_tokens": 127051917.0,
      "step": 1667
    },
    {
      "epoch": 2.079850280723643,
      "grad_norm": 0.10596465405559069,
      "learning_rate": 3.0502255578599594e-06,
      "loss": 0.0286,
      "num_tokens": 127126581.0,
      "step": 1668
    },
    {
      "epoch": 2.08109794135995,
      "grad_norm": 0.12147255599548454,
      "learning_rate": 3.0451447664357005e-06,
      "loss": 0.0305,
      "num_tokens": 127203055.0,
      "step": 1669
    },
    {
      "epoch": 2.082345601996257,
      "grad_norm": 0.11456004309224181,
      "learning_rate": 3.040068426405976e-06,
      "loss": 0.0289,
      "num_tokens": 127279082.0,
      "step": 1670
    },
    {
      "epoch": 2.083593262632564,
      "grad_norm": 0.1099209222662035,
      "learning_rate": 3.0349965469757283e-06,
      "loss": 0.0288,
      "num_tokens": 127354973.0,
      "step": 1671
    },
    {
      "epoch": 2.084840923268871,
      "grad_norm": 0.1092793466680077,
      "learning_rate": 3.0299291373418038e-06,
      "loss": 0.0294,
      "num_tokens": 127430677.0,
      "step": 1672
    },
    {
      "epoch": 2.0860885839051777,
      "grad_norm": 0.1140484049265834,
      "learning_rate": 3.024866206692953e-06,
      "loss": 0.029,
      "num_tokens": 127508449.0,
      "step": 1673
    },
    {
      "epoch": 2.087336244541485,
      "grad_norm": 0.11781551488387931,
      "learning_rate": 3.0198077642097945e-06,
      "loss": 0.0299,
      "num_tokens": 127584320.0,
      "step": 1674
    },
    {
      "epoch": 2.0885839051777917,
      "grad_norm": 0.11904168811025281,
      "learning_rate": 3.014753819064817e-06,
      "loss": 0.0293,
      "num_tokens": 127661488.0,
      "step": 1675
    },
    {
      "epoch": 2.0898315658140985,
      "grad_norm": 0.11711225032175877,
      "learning_rate": 3.009704380422348e-06,
      "loss": 0.0291,
      "num_tokens": 127737991.0,
      "step": 1676
    },
    {
      "epoch": 2.0910792264504057,
      "grad_norm": 0.10612396078457137,
      "learning_rate": 3.004659457438548e-06,
      "loss": 0.0282,
      "num_tokens": 127814066.0,
      "step": 1677
    },
    {
      "epoch": 2.0923268870867124,
      "grad_norm": 0.11501420629637818,
      "learning_rate": 2.999619059261387e-06,
      "loss": 0.0294,
      "num_tokens": 127890907.0,
      "step": 1678
    },
    {
      "epoch": 2.093574547723019,
      "grad_norm": 0.11619371324958405,
      "learning_rate": 2.9945831950306285e-06,
      "loss": 0.029,
      "num_tokens": 127966833.0,
      "step": 1679
    },
    {
      "epoch": 2.0948222083593264,
      "grad_norm": 0.1344217605260488,
      "learning_rate": 2.9895518738778196e-06,
      "loss": 0.0303,
      "num_tokens": 128043513.0,
      "step": 1680
    },
    {
      "epoch": 2.096069868995633,
      "grad_norm": 0.11509467874428474,
      "learning_rate": 2.984525104926262e-06,
      "loss": 0.0287,
      "num_tokens": 128120826.0,
      "step": 1681
    },
    {
      "epoch": 2.09731752963194,
      "grad_norm": 0.13048016716661157,
      "learning_rate": 2.97950289729101e-06,
      "loss": 0.0285,
      "num_tokens": 128197592.0,
      "step": 1682
    },
    {
      "epoch": 2.098565190268247,
      "grad_norm": 0.11403952953370168,
      "learning_rate": 2.974485260078846e-06,
      "loss": 0.0284,
      "num_tokens": 128273160.0,
      "step": 1683
    },
    {
      "epoch": 2.099812850904554,
      "grad_norm": 0.11301234594867011,
      "learning_rate": 2.9694722023882607e-06,
      "loss": 0.0288,
      "num_tokens": 128349131.0,
      "step": 1684
    },
    {
      "epoch": 2.1010605115408607,
      "grad_norm": 0.11308339383694495,
      "learning_rate": 2.9644637333094404e-06,
      "loss": 0.0286,
      "num_tokens": 128424743.0,
      "step": 1685
    },
    {
      "epoch": 2.102308172177168,
      "grad_norm": 0.13222972121674195,
      "learning_rate": 2.959459861924258e-06,
      "loss": 0.0321,
      "num_tokens": 128500534.0,
      "step": 1686
    },
    {
      "epoch": 2.1035558328134747,
      "grad_norm": 0.11118702677552449,
      "learning_rate": 2.954460597306242e-06,
      "loss": 0.028,
      "num_tokens": 128576822.0,
      "step": 1687
    },
    {
      "epoch": 2.1048034934497815,
      "grad_norm": 0.12542704234184968,
      "learning_rate": 2.9494659485205683e-06,
      "loss": 0.0295,
      "num_tokens": 128653520.0,
      "step": 1688
    },
    {
      "epoch": 2.1060511540860887,
      "grad_norm": 0.11687630320165059,
      "learning_rate": 2.9444759246240505e-06,
      "loss": 0.0291,
      "num_tokens": 128730266.0,
      "step": 1689
    },
    {
      "epoch": 2.1072988147223954,
      "grad_norm": 0.11704520647268402,
      "learning_rate": 2.939490534665107e-06,
      "loss": 0.0303,
      "num_tokens": 128806281.0,
      "step": 1690
    },
    {
      "epoch": 2.108546475358702,
      "grad_norm": 0.12007099225710577,
      "learning_rate": 2.934509787683755e-06,
      "loss": 0.0296,
      "num_tokens": 128882197.0,
      "step": 1691
    },
    {
      "epoch": 2.1097941359950094,
      "grad_norm": 0.12894656412831512,
      "learning_rate": 2.929533692711598e-06,
      "loss": 0.0288,
      "num_tokens": 128958255.0,
      "step": 1692
    },
    {
      "epoch": 2.111041796631316,
      "grad_norm": 0.13635018499385967,
      "learning_rate": 2.9245622587717982e-06,
      "loss": 0.0317,
      "num_tokens": 129034456.0,
      "step": 1693
    },
    {
      "epoch": 2.1122894572676234,
      "grad_norm": 0.1188141113279919,
      "learning_rate": 2.919595494879065e-06,
      "loss": 0.0288,
      "num_tokens": 129111270.0,
      "step": 1694
    },
    {
      "epoch": 2.11353711790393,
      "grad_norm": 0.1218554043702889,
      "learning_rate": 2.9146334100396474e-06,
      "loss": 0.0281,
      "num_tokens": 129187037.0,
      "step": 1695
    },
    {
      "epoch": 2.114784778540237,
      "grad_norm": 0.11375072712332324,
      "learning_rate": 2.9096760132513036e-06,
      "loss": 0.0292,
      "num_tokens": 129263591.0,
      "step": 1696
    },
    {
      "epoch": 2.116032439176544,
      "grad_norm": 0.1226238587716704,
      "learning_rate": 2.9047233135032927e-06,
      "loss": 0.0286,
      "num_tokens": 129339359.0,
      "step": 1697
    },
    {
      "epoch": 2.117280099812851,
      "grad_norm": 0.1144029375418943,
      "learning_rate": 2.8997753197763532e-06,
      "loss": 0.0303,
      "num_tokens": 129415566.0,
      "step": 1698
    },
    {
      "epoch": 2.1185277604491577,
      "grad_norm": 0.12621177470688724,
      "learning_rate": 2.894832041042699e-06,
      "loss": 0.0302,
      "num_tokens": 129492927.0,
      "step": 1699
    },
    {
      "epoch": 2.119775421085465,
      "grad_norm": 0.13278516666491638,
      "learning_rate": 2.8898934862659823e-06,
      "loss": 0.0463,
      "num_tokens": 129570623.0,
      "step": 1700
    },
    {
      "epoch": 2.1210230817217717,
      "grad_norm": 0.12206158358190072,
      "learning_rate": 2.8849596644013e-06,
      "loss": 0.0303,
      "num_tokens": 129646983.0,
      "step": 1701
    },
    {
      "epoch": 2.1222707423580784,
      "grad_norm": 0.15346776037825924,
      "learning_rate": 2.880030584395162e-06,
      "loss": 0.029,
      "num_tokens": 129723058.0,
      "step": 1702
    },
    {
      "epoch": 2.1235184029943857,
      "grad_norm": 0.13346485239158706,
      "learning_rate": 2.8751062551854775e-06,
      "loss": 0.0292,
      "num_tokens": 129799644.0,
      "step": 1703
    },
    {
      "epoch": 2.1247660636306924,
      "grad_norm": 0.12080827755014605,
      "learning_rate": 2.870186685701545e-06,
      "loss": 0.0296,
      "num_tokens": 129875807.0,
      "step": 1704
    },
    {
      "epoch": 2.126013724266999,
      "grad_norm": 0.11778339398723604,
      "learning_rate": 2.8652718848640337e-06,
      "loss": 0.0281,
      "num_tokens": 129952062.0,
      "step": 1705
    },
    {
      "epoch": 2.1272613849033064,
      "grad_norm": 0.12815407188577152,
      "learning_rate": 2.8603618615849603e-06,
      "loss": 0.0292,
      "num_tokens": 130029889.0,
      "step": 1706
    },
    {
      "epoch": 2.128509045539613,
      "grad_norm": 0.12170030398585802,
      "learning_rate": 2.8554566247676806e-06,
      "loss": 0.0304,
      "num_tokens": 130106278.0,
      "step": 1707
    },
    {
      "epoch": 2.12975670617592,
      "grad_norm": 0.13564222415311583,
      "learning_rate": 2.850556183306874e-06,
      "loss": 0.0298,
      "num_tokens": 130182742.0,
      "step": 1708
    },
    {
      "epoch": 2.131004366812227,
      "grad_norm": 0.12092186918308077,
      "learning_rate": 2.845660546088519e-06,
      "loss": 0.0292,
      "num_tokens": 130258018.0,
      "step": 1709
    },
    {
      "epoch": 2.132252027448534,
      "grad_norm": 0.12388745329777386,
      "learning_rate": 2.8407697219898865e-06,
      "loss": 0.0292,
      "num_tokens": 130334446.0,
      "step": 1710
    },
    {
      "epoch": 2.133499688084841,
      "grad_norm": 0.12187584450348393,
      "learning_rate": 2.8358837198795223e-06,
      "loss": 0.0292,
      "num_tokens": 130409895.0,
      "step": 1711
    },
    {
      "epoch": 2.134747348721148,
      "grad_norm": 0.11869077268783002,
      "learning_rate": 2.8310025486172223e-06,
      "loss": 0.0296,
      "num_tokens": 130486706.0,
      "step": 1712
    },
    {
      "epoch": 2.1359950093574547,
      "grad_norm": 0.10293036941377427,
      "learning_rate": 2.8261262170540242e-06,
      "loss": 0.028,
      "num_tokens": 130562325.0,
      "step": 1713
    },
    {
      "epoch": 2.137242669993762,
      "grad_norm": 0.1128646409689079,
      "learning_rate": 2.821254734032194e-06,
      "loss": 0.0292,
      "num_tokens": 130638649.0,
      "step": 1714
    },
    {
      "epoch": 2.1384903306300687,
      "grad_norm": 0.1206922280211655,
      "learning_rate": 2.8163881083852e-06,
      "loss": 0.0284,
      "num_tokens": 130715765.0,
      "step": 1715
    },
    {
      "epoch": 2.1397379912663754,
      "grad_norm": 0.1110611156235701,
      "learning_rate": 2.811526348937706e-06,
      "loss": 0.0293,
      "num_tokens": 130791789.0,
      "step": 1716
    },
    {
      "epoch": 2.1409856519026826,
      "grad_norm": 0.12851385903558105,
      "learning_rate": 2.806669464505552e-06,
      "loss": 0.0291,
      "num_tokens": 130868440.0,
      "step": 1717
    },
    {
      "epoch": 2.1422333125389894,
      "grad_norm": 0.10535176804674477,
      "learning_rate": 2.80181746389574e-06,
      "loss": 0.028,
      "num_tokens": 130945015.0,
      "step": 1718
    },
    {
      "epoch": 2.143480973175296,
      "grad_norm": 0.11822191143631985,
      "learning_rate": 2.7969703559064076e-06,
      "loss": 0.0289,
      "num_tokens": 131022857.0,
      "step": 1719
    },
    {
      "epoch": 2.1447286338116034,
      "grad_norm": 0.12991511557728588,
      "learning_rate": 2.792128149326833e-06,
      "loss": 0.0305,
      "num_tokens": 131099510.0,
      "step": 1720
    },
    {
      "epoch": 2.14597629444791,
      "grad_norm": 0.12718239799706485,
      "learning_rate": 2.7872908529373976e-06,
      "loss": 0.0301,
      "num_tokens": 131176096.0,
      "step": 1721
    },
    {
      "epoch": 2.147223955084217,
      "grad_norm": 0.11554797810203757,
      "learning_rate": 2.782458475509581e-06,
      "loss": 0.0302,
      "num_tokens": 131252380.0,
      "step": 1722
    },
    {
      "epoch": 2.148471615720524,
      "grad_norm": 0.12320833041672591,
      "learning_rate": 2.7776310258059447e-06,
      "loss": 0.03,
      "num_tokens": 131328378.0,
      "step": 1723
    },
    {
      "epoch": 2.149719276356831,
      "grad_norm": 0.11326383473860072,
      "learning_rate": 2.772808512580114e-06,
      "loss": 0.0288,
      "num_tokens": 131405405.0,
      "step": 1724
    },
    {
      "epoch": 2.1509669369931377,
      "grad_norm": 0.11951009210803026,
      "learning_rate": 2.767990944576763e-06,
      "loss": 0.0286,
      "num_tokens": 131481573.0,
      "step": 1725
    },
    {
      "epoch": 2.152214597629445,
      "grad_norm": 0.11801895008095241,
      "learning_rate": 2.7631783305316017e-06,
      "loss": 0.029,
      "num_tokens": 131557531.0,
      "step": 1726
    },
    {
      "epoch": 2.1534622582657517,
      "grad_norm": 0.10989109836822063,
      "learning_rate": 2.7583706791713503e-06,
      "loss": 0.0287,
      "num_tokens": 131633233.0,
      "step": 1727
    },
    {
      "epoch": 2.154709918902059,
      "grad_norm": 0.11913978011568563,
      "learning_rate": 2.7535679992137338e-06,
      "loss": 0.0277,
      "num_tokens": 131709262.0,
      "step": 1728
    },
    {
      "epoch": 2.1559575795383656,
      "grad_norm": 0.12802965808075728,
      "learning_rate": 2.7487702993674647e-06,
      "loss": 0.0298,
      "num_tokens": 131785431.0,
      "step": 1729
    },
    {
      "epoch": 2.1572052401746724,
      "grad_norm": 0.11993178026046523,
      "learning_rate": 2.7439775883322228e-06,
      "loss": 0.0293,
      "num_tokens": 131861208.0,
      "step": 1730
    },
    {
      "epoch": 2.158452900810979,
      "grad_norm": 0.11896934214964502,
      "learning_rate": 2.739189874798639e-06,
      "loss": 0.0302,
      "num_tokens": 131937674.0,
      "step": 1731
    },
    {
      "epoch": 2.1597005614472864,
      "grad_norm": 0.13073507570615797,
      "learning_rate": 2.7344071674482874e-06,
      "loss": 0.029,
      "num_tokens": 132013613.0,
      "step": 1732
    },
    {
      "epoch": 2.160948222083593,
      "grad_norm": 0.12538080148618927,
      "learning_rate": 2.729629474953662e-06,
      "loss": 0.0299,
      "num_tokens": 132089741.0,
      "step": 1733
    },
    {
      "epoch": 2.1621958827199004,
      "grad_norm": 0.11470530230982613,
      "learning_rate": 2.7248568059781654e-06,
      "loss": 0.0283,
      "num_tokens": 132164908.0,
      "step": 1734
    },
    {
      "epoch": 2.163443543356207,
      "grad_norm": 0.12364100713980067,
      "learning_rate": 2.7200891691760838e-06,
      "loss": 0.0294,
      "num_tokens": 132242419.0,
      "step": 1735
    },
    {
      "epoch": 2.164691203992514,
      "grad_norm": 0.1124677553566527,
      "learning_rate": 2.715326573192588e-06,
      "loss": 0.0286,
      "num_tokens": 132318249.0,
      "step": 1736
    },
    {
      "epoch": 2.165938864628821,
      "grad_norm": 0.1297173190208866,
      "learning_rate": 2.710569026663702e-06,
      "loss": 0.0303,
      "num_tokens": 132394392.0,
      "step": 1737
    },
    {
      "epoch": 2.167186525265128,
      "grad_norm": 0.11636204301816074,
      "learning_rate": 2.705816538216296e-06,
      "loss": 0.0286,
      "num_tokens": 132471185.0,
      "step": 1738
    },
    {
      "epoch": 2.1684341859014347,
      "grad_norm": 0.11967082027832546,
      "learning_rate": 2.7010691164680696e-06,
      "loss": 0.0291,
      "num_tokens": 132547646.0,
      "step": 1739
    },
    {
      "epoch": 2.169681846537742,
      "grad_norm": 0.11365392254175426,
      "learning_rate": 2.696326770027533e-06,
      "loss": 0.0294,
      "num_tokens": 132622649.0,
      "step": 1740
    },
    {
      "epoch": 2.1709295071740486,
      "grad_norm": 0.13542950967226994,
      "learning_rate": 2.6915895074939912e-06,
      "loss": 0.0304,
      "num_tokens": 132699810.0,
      "step": 1741
    },
    {
      "epoch": 2.1721771678103554,
      "grad_norm": 0.1149886342950311,
      "learning_rate": 2.6868573374575356e-06,
      "loss": 0.0296,
      "num_tokens": 132775540.0,
      "step": 1742
    },
    {
      "epoch": 2.1734248284466626,
      "grad_norm": 0.13013765441316502,
      "learning_rate": 2.6821302684990204e-06,
      "loss": 0.0297,
      "num_tokens": 132851557.0,
      "step": 1743
    },
    {
      "epoch": 2.1746724890829694,
      "grad_norm": 0.12482263669044058,
      "learning_rate": 2.677408309190049e-06,
      "loss": 0.0292,
      "num_tokens": 132927251.0,
      "step": 1744
    },
    {
      "epoch": 2.175920149719276,
      "grad_norm": 0.12082906222739927,
      "learning_rate": 2.672691468092963e-06,
      "loss": 0.029,
      "num_tokens": 133004097.0,
      "step": 1745
    },
    {
      "epoch": 2.1771678103555834,
      "grad_norm": 0.11920653050743499,
      "learning_rate": 2.6679797537608184e-06,
      "loss": 0.0289,
      "num_tokens": 133079389.0,
      "step": 1746
    },
    {
      "epoch": 2.17841547099189,
      "grad_norm": 0.11613342645530199,
      "learning_rate": 2.6632731747373785e-06,
      "loss": 0.0288,
      "num_tokens": 133155198.0,
      "step": 1747
    },
    {
      "epoch": 2.179663131628197,
      "grad_norm": 0.123898197898323,
      "learning_rate": 2.658571739557096e-06,
      "loss": 0.0292,
      "num_tokens": 133231097.0,
      "step": 1748
    },
    {
      "epoch": 2.180910792264504,
      "grad_norm": 0.12283293422724019,
      "learning_rate": 2.653875456745092e-06,
      "loss": 0.0294,
      "num_tokens": 133308155.0,
      "step": 1749
    },
    {
      "epoch": 2.182158452900811,
      "grad_norm": 0.11096661481389018,
      "learning_rate": 2.6491843348171455e-06,
      "loss": 0.0285,
      "num_tokens": 133384157.0,
      "step": 1750
    },
    {
      "epoch": 2.183406113537118,
      "grad_norm": 0.1323212456611913,
      "learning_rate": 2.644498382279681e-06,
      "loss": 0.0317,
      "num_tokens": 133461262.0,
      "step": 1751
    },
    {
      "epoch": 2.184653774173425,
      "grad_norm": 0.11301431248801659,
      "learning_rate": 2.639817607629745e-06,
      "loss": 0.0293,
      "num_tokens": 133538603.0,
      "step": 1752
    },
    {
      "epoch": 2.1859014348097316,
      "grad_norm": 0.12487264376417616,
      "learning_rate": 2.635142019354995e-06,
      "loss": 0.0298,
      "num_tokens": 133614939.0,
      "step": 1753
    },
    {
      "epoch": 2.187149095446039,
      "grad_norm": 0.12391443914690654,
      "learning_rate": 2.6304716259336903e-06,
      "loss": 0.0301,
      "num_tokens": 133691509.0,
      "step": 1754
    },
    {
      "epoch": 2.1883967560823456,
      "grad_norm": 0.11436807139379579,
      "learning_rate": 2.6258064358346642e-06,
      "loss": 0.0293,
      "num_tokens": 133769447.0,
      "step": 1755
    },
    {
      "epoch": 2.1896444167186524,
      "grad_norm": 0.11605243292138384,
      "learning_rate": 2.621146457517314e-06,
      "loss": 0.0298,
      "num_tokens": 133845741.0,
      "step": 1756
    },
    {
      "epoch": 2.1908920773549596,
      "grad_norm": 0.12762818150692948,
      "learning_rate": 2.6164916994315916e-06,
      "loss": 0.0293,
      "num_tokens": 133922231.0,
      "step": 1757
    },
    {
      "epoch": 2.1921397379912664,
      "grad_norm": 0.10934869459664252,
      "learning_rate": 2.6118421700179795e-06,
      "loss": 0.0286,
      "num_tokens": 133997898.0,
      "step": 1758
    },
    {
      "epoch": 2.193387398627573,
      "grad_norm": 0.11556111476881359,
      "learning_rate": 2.6071978777074796e-06,
      "loss": 0.0285,
      "num_tokens": 134074139.0,
      "step": 1759
    },
    {
      "epoch": 2.1946350592638804,
      "grad_norm": 0.11684215494013445,
      "learning_rate": 2.6025588309215975e-06,
      "loss": 0.0288,
      "num_tokens": 134151443.0,
      "step": 1760
    },
    {
      "epoch": 2.195882719900187,
      "grad_norm": 0.10859546600803964,
      "learning_rate": 2.5979250380723287e-06,
      "loss": 0.0291,
      "num_tokens": 134228671.0,
      "step": 1761
    },
    {
      "epoch": 2.197130380536494,
      "grad_norm": 0.1184786660169172,
      "learning_rate": 2.5932965075621376e-06,
      "loss": 0.0295,
      "num_tokens": 134305328.0,
      "step": 1762
    },
    {
      "epoch": 2.198378041172801,
      "grad_norm": 0.11694769918714469,
      "learning_rate": 2.5886732477839514e-06,
      "loss": 0.0287,
      "num_tokens": 134381607.0,
      "step": 1763
    },
    {
      "epoch": 2.199625701809108,
      "grad_norm": 0.10869599540703445,
      "learning_rate": 2.584055267121137e-06,
      "loss": 0.0285,
      "num_tokens": 134457944.0,
      "step": 1764
    },
    {
      "epoch": 2.2008733624454146,
      "grad_norm": 0.11388521689730151,
      "learning_rate": 2.579442573947488e-06,
      "loss": 0.0292,
      "num_tokens": 134534177.0,
      "step": 1765
    },
    {
      "epoch": 2.202121023081722,
      "grad_norm": 0.12019491039099481,
      "learning_rate": 2.5748351766272127e-06,
      "loss": 0.0324,
      "num_tokens": 134610985.0,
      "step": 1766
    },
    {
      "epoch": 2.2033686837180286,
      "grad_norm": 0.12572287691368844,
      "learning_rate": 2.5702330835149137e-06,
      "loss": 0.0302,
      "num_tokens": 134686523.0,
      "step": 1767
    },
    {
      "epoch": 2.204616344354336,
      "grad_norm": 0.11387811266227701,
      "learning_rate": 2.5656363029555788e-06,
      "loss": 0.029,
      "num_tokens": 134762744.0,
      "step": 1768
    },
    {
      "epoch": 2.2058640049906426,
      "grad_norm": 0.13846833190116084,
      "learning_rate": 2.561044843284558e-06,
      "loss": 0.0296,
      "num_tokens": 134839137.0,
      "step": 1769
    },
    {
      "epoch": 2.2071116656269494,
      "grad_norm": 0.1555816479613367,
      "learning_rate": 2.556458712827558e-06,
      "loss": 0.0299,
      "num_tokens": 134914750.0,
      "step": 1770
    },
    {
      "epoch": 2.2083593262632566,
      "grad_norm": 0.11209797843880384,
      "learning_rate": 2.551877919900619e-06,
      "loss": 0.0289,
      "num_tokens": 134990474.0,
      "step": 1771
    },
    {
      "epoch": 2.2096069868995634,
      "grad_norm": 0.11661372301986613,
      "learning_rate": 2.5473024728101004e-06,
      "loss": 0.0294,
      "num_tokens": 135066743.0,
      "step": 1772
    },
    {
      "epoch": 2.21085464753587,
      "grad_norm": 0.12963541577443485,
      "learning_rate": 2.5427323798526747e-06,
      "loss": 0.0317,
      "num_tokens": 135142093.0,
      "step": 1773
    },
    {
      "epoch": 2.2121023081721773,
      "grad_norm": 0.11731994857913773,
      "learning_rate": 2.538167649315298e-06,
      "loss": 0.0281,
      "num_tokens": 135217549.0,
      "step": 1774
    },
    {
      "epoch": 2.213349968808484,
      "grad_norm": 0.11727497789837275,
      "learning_rate": 2.5336082894752084e-06,
      "loss": 0.028,
      "num_tokens": 135292208.0,
      "step": 1775
    },
    {
      "epoch": 2.214597629444791,
      "grad_norm": 0.10924884561562678,
      "learning_rate": 2.529054308599906e-06,
      "loss": 0.0285,
      "num_tokens": 135368860.0,
      "step": 1776
    },
    {
      "epoch": 2.215845290081098,
      "grad_norm": 0.1237241496137088,
      "learning_rate": 2.524505714947131e-06,
      "loss": 0.0302,
      "num_tokens": 135445704.0,
      "step": 1777
    },
    {
      "epoch": 2.217092950717405,
      "grad_norm": 0.14377566320772303,
      "learning_rate": 2.5199625167648576e-06,
      "loss": 0.0334,
      "num_tokens": 135521961.0,
      "step": 1778
    },
    {
      "epoch": 2.2183406113537116,
      "grad_norm": 0.11347892313850486,
      "learning_rate": 2.515424722291282e-06,
      "loss": 0.0283,
      "num_tokens": 135596262.0,
      "step": 1779
    },
    {
      "epoch": 2.219588271990019,
      "grad_norm": 0.12225004311131818,
      "learning_rate": 2.5108923397547934e-06,
      "loss": 0.0298,
      "num_tokens": 135673824.0,
      "step": 1780
    },
    {
      "epoch": 2.2208359326263256,
      "grad_norm": 0.11499506956311069,
      "learning_rate": 2.5063653773739705e-06,
      "loss": 0.0294,
      "num_tokens": 135749231.0,
      "step": 1781
    },
    {
      "epoch": 2.2220835932626324,
      "grad_norm": 0.1221359406411641,
      "learning_rate": 2.501843843357568e-06,
      "loss": 0.0291,
      "num_tokens": 135825518.0,
      "step": 1782
    },
    {
      "epoch": 2.2233312538989396,
      "grad_norm": 0.11102792400168941,
      "learning_rate": 2.4973277459044927e-06,
      "loss": 0.0288,
      "num_tokens": 135900872.0,
      "step": 1783
    },
    {
      "epoch": 2.2245789145352464,
      "grad_norm": 0.11593699845823963,
      "learning_rate": 2.4928170932037916e-06,
      "loss": 0.0288,
      "num_tokens": 135977380.0,
      "step": 1784
    },
    {
      "epoch": 2.225826575171553,
      "grad_norm": 0.1201879734680093,
      "learning_rate": 2.4883118934346446e-06,
      "loss": 0.0307,
      "num_tokens": 136052351.0,
      "step": 1785
    },
    {
      "epoch": 2.2270742358078603,
      "grad_norm": 0.11414556232554415,
      "learning_rate": 2.48381215476634e-06,
      "loss": 0.0287,
      "num_tokens": 136128645.0,
      "step": 1786
    },
    {
      "epoch": 2.228321896444167,
      "grad_norm": 0.11889689490126032,
      "learning_rate": 2.4793178853582624e-06,
      "loss": 0.0277,
      "num_tokens": 136207824.0,
      "step": 1787
    },
    {
      "epoch": 2.229569557080474,
      "grad_norm": 0.12178468802022104,
      "learning_rate": 2.474829093359881e-06,
      "loss": 0.0286,
      "num_tokens": 136283777.0,
      "step": 1788
    },
    {
      "epoch": 2.230817217716781,
      "grad_norm": 0.11219961060063727,
      "learning_rate": 2.4703457869107346e-06,
      "loss": 0.0284,
      "num_tokens": 136359801.0,
      "step": 1789
    },
    {
      "epoch": 2.232064878353088,
      "grad_norm": 0.11319967454171764,
      "learning_rate": 2.4658679741404106e-06,
      "loss": 0.0283,
      "num_tokens": 136436800.0,
      "step": 1790
    },
    {
      "epoch": 2.233312538989395,
      "grad_norm": 0.10577538482285624,
      "learning_rate": 2.461395663168539e-06,
      "loss": 0.0284,
      "num_tokens": 136512719.0,
      "step": 1791
    },
    {
      "epoch": 2.234560199625702,
      "grad_norm": 0.11201254599703794,
      "learning_rate": 2.4569288621047704e-06,
      "loss": 0.0276,
      "num_tokens": 136588384.0,
      "step": 1792
    },
    {
      "epoch": 2.2358078602620086,
      "grad_norm": 0.13162060846047863,
      "learning_rate": 2.452467579048764e-06,
      "loss": 0.0289,
      "num_tokens": 136665095.0,
      "step": 1793
    },
    {
      "epoch": 2.237055520898316,
      "grad_norm": 0.1139883066704449,
      "learning_rate": 2.4480118220901764e-06,
      "loss": 0.0284,
      "num_tokens": 136741352.0,
      "step": 1794
    },
    {
      "epoch": 2.2383031815346226,
      "grad_norm": 0.11192665390675736,
      "learning_rate": 2.4435615993086414e-06,
      "loss": 0.0284,
      "num_tokens": 136817373.0,
      "step": 1795
    },
    {
      "epoch": 2.2395508421709294,
      "grad_norm": 0.11005654494213833,
      "learning_rate": 2.4391169187737555e-06,
      "loss": 0.0294,
      "num_tokens": 136892152.0,
      "step": 1796
    },
    {
      "epoch": 2.2407985028072366,
      "grad_norm": 0.11372508392162389,
      "learning_rate": 2.434677788545071e-06,
      "loss": 0.03,
      "num_tokens": 136967795.0,
      "step": 1797
    },
    {
      "epoch": 2.2420461634435433,
      "grad_norm": 0.11176146479955043,
      "learning_rate": 2.4302442166720723e-06,
      "loss": 0.029,
      "num_tokens": 137043516.0,
      "step": 1798
    },
    {
      "epoch": 2.24329382407985,
      "grad_norm": 0.1313817503178842,
      "learning_rate": 2.4258162111941634e-06,
      "loss": 0.0323,
      "num_tokens": 137119810.0,
      "step": 1799
    },
    {
      "epoch": 2.2445414847161573,
      "grad_norm": 0.11693831324465238,
      "learning_rate": 2.42139378014066e-06,
      "loss": 0.0283,
      "num_tokens": 137195390.0,
      "step": 1800
    },
    {
      "epoch": 2.245789145352464,
      "grad_norm": 0.10613607438236099,
      "learning_rate": 2.416976931530764e-06,
      "loss": 0.0279,
      "num_tokens": 137270115.0,
      "step": 1801
    },
    {
      "epoch": 2.247036805988771,
      "grad_norm": 0.12911652294793682,
      "learning_rate": 2.4125656733735554e-06,
      "loss": 0.0299,
      "num_tokens": 137347278.0,
      "step": 1802
    },
    {
      "epoch": 2.248284466625078,
      "grad_norm": 0.1118436847971541,
      "learning_rate": 2.4081600136679805e-06,
      "loss": 0.0293,
      "num_tokens": 137422667.0,
      "step": 1803
    },
    {
      "epoch": 2.249532127261385,
      "grad_norm": 0.12017153749591643,
      "learning_rate": 2.403759960402834e-06,
      "loss": 0.0306,
      "num_tokens": 137499970.0,
      "step": 1804
    },
    {
      "epoch": 2.2507797878976916,
      "grad_norm": 0.11762476655558339,
      "learning_rate": 2.39936552155674e-06,
      "loss": 0.0285,
      "num_tokens": 137577581.0,
      "step": 1805
    },
    {
      "epoch": 2.252027448533999,
      "grad_norm": 0.1257011326248305,
      "learning_rate": 2.394976705098143e-06,
      "loss": 0.0293,
      "num_tokens": 137653132.0,
      "step": 1806
    },
    {
      "epoch": 2.2532751091703056,
      "grad_norm": 0.12517247918878088,
      "learning_rate": 2.3905935189852967e-06,
      "loss": 0.0301,
      "num_tokens": 137730362.0,
      "step": 1807
    },
    {
      "epoch": 2.254522769806613,
      "grad_norm": 0.11159522749229389,
      "learning_rate": 2.386215971166242e-06,
      "loss": 0.029,
      "num_tokens": 137806677.0,
      "step": 1808
    },
    {
      "epoch": 2.2557704304429196,
      "grad_norm": 0.1148546738045,
      "learning_rate": 2.381844069578793e-06,
      "loss": 0.0286,
      "num_tokens": 137882602.0,
      "step": 1809
    },
    {
      "epoch": 2.2570180910792264,
      "grad_norm": 0.11440912221708367,
      "learning_rate": 2.3774778221505316e-06,
      "loss": 0.0283,
      "num_tokens": 137958543.0,
      "step": 1810
    },
    {
      "epoch": 2.2582657517155336,
      "grad_norm": 0.1306981591104985,
      "learning_rate": 2.3731172367987856e-06,
      "loss": 0.046,
      "num_tokens": 138034225.0,
      "step": 1811
    },
    {
      "epoch": 2.2595134123518403,
      "grad_norm": 0.11196369587951115,
      "learning_rate": 2.3687623214306096e-06,
      "loss": 0.0289,
      "num_tokens": 138109276.0,
      "step": 1812
    },
    {
      "epoch": 2.260761072988147,
      "grad_norm": 0.11481976617907198,
      "learning_rate": 2.364413083942787e-06,
      "loss": 0.0286,
      "num_tokens": 138185059.0,
      "step": 1813
    },
    {
      "epoch": 2.2620087336244543,
      "grad_norm": 0.1257854519258705,
      "learning_rate": 2.3600695322217965e-06,
      "loss": 0.0302,
      "num_tokens": 138262086.0,
      "step": 1814
    },
    {
      "epoch": 2.263256394260761,
      "grad_norm": 0.12369190094856691,
      "learning_rate": 2.355731674143809e-06,
      "loss": 0.0301,
      "num_tokens": 138338212.0,
      "step": 1815
    },
    {
      "epoch": 2.264504054897068,
      "grad_norm": 0.12247698026358372,
      "learning_rate": 2.3513995175746757e-06,
      "loss": 0.0291,
      "num_tokens": 138414401.0,
      "step": 1816
    },
    {
      "epoch": 2.265751715533375,
      "grad_norm": 0.1271346776570538,
      "learning_rate": 2.3470730703699034e-06,
      "loss": 0.0297,
      "num_tokens": 138491333.0,
      "step": 1817
    },
    {
      "epoch": 2.266999376169682,
      "grad_norm": 0.11370340212148224,
      "learning_rate": 2.3427523403746496e-06,
      "loss": 0.0291,
      "num_tokens": 138567428.0,
      "step": 1818
    },
    {
      "epoch": 2.2682470368059886,
      "grad_norm": 0.11295824445642685,
      "learning_rate": 2.338437335423705e-06,
      "loss": 0.0288,
      "num_tokens": 138643908.0,
      "step": 1819
    },
    {
      "epoch": 2.269494697442296,
      "grad_norm": 0.11800330007227519,
      "learning_rate": 2.3341280633414763e-06,
      "loss": 0.0296,
      "num_tokens": 138718747.0,
      "step": 1820
    },
    {
      "epoch": 2.2707423580786026,
      "grad_norm": 0.11949774975707976,
      "learning_rate": 2.3298245319419755e-06,
      "loss": 0.0294,
      "num_tokens": 138795371.0,
      "step": 1821
    },
    {
      "epoch": 2.2719900187149094,
      "grad_norm": 0.11244072126302489,
      "learning_rate": 2.325526749028808e-06,
      "loss": 0.0287,
      "num_tokens": 138870711.0,
      "step": 1822
    },
    {
      "epoch": 2.2732376793512166,
      "grad_norm": 0.10880477678477739,
      "learning_rate": 2.321234722395152e-06,
      "loss": 0.0285,
      "num_tokens": 138946056.0,
      "step": 1823
    },
    {
      "epoch": 2.2744853399875233,
      "grad_norm": 0.12263235673529867,
      "learning_rate": 2.3169484598237484e-06,
      "loss": 0.0295,
      "num_tokens": 139022433.0,
      "step": 1824
    },
    {
      "epoch": 2.2757330006238305,
      "grad_norm": 0.11275334804297876,
      "learning_rate": 2.312667969086887e-06,
      "loss": 0.0291,
      "num_tokens": 139098812.0,
      "step": 1825
    },
    {
      "epoch": 2.2769806612601373,
      "grad_norm": 0.12780973779129554,
      "learning_rate": 2.308393257946393e-06,
      "loss": 0.0293,
      "num_tokens": 139174520.0,
      "step": 1826
    },
    {
      "epoch": 2.278228321896444,
      "grad_norm": 0.11873970508053847,
      "learning_rate": 2.304124334153608e-06,
      "loss": 0.0298,
      "num_tokens": 139251494.0,
      "step": 1827
    },
    {
      "epoch": 2.279475982532751,
      "grad_norm": 0.12524822945833192,
      "learning_rate": 2.2998612054493827e-06,
      "loss": 0.0293,
      "num_tokens": 139327208.0,
      "step": 1828
    },
    {
      "epoch": 2.280723643169058,
      "grad_norm": 0.12246702316926837,
      "learning_rate": 2.2956038795640573e-06,
      "loss": 0.0289,
      "num_tokens": 139402782.0,
      "step": 1829
    },
    {
      "epoch": 2.281971303805365,
      "grad_norm": 0.11636388498884984,
      "learning_rate": 2.291352364217449e-06,
      "loss": 0.0293,
      "num_tokens": 139479741.0,
      "step": 1830
    },
    {
      "epoch": 2.283218964441672,
      "grad_norm": 0.11228895431486227,
      "learning_rate": 2.287106667118841e-06,
      "loss": 0.0283,
      "num_tokens": 139554910.0,
      "step": 1831
    },
    {
      "epoch": 2.284466625077979,
      "grad_norm": 0.12324954885696872,
      "learning_rate": 2.2828667959669674e-06,
      "loss": 0.0298,
      "num_tokens": 139632084.0,
      "step": 1832
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 0.11630414434749232,
      "learning_rate": 2.2786327584499944e-06,
      "loss": 0.0286,
      "num_tokens": 139708348.0,
      "step": 1833
    },
    {
      "epoch": 2.286961946350593,
      "grad_norm": 0.11645485108839593,
      "learning_rate": 2.2744045622455112e-06,
      "loss": 0.0297,
      "num_tokens": 139785741.0,
      "step": 1834
    },
    {
      "epoch": 2.2882096069868996,
      "grad_norm": 0.12845198074631714,
      "learning_rate": 2.270182215020517e-06,
      "loss": 0.0296,
      "num_tokens": 139861955.0,
      "step": 1835
    },
    {
      "epoch": 2.2894572676232063,
      "grad_norm": 0.11029976660171407,
      "learning_rate": 2.2659657244314017e-06,
      "loss": 0.0285,
      "num_tokens": 139938627.0,
      "step": 1836
    },
    {
      "epoch": 2.2907049282595136,
      "grad_norm": 0.11686652041244248,
      "learning_rate": 2.26175509812394e-06,
      "loss": 0.0291,
      "num_tokens": 140016141.0,
      "step": 1837
    },
    {
      "epoch": 2.2919525888958203,
      "grad_norm": 0.1125996988495152,
      "learning_rate": 2.2575503437332677e-06,
      "loss": 0.0291,
      "num_tokens": 140091465.0,
      "step": 1838
    },
    {
      "epoch": 2.293200249532127,
      "grad_norm": 0.11916698980967634,
      "learning_rate": 2.2533514688838755e-06,
      "loss": 0.0295,
      "num_tokens": 140168133.0,
      "step": 1839
    },
    {
      "epoch": 2.2944479101684343,
      "grad_norm": 0.11855479850727757,
      "learning_rate": 2.2491584811895927e-06,
      "loss": 0.0306,
      "num_tokens": 140245782.0,
      "step": 1840
    },
    {
      "epoch": 2.295695570804741,
      "grad_norm": 0.1135002870486832,
      "learning_rate": 2.244971388253576e-06,
      "loss": 0.029,
      "num_tokens": 140321664.0,
      "step": 1841
    },
    {
      "epoch": 2.2969432314410483,
      "grad_norm": 0.12754892462978165,
      "learning_rate": 2.2407901976682884e-06,
      "loss": 0.0287,
      "num_tokens": 140398059.0,
      "step": 1842
    },
    {
      "epoch": 2.298190892077355,
      "grad_norm": 0.10743682818469019,
      "learning_rate": 2.2366149170154907e-06,
      "loss": 0.0289,
      "num_tokens": 140474424.0,
      "step": 1843
    },
    {
      "epoch": 2.299438552713662,
      "grad_norm": 0.1085712382920245,
      "learning_rate": 2.232445553866231e-06,
      "loss": 0.0282,
      "num_tokens": 140548741.0,
      "step": 1844
    },
    {
      "epoch": 2.3006862133499686,
      "grad_norm": 0.11188206978220724,
      "learning_rate": 2.228282115780824e-06,
      "loss": 0.0285,
      "num_tokens": 140626013.0,
      "step": 1845
    },
    {
      "epoch": 2.301933873986276,
      "grad_norm": 0.11573511063173106,
      "learning_rate": 2.22412461030884e-06,
      "loss": 0.0295,
      "num_tokens": 140702145.0,
      "step": 1846
    },
    {
      "epoch": 2.3031815346225826,
      "grad_norm": 0.11974119831273712,
      "learning_rate": 2.2199730449890964e-06,
      "loss": 0.029,
      "num_tokens": 140777410.0,
      "step": 1847
    },
    {
      "epoch": 2.30442919525889,
      "grad_norm": 0.12021299737556257,
      "learning_rate": 2.215827427349635e-06,
      "loss": 0.0292,
      "num_tokens": 140853428.0,
      "step": 1848
    },
    {
      "epoch": 2.3056768558951966,
      "grad_norm": 0.12185882384047894,
      "learning_rate": 2.211687764907711e-06,
      "loss": 0.029,
      "num_tokens": 140929302.0,
      "step": 1849
    },
    {
      "epoch": 2.3069245165315033,
      "grad_norm": 0.12520481675496664,
      "learning_rate": 2.2075540651697873e-06,
      "loss": 0.0299,
      "num_tokens": 141007578.0,
      "step": 1850
    },
    {
      "epoch": 2.3081721771678105,
      "grad_norm": 0.12383609326667641,
      "learning_rate": 2.2034263356315087e-06,
      "loss": 0.0294,
      "num_tokens": 141084153.0,
      "step": 1851
    },
    {
      "epoch": 2.3094198378041173,
      "grad_norm": 0.11399711978377276,
      "learning_rate": 2.1993045837776957e-06,
      "loss": 0.0276,
      "num_tokens": 141161746.0,
      "step": 1852
    },
    {
      "epoch": 2.310667498440424,
      "grad_norm": 0.12307731559682583,
      "learning_rate": 2.195188817082331e-06,
      "loss": 0.0292,
      "num_tokens": 141236998.0,
      "step": 1853
    },
    {
      "epoch": 2.3119151590767313,
      "grad_norm": 0.12325195165579966,
      "learning_rate": 2.1910790430085465e-06,
      "loss": 0.0285,
      "num_tokens": 141313798.0,
      "step": 1854
    },
    {
      "epoch": 2.313162819713038,
      "grad_norm": 0.12353402922323736,
      "learning_rate": 2.1869752690086e-06,
      "loss": 0.0296,
      "num_tokens": 141391359.0,
      "step": 1855
    },
    {
      "epoch": 2.314410480349345,
      "grad_norm": 0.11990914943000852,
      "learning_rate": 2.1828775025238787e-06,
      "loss": 0.0299,
      "num_tokens": 141469951.0,
      "step": 1856
    },
    {
      "epoch": 2.315658140985652,
      "grad_norm": 0.11789036943836603,
      "learning_rate": 2.1787857509848693e-06,
      "loss": 0.0294,
      "num_tokens": 141546155.0,
      "step": 1857
    },
    {
      "epoch": 2.316905801621959,
      "grad_norm": 0.11643916471269543,
      "learning_rate": 2.174700021811153e-06,
      "loss": 0.029,
      "num_tokens": 141622896.0,
      "step": 1858
    },
    {
      "epoch": 2.3181534622582656,
      "grad_norm": 0.11095738581724128,
      "learning_rate": 2.1706203224113944e-06,
      "loss": 0.0286,
      "num_tokens": 141699035.0,
      "step": 1859
    },
    {
      "epoch": 2.319401122894573,
      "grad_norm": 0.12453893497423184,
      "learning_rate": 2.1665466601833197e-06,
      "loss": 0.03,
      "num_tokens": 141776543.0,
      "step": 1860
    },
    {
      "epoch": 2.3206487835308796,
      "grad_norm": 0.11909510325046865,
      "learning_rate": 2.162479042513711e-06,
      "loss": 0.0295,
      "num_tokens": 141852198.0,
      "step": 1861
    },
    {
      "epoch": 2.3218964441671863,
      "grad_norm": 0.11584200227594135,
      "learning_rate": 2.158417476778388e-06,
      "loss": 0.029,
      "num_tokens": 141926871.0,
      "step": 1862
    },
    {
      "epoch": 2.3231441048034935,
      "grad_norm": 0.11689424136269533,
      "learning_rate": 2.1543619703421975e-06,
      "loss": 0.0293,
      "num_tokens": 142003627.0,
      "step": 1863
    },
    {
      "epoch": 2.3243917654398003,
      "grad_norm": 0.12009870206278882,
      "learning_rate": 2.1503125305589976e-06,
      "loss": 0.0292,
      "num_tokens": 142079455.0,
      "step": 1864
    },
    {
      "epoch": 2.3256394260761075,
      "grad_norm": 0.11985517420175247,
      "learning_rate": 2.146269164771648e-06,
      "loss": 0.0298,
      "num_tokens": 142156920.0,
      "step": 1865
    },
    {
      "epoch": 2.3268870867124143,
      "grad_norm": 0.10358166580215797,
      "learning_rate": 2.142231880311992e-06,
      "loss": 0.0275,
      "num_tokens": 142232961.0,
      "step": 1866
    },
    {
      "epoch": 2.328134747348721,
      "grad_norm": 0.11680882154313593,
      "learning_rate": 2.1382006845008456e-06,
      "loss": 0.0299,
      "num_tokens": 142308930.0,
      "step": 1867
    },
    {
      "epoch": 2.329382407985028,
      "grad_norm": 0.11287699434889473,
      "learning_rate": 2.1341755846479868e-06,
      "loss": 0.0288,
      "num_tokens": 142385073.0,
      "step": 1868
    },
    {
      "epoch": 2.330630068621335,
      "grad_norm": 0.11922671998430709,
      "learning_rate": 2.1301565880521387e-06,
      "loss": 0.0299,
      "num_tokens": 142461069.0,
      "step": 1869
    },
    {
      "epoch": 2.331877729257642,
      "grad_norm": 0.11411670675268805,
      "learning_rate": 2.1261437020009565e-06,
      "loss": 0.0288,
      "num_tokens": 142536513.0,
      "step": 1870
    },
    {
      "epoch": 2.333125389893949,
      "grad_norm": 0.11086492410716439,
      "learning_rate": 2.122136933771014e-06,
      "loss": 0.0275,
      "num_tokens": 142613035.0,
      "step": 1871
    },
    {
      "epoch": 2.334373050530256,
      "grad_norm": 0.11355073632493284,
      "learning_rate": 2.118136290627795e-06,
      "loss": 0.0281,
      "num_tokens": 142687991.0,
      "step": 1872
    },
    {
      "epoch": 2.3356207111665626,
      "grad_norm": 0.13046047958341747,
      "learning_rate": 2.114141779825674e-06,
      "loss": 0.03,
      "num_tokens": 142765175.0,
      "step": 1873
    },
    {
      "epoch": 2.3368683718028698,
      "grad_norm": 0.12467814349298817,
      "learning_rate": 2.110153408607904e-06,
      "loss": 0.0289,
      "num_tokens": 142840996.0,
      "step": 1874
    },
    {
      "epoch": 2.3381160324391765,
      "grad_norm": 0.11966459047710178,
      "learning_rate": 2.1061711842066124e-06,
      "loss": 0.0296,
      "num_tokens": 142916895.0,
      "step": 1875
    },
    {
      "epoch": 2.3393636930754833,
      "grad_norm": 0.11610628096770173,
      "learning_rate": 2.1021951138427736e-06,
      "loss": 0.0289,
      "num_tokens": 142992632.0,
      "step": 1876
    },
    {
      "epoch": 2.3406113537117905,
      "grad_norm": 0.10611006288527419,
      "learning_rate": 2.0982252047262025e-06,
      "loss": 0.0289,
      "num_tokens": 143068924.0,
      "step": 1877
    },
    {
      "epoch": 2.3418590143480973,
      "grad_norm": 0.11261369317367498,
      "learning_rate": 2.094261464055548e-06,
      "loss": 0.0284,
      "num_tokens": 143144300.0,
      "step": 1878
    },
    {
      "epoch": 2.343106674984404,
      "grad_norm": 0.11413669262158226,
      "learning_rate": 2.0903038990182684e-06,
      "loss": 0.0286,
      "num_tokens": 143220726.0,
      "step": 1879
    },
    {
      "epoch": 2.3443543356207113,
      "grad_norm": 0.12312252563671286,
      "learning_rate": 2.086352516790624e-06,
      "loss": 0.0292,
      "num_tokens": 143296504.0,
      "step": 1880
    },
    {
      "epoch": 2.345601996257018,
      "grad_norm": 0.12301300351069973,
      "learning_rate": 2.082407324537668e-06,
      "loss": 0.03,
      "num_tokens": 143373399.0,
      "step": 1881
    },
    {
      "epoch": 2.3468496568933253,
      "grad_norm": 0.11366967533685554,
      "learning_rate": 2.078468329413223e-06,
      "loss": 0.0284,
      "num_tokens": 143450300.0,
      "step": 1882
    },
    {
      "epoch": 2.348097317529632,
      "grad_norm": 0.12741351497399822,
      "learning_rate": 2.07453553855988e-06,
      "loss": 0.0298,
      "num_tokens": 143526458.0,
      "step": 1883
    },
    {
      "epoch": 2.349344978165939,
      "grad_norm": 0.1264248370411575,
      "learning_rate": 2.0706089591089785e-06,
      "loss": 0.0296,
      "num_tokens": 143602427.0,
      "step": 1884
    },
    {
      "epoch": 2.3505926388022456,
      "grad_norm": 0.11387889117213461,
      "learning_rate": 2.0666885981805916e-06,
      "loss": 0.0284,
      "num_tokens": 143679216.0,
      "step": 1885
    },
    {
      "epoch": 2.3518402994385528,
      "grad_norm": 0.1313321787677014,
      "learning_rate": 2.0627744628835196e-06,
      "loss": 0.0291,
      "num_tokens": 143755004.0,
      "step": 1886
    },
    {
      "epoch": 2.3530879600748595,
      "grad_norm": 0.11866227579559852,
      "learning_rate": 2.058866560315273e-06,
      "loss": 0.0284,
      "num_tokens": 143831363.0,
      "step": 1887
    },
    {
      "epoch": 2.3543356207111668,
      "grad_norm": 0.11693130482462902,
      "learning_rate": 2.054964897562061e-06,
      "loss": 0.0286,
      "num_tokens": 143906823.0,
      "step": 1888
    },
    {
      "epoch": 2.3555832813474735,
      "grad_norm": 0.12387166405220712,
      "learning_rate": 2.0510694816987724e-06,
      "loss": 0.0292,
      "num_tokens": 143982490.0,
      "step": 1889
    },
    {
      "epoch": 2.3568309419837803,
      "grad_norm": 0.11625853498164135,
      "learning_rate": 2.047180319788981e-06,
      "loss": 0.0293,
      "num_tokens": 144059126.0,
      "step": 1890
    },
    {
      "epoch": 2.3580786026200875,
      "grad_norm": 0.11795665021289878,
      "learning_rate": 2.0432974188849103e-06,
      "loss": 0.0294,
      "num_tokens": 144134816.0,
      "step": 1891
    },
    {
      "epoch": 2.3593262632563943,
      "grad_norm": 0.11653403992770552,
      "learning_rate": 2.0394207860274304e-06,
      "loss": 0.029,
      "num_tokens": 144210634.0,
      "step": 1892
    },
    {
      "epoch": 2.360573923892701,
      "grad_norm": 0.11661988331669815,
      "learning_rate": 2.035550428246053e-06,
      "loss": 0.0302,
      "num_tokens": 144286780.0,
      "step": 1893
    },
    {
      "epoch": 2.3618215845290083,
      "grad_norm": 0.10878502664004895,
      "learning_rate": 2.0316863525589037e-06,
      "loss": 0.028,
      "num_tokens": 144364570.0,
      "step": 1894
    },
    {
      "epoch": 2.363069245165315,
      "grad_norm": 0.11560765133919251,
      "learning_rate": 2.0278285659727187e-06,
      "loss": 0.0283,
      "num_tokens": 144438936.0,
      "step": 1895
    },
    {
      "epoch": 2.364316905801622,
      "grad_norm": 0.12027357703372889,
      "learning_rate": 2.023977075482833e-06,
      "loss": 0.0297,
      "num_tokens": 144515667.0,
      "step": 1896
    },
    {
      "epoch": 2.365564566437929,
      "grad_norm": 0.11720329585672999,
      "learning_rate": 2.0201318880731633e-06,
      "loss": 0.0289,
      "num_tokens": 144592784.0,
      "step": 1897
    },
    {
      "epoch": 2.3668122270742358,
      "grad_norm": 0.12262761391702148,
      "learning_rate": 2.0162930107161963e-06,
      "loss": 0.0287,
      "num_tokens": 144668333.0,
      "step": 1898
    },
    {
      "epoch": 2.3680598877105425,
      "grad_norm": 0.11209003719114731,
      "learning_rate": 2.012460450372976e-06,
      "loss": 0.0288,
      "num_tokens": 144744725.0,
      "step": 1899
    },
    {
      "epoch": 2.3693075483468498,
      "grad_norm": 0.12214495630165384,
      "learning_rate": 2.0086342139930932e-06,
      "loss": 0.0298,
      "num_tokens": 144821579.0,
      "step": 1900
    },
    {
      "epoch": 2.3705552089831565,
      "grad_norm": 0.12046060228472436,
      "learning_rate": 2.004814308514671e-06,
      "loss": 0.0287,
      "num_tokens": 144897323.0,
      "step": 1901
    },
    {
      "epoch": 2.3718028696194633,
      "grad_norm": 0.11069917239375722,
      "learning_rate": 2.001000740864353e-06,
      "loss": 0.0277,
      "num_tokens": 144972434.0,
      "step": 1902
    },
    {
      "epoch": 2.3730505302557705,
      "grad_norm": 0.11894361619187248,
      "learning_rate": 1.9971935179572893e-06,
      "loss": 0.0293,
      "num_tokens": 145048953.0,
      "step": 1903
    },
    {
      "epoch": 2.3742981908920773,
      "grad_norm": 0.11522118972823835,
      "learning_rate": 1.993392646697127e-06,
      "loss": 0.0287,
      "num_tokens": 145124171.0,
      "step": 1904
    },
    {
      "epoch": 2.3755458515283845,
      "grad_norm": 0.11993473456085262,
      "learning_rate": 1.9895981339759927e-06,
      "loss": 0.0298,
      "num_tokens": 145200923.0,
      "step": 1905
    },
    {
      "epoch": 2.3767935121646913,
      "grad_norm": 0.11659064385891797,
      "learning_rate": 1.985809986674487e-06,
      "loss": 0.0304,
      "num_tokens": 145277642.0,
      "step": 1906
    },
    {
      "epoch": 2.378041172800998,
      "grad_norm": 0.11233951997406662,
      "learning_rate": 1.982028211661665e-06,
      "loss": 0.0282,
      "num_tokens": 145354802.0,
      "step": 1907
    },
    {
      "epoch": 2.3792888334373052,
      "grad_norm": 0.12409125074153357,
      "learning_rate": 1.9782528157950266e-06,
      "loss": 0.0289,
      "num_tokens": 145430480.0,
      "step": 1908
    },
    {
      "epoch": 2.380536494073612,
      "grad_norm": 0.11627883738278946,
      "learning_rate": 1.974483805920508e-06,
      "loss": 0.0292,
      "num_tokens": 145507192.0,
      "step": 1909
    },
    {
      "epoch": 2.3817841547099188,
      "grad_norm": 0.12734639465992706,
      "learning_rate": 1.970721188872461e-06,
      "loss": 0.0302,
      "num_tokens": 145583500.0,
      "step": 1910
    },
    {
      "epoch": 2.383031815346226,
      "grad_norm": 0.11009739257647629,
      "learning_rate": 1.966964971473649e-06,
      "loss": 0.0282,
      "num_tokens": 145659305.0,
      "step": 1911
    },
    {
      "epoch": 2.3842794759825328,
      "grad_norm": 0.12085112973508644,
      "learning_rate": 1.9632151605352296e-06,
      "loss": 0.0288,
      "num_tokens": 145735646.0,
      "step": 1912
    },
    {
      "epoch": 2.3855271366188395,
      "grad_norm": 0.11074193233911775,
      "learning_rate": 1.9594717628567432e-06,
      "loss": 0.029,
      "num_tokens": 145810718.0,
      "step": 1913
    },
    {
      "epoch": 2.3867747972551467,
      "grad_norm": 0.12306001735547971,
      "learning_rate": 1.9557347852261007e-06,
      "loss": 0.0294,
      "num_tokens": 145886926.0,
      "step": 1914
    },
    {
      "epoch": 2.3880224578914535,
      "grad_norm": 0.11756178292941767,
      "learning_rate": 1.9520042344195727e-06,
      "loss": 0.0297,
      "num_tokens": 145963538.0,
      "step": 1915
    },
    {
      "epoch": 2.3892701185277603,
      "grad_norm": 0.12225103966132156,
      "learning_rate": 1.9482801172017758e-06,
      "loss": 0.0288,
      "num_tokens": 146039990.0,
      "step": 1916
    },
    {
      "epoch": 2.3905177791640675,
      "grad_norm": 0.11059553718727183,
      "learning_rate": 1.9445624403256576e-06,
      "loss": 0.0285,
      "num_tokens": 146116517.0,
      "step": 1917
    },
    {
      "epoch": 2.3917654398003743,
      "grad_norm": 0.12578612156742616,
      "learning_rate": 1.940851210532493e-06,
      "loss": 0.0298,
      "num_tokens": 146192937.0,
      "step": 1918
    },
    {
      "epoch": 2.393013100436681,
      "grad_norm": 0.11893920438902063,
      "learning_rate": 1.937146434551863e-06,
      "loss": 0.0283,
      "num_tokens": 146267536.0,
      "step": 1919
    },
    {
      "epoch": 2.3942607610729882,
      "grad_norm": 0.11149478865971485,
      "learning_rate": 1.933448119101644e-06,
      "loss": 0.0289,
      "num_tokens": 146344824.0,
      "step": 1920
    },
    {
      "epoch": 2.395508421709295,
      "grad_norm": 0.11882385786045513,
      "learning_rate": 1.929756270888003e-06,
      "loss": 0.0289,
      "num_tokens": 146421929.0,
      "step": 1921
    },
    {
      "epoch": 2.3967560823456022,
      "grad_norm": 0.10727470152443296,
      "learning_rate": 1.9260708966053744e-06,
      "loss": 0.0283,
      "num_tokens": 146497708.0,
      "step": 1922
    },
    {
      "epoch": 2.398003742981909,
      "grad_norm": 0.1094868205451024,
      "learning_rate": 1.9223920029364555e-06,
      "loss": 0.0283,
      "num_tokens": 146573739.0,
      "step": 1923
    },
    {
      "epoch": 2.3992514036182158,
      "grad_norm": 0.11691359954601226,
      "learning_rate": 1.9187195965521934e-06,
      "loss": 0.0293,
      "num_tokens": 146650698.0,
      "step": 1924
    },
    {
      "epoch": 2.4004990642545225,
      "grad_norm": 0.10662908498967849,
      "learning_rate": 1.9150536841117713e-06,
      "loss": 0.0286,
      "num_tokens": 146726679.0,
      "step": 1925
    },
    {
      "epoch": 2.4017467248908297,
      "grad_norm": 0.14014329804407294,
      "learning_rate": 1.911394272262595e-06,
      "loss": 0.0294,
      "num_tokens": 146804849.0,
      "step": 1926
    },
    {
      "epoch": 2.4029943855271365,
      "grad_norm": 0.12226614216324987,
      "learning_rate": 1.907741367640286e-06,
      "loss": 0.0299,
      "num_tokens": 146881226.0,
      "step": 1927
    },
    {
      "epoch": 2.4042420461634437,
      "grad_norm": 0.13585310316541752,
      "learning_rate": 1.9040949768686646e-06,
      "loss": 0.0305,
      "num_tokens": 146957259.0,
      "step": 1928
    },
    {
      "epoch": 2.4054897067997505,
      "grad_norm": 0.10302972876901556,
      "learning_rate": 1.900455106559737e-06,
      "loss": 0.0273,
      "num_tokens": 147031920.0,
      "step": 1929
    },
    {
      "epoch": 2.4067373674360573,
      "grad_norm": 0.1215937415585842,
      "learning_rate": 1.8968217633136909e-06,
      "loss": 0.0287,
      "num_tokens": 147109354.0,
      "step": 1930
    },
    {
      "epoch": 2.4079850280723645,
      "grad_norm": 0.11880054224376227,
      "learning_rate": 1.893194953718875e-06,
      "loss": 0.0287,
      "num_tokens": 147185080.0,
      "step": 1931
    },
    {
      "epoch": 2.4092326887086712,
      "grad_norm": 0.11355682893486546,
      "learning_rate": 1.8895746843517892e-06,
      "loss": 0.0291,
      "num_tokens": 147260147.0,
      "step": 1932
    },
    {
      "epoch": 2.410480349344978,
      "grad_norm": 0.1210669123668994,
      "learning_rate": 1.8859609617770786e-06,
      "loss": 0.0302,
      "num_tokens": 147336387.0,
      "step": 1933
    },
    {
      "epoch": 2.4117280099812852,
      "grad_norm": 0.11208579999965458,
      "learning_rate": 1.8823537925475143e-06,
      "loss": 0.0283,
      "num_tokens": 147411743.0,
      "step": 1934
    },
    {
      "epoch": 2.412975670617592,
      "grad_norm": 0.12438861535121437,
      "learning_rate": 1.8787531832039846e-06,
      "loss": 0.0294,
      "num_tokens": 147487099.0,
      "step": 1935
    },
    {
      "epoch": 2.4142233312538988,
      "grad_norm": 0.11521486060217191,
      "learning_rate": 1.8751591402754802e-06,
      "loss": 0.0295,
      "num_tokens": 147562576.0,
      "step": 1936
    },
    {
      "epoch": 2.415470991890206,
      "grad_norm": 0.12833118858246773,
      "learning_rate": 1.8715716702790903e-06,
      "loss": 0.029,
      "num_tokens": 147639867.0,
      "step": 1937
    },
    {
      "epoch": 2.4167186525265127,
      "grad_norm": 0.1232632765835214,
      "learning_rate": 1.8679907797199798e-06,
      "loss": 0.0289,
      "num_tokens": 147716321.0,
      "step": 1938
    },
    {
      "epoch": 2.41796631316282,
      "grad_norm": 0.12388893166951646,
      "learning_rate": 1.8644164750913868e-06,
      "loss": 0.029,
      "num_tokens": 147792320.0,
      "step": 1939
    },
    {
      "epoch": 2.4192139737991267,
      "grad_norm": 0.11673892717341924,
      "learning_rate": 1.8608487628746072e-06,
      "loss": 0.0288,
      "num_tokens": 147867833.0,
      "step": 1940
    },
    {
      "epoch": 2.4204616344354335,
      "grad_norm": 0.12016233695124004,
      "learning_rate": 1.8572876495389808e-06,
      "loss": 0.0289,
      "num_tokens": 147944585.0,
      "step": 1941
    },
    {
      "epoch": 2.4217092950717403,
      "grad_norm": 0.12686819493850526,
      "learning_rate": 1.8537331415418802e-06,
      "loss": 0.0294,
      "num_tokens": 148020225.0,
      "step": 1942
    },
    {
      "epoch": 2.4229569557080475,
      "grad_norm": 0.11875350996516942,
      "learning_rate": 1.8501852453287056e-06,
      "loss": 0.03,
      "num_tokens": 148096408.0,
      "step": 1943
    },
    {
      "epoch": 2.4242046163443542,
      "grad_norm": 0.12486101318734383,
      "learning_rate": 1.846643967332865e-06,
      "loss": 0.0284,
      "num_tokens": 148172670.0,
      "step": 1944
    },
    {
      "epoch": 2.4254522769806615,
      "grad_norm": 0.11059777297523263,
      "learning_rate": 1.8431093139757635e-06,
      "loss": 0.0277,
      "num_tokens": 148248603.0,
      "step": 1945
    },
    {
      "epoch": 2.4266999376169682,
      "grad_norm": 0.10968511209097837,
      "learning_rate": 1.8395812916667974e-06,
      "loss": 0.0275,
      "num_tokens": 148324109.0,
      "step": 1946
    },
    {
      "epoch": 2.427947598253275,
      "grad_norm": 0.12136214189843847,
      "learning_rate": 1.836059906803339e-06,
      "loss": 0.0288,
      "num_tokens": 148400040.0,
      "step": 1947
    },
    {
      "epoch": 2.429195258889582,
      "grad_norm": 0.1181035158906749,
      "learning_rate": 1.832545165770721e-06,
      "loss": 0.0281,
      "num_tokens": 148477225.0,
      "step": 1948
    },
    {
      "epoch": 2.430442919525889,
      "grad_norm": 0.1259344854755692,
      "learning_rate": 1.8290370749422327e-06,
      "loss": 0.0295,
      "num_tokens": 148554162.0,
      "step": 1949
    },
    {
      "epoch": 2.4316905801621957,
      "grad_norm": 0.11410632203785342,
      "learning_rate": 1.8255356406791036e-06,
      "loss": 0.0282,
      "num_tokens": 148629585.0,
      "step": 1950
    },
    {
      "epoch": 2.432938240798503,
      "grad_norm": 0.12856021094192358,
      "learning_rate": 1.82204086933049e-06,
      "loss": 0.0296,
      "num_tokens": 148706439.0,
      "step": 1951
    },
    {
      "epoch": 2.4341859014348097,
      "grad_norm": 0.11923685601983912,
      "learning_rate": 1.8185527672334712e-06,
      "loss": 0.0294,
      "num_tokens": 148781899.0,
      "step": 1952
    },
    {
      "epoch": 2.4354335620711165,
      "grad_norm": 0.12128121310656399,
      "learning_rate": 1.8150713407130283e-06,
      "loss": 0.0285,
      "num_tokens": 148857892.0,
      "step": 1953
    },
    {
      "epoch": 2.4366812227074237,
      "grad_norm": 0.11220293138970718,
      "learning_rate": 1.8115965960820414e-06,
      "loss": 0.0287,
      "num_tokens": 148933540.0,
      "step": 1954
    },
    {
      "epoch": 2.4379288833437305,
      "grad_norm": 0.12701814936674094,
      "learning_rate": 1.8081285396412738e-06,
      "loss": 0.0309,
      "num_tokens": 149009410.0,
      "step": 1955
    },
    {
      "epoch": 2.4391765439800372,
      "grad_norm": 0.12316860118736007,
      "learning_rate": 1.8046671776793584e-06,
      "loss": 0.0292,
      "num_tokens": 149085567.0,
      "step": 1956
    },
    {
      "epoch": 2.4404242046163445,
      "grad_norm": 0.12579245754407814,
      "learning_rate": 1.80121251647279e-06,
      "loss": 0.0293,
      "num_tokens": 149160818.0,
      "step": 1957
    },
    {
      "epoch": 2.4416718652526512,
      "grad_norm": 0.11864506441301292,
      "learning_rate": 1.7977645622859157e-06,
      "loss": 0.0295,
      "num_tokens": 149237095.0,
      "step": 1958
    },
    {
      "epoch": 2.442919525888958,
      "grad_norm": 0.12128068105962087,
      "learning_rate": 1.7943233213709173e-06,
      "loss": 0.0291,
      "num_tokens": 149313199.0,
      "step": 1959
    },
    {
      "epoch": 2.444167186525265,
      "grad_norm": 0.11867438858434401,
      "learning_rate": 1.7908887999678046e-06,
      "loss": 0.0286,
      "num_tokens": 149390816.0,
      "step": 1960
    },
    {
      "epoch": 2.445414847161572,
      "grad_norm": 0.10792932976220691,
      "learning_rate": 1.7874610043044027e-06,
      "loss": 0.0281,
      "num_tokens": 149466985.0,
      "step": 1961
    },
    {
      "epoch": 2.446662507797879,
      "grad_norm": 0.1561122611017201,
      "learning_rate": 1.7840399405963432e-06,
      "loss": 0.0346,
      "num_tokens": 149544590.0,
      "step": 1962
    },
    {
      "epoch": 2.447910168434186,
      "grad_norm": 0.12123511730152045,
      "learning_rate": 1.7806256150470472e-06,
      "loss": 0.0286,
      "num_tokens": 149619730.0,
      "step": 1963
    },
    {
      "epoch": 2.4491578290704927,
      "grad_norm": 0.10870931536325534,
      "learning_rate": 1.7772180338477173e-06,
      "loss": 0.0281,
      "num_tokens": 149694700.0,
      "step": 1964
    },
    {
      "epoch": 2.4504054897068,
      "grad_norm": 0.11432026693551328,
      "learning_rate": 1.7738172031773322e-06,
      "loss": 0.0288,
      "num_tokens": 149770875.0,
      "step": 1965
    },
    {
      "epoch": 2.4516531503431067,
      "grad_norm": 0.1237259018436631,
      "learning_rate": 1.7704231292026219e-06,
      "loss": 0.0286,
      "num_tokens": 149847328.0,
      "step": 1966
    },
    {
      "epoch": 2.4529008109794135,
      "grad_norm": 0.12251333691455961,
      "learning_rate": 1.76703581807807e-06,
      "loss": 0.0283,
      "num_tokens": 149923014.0,
      "step": 1967
    },
    {
      "epoch": 2.4541484716157207,
      "grad_norm": 0.13592170557138702,
      "learning_rate": 1.7636552759458963e-06,
      "loss": 0.0295,
      "num_tokens": 150000726.0,
      "step": 1968
    },
    {
      "epoch": 2.4553961322520275,
      "grad_norm": 0.10644631287643523,
      "learning_rate": 1.760281508936045e-06,
      "loss": 0.029,
      "num_tokens": 150076628.0,
      "step": 1969
    },
    {
      "epoch": 2.4566437928883342,
      "grad_norm": 0.11637380854434433,
      "learning_rate": 1.7569145231661738e-06,
      "loss": 0.029,
      "num_tokens": 150153366.0,
      "step": 1970
    },
    {
      "epoch": 2.4578914535246414,
      "grad_norm": 0.11721908963690945,
      "learning_rate": 1.753554324741648e-06,
      "loss": 0.0288,
      "num_tokens": 150229286.0,
      "step": 1971
    },
    {
      "epoch": 2.459139114160948,
      "grad_norm": 0.11861330158718068,
      "learning_rate": 1.7502009197555215e-06,
      "loss": 0.0283,
      "num_tokens": 150304975.0,
      "step": 1972
    },
    {
      "epoch": 2.460386774797255,
      "grad_norm": 0.11613107309040814,
      "learning_rate": 1.7468543142885308e-06,
      "loss": 0.0288,
      "num_tokens": 150380574.0,
      "step": 1973
    },
    {
      "epoch": 2.461634435433562,
      "grad_norm": 0.11791945836868803,
      "learning_rate": 1.7435145144090852e-06,
      "loss": 0.0297,
      "num_tokens": 150456868.0,
      "step": 1974
    },
    {
      "epoch": 2.462882096069869,
      "grad_norm": 0.10865521562257276,
      "learning_rate": 1.740181526173248e-06,
      "loss": 0.028,
      "num_tokens": 150532783.0,
      "step": 1975
    },
    {
      "epoch": 2.4641297567061757,
      "grad_norm": 0.12013505293295967,
      "learning_rate": 1.736855355624737e-06,
      "loss": 0.0289,
      "num_tokens": 150608207.0,
      "step": 1976
    },
    {
      "epoch": 2.465377417342483,
      "grad_norm": 0.1190130952026185,
      "learning_rate": 1.7335360087949048e-06,
      "loss": 0.0291,
      "num_tokens": 150683923.0,
      "step": 1977
    },
    {
      "epoch": 2.4666250779787897,
      "grad_norm": 0.12021618470381827,
      "learning_rate": 1.73022349170273e-06,
      "loss": 0.0295,
      "num_tokens": 150759319.0,
      "step": 1978
    },
    {
      "epoch": 2.467872738615097,
      "grad_norm": 0.1188294278109119,
      "learning_rate": 1.7269178103548057e-06,
      "loss": 0.0298,
      "num_tokens": 150836584.0,
      "step": 1979
    },
    {
      "epoch": 2.4691203992514037,
      "grad_norm": 0.11623486387948447,
      "learning_rate": 1.723618970745334e-06,
      "loss": 0.0293,
      "num_tokens": 150911549.0,
      "step": 1980
    },
    {
      "epoch": 2.4703680598877105,
      "grad_norm": 0.1176345828523494,
      "learning_rate": 1.7203269788561067e-06,
      "loss": 0.0291,
      "num_tokens": 150987402.0,
      "step": 1981
    },
    {
      "epoch": 2.4716157205240172,
      "grad_norm": 0.1170956124313601,
      "learning_rate": 1.7170418406564982e-06,
      "loss": 0.0288,
      "num_tokens": 151065490.0,
      "step": 1982
    },
    {
      "epoch": 2.4728633811603244,
      "grad_norm": 0.11242037500890412,
      "learning_rate": 1.7137635621034614e-06,
      "loss": 0.0283,
      "num_tokens": 151140755.0,
      "step": 1983
    },
    {
      "epoch": 2.474111041796631,
      "grad_norm": 0.12957828390475865,
      "learning_rate": 1.7104921491415038e-06,
      "loss": 0.0305,
      "num_tokens": 151219039.0,
      "step": 1984
    },
    {
      "epoch": 2.4753587024329384,
      "grad_norm": 0.1268773509148947,
      "learning_rate": 1.7072276077026856e-06,
      "loss": 0.0295,
      "num_tokens": 151296866.0,
      "step": 1985
    },
    {
      "epoch": 2.476606363069245,
      "grad_norm": 0.11827505548209889,
      "learning_rate": 1.7039699437066076e-06,
      "loss": 0.0289,
      "num_tokens": 151374496.0,
      "step": 1986
    },
    {
      "epoch": 2.477854023705552,
      "grad_norm": 0.12394239832221975,
      "learning_rate": 1.7007191630604003e-06,
      "loss": 0.0297,
      "num_tokens": 151449768.0,
      "step": 1987
    },
    {
      "epoch": 2.479101684341859,
      "grad_norm": 0.10729321452712806,
      "learning_rate": 1.6974752716587092e-06,
      "loss": 0.0281,
      "num_tokens": 151525242.0,
      "step": 1988
    },
    {
      "epoch": 2.480349344978166,
      "grad_norm": 0.11244564865318404,
      "learning_rate": 1.6942382753836912e-06,
      "loss": 0.0276,
      "num_tokens": 151600019.0,
      "step": 1989
    },
    {
      "epoch": 2.4815970056144727,
      "grad_norm": 0.11260476703513686,
      "learning_rate": 1.691008180105e-06,
      "loss": 0.0286,
      "num_tokens": 151676641.0,
      "step": 1990
    },
    {
      "epoch": 2.48284466625078,
      "grad_norm": 0.11508491747273103,
      "learning_rate": 1.6877849916797728e-06,
      "loss": 0.0287,
      "num_tokens": 151752758.0,
      "step": 1991
    },
    {
      "epoch": 2.4840923268870867,
      "grad_norm": 0.12295722659763934,
      "learning_rate": 1.684568715952626e-06,
      "loss": 0.0292,
      "num_tokens": 151828614.0,
      "step": 1992
    },
    {
      "epoch": 2.4853399875233935,
      "grad_norm": 0.11201936145104541,
      "learning_rate": 1.6813593587556392e-06,
      "loss": 0.0289,
      "num_tokens": 151904503.0,
      "step": 1993
    },
    {
      "epoch": 2.4865876481597007,
      "grad_norm": 0.11197536265010408,
      "learning_rate": 1.6781569259083463e-06,
      "loss": 0.0281,
      "num_tokens": 151980779.0,
      "step": 1994
    },
    {
      "epoch": 2.4878353087960074,
      "grad_norm": 0.127443601047076,
      "learning_rate": 1.6749614232177273e-06,
      "loss": 0.0294,
      "num_tokens": 152057218.0,
      "step": 1995
    },
    {
      "epoch": 2.489082969432314,
      "grad_norm": 0.1272175699987119,
      "learning_rate": 1.6717728564781927e-06,
      "loss": 0.0297,
      "num_tokens": 152134969.0,
      "step": 1996
    },
    {
      "epoch": 2.4903306300686214,
      "grad_norm": 0.11830686372514893,
      "learning_rate": 1.6685912314715797e-06,
      "loss": 0.0286,
      "num_tokens": 152209799.0,
      "step": 1997
    },
    {
      "epoch": 2.491578290704928,
      "grad_norm": 0.11785531730640933,
      "learning_rate": 1.6654165539671342e-06,
      "loss": 0.0292,
      "num_tokens": 152286463.0,
      "step": 1998
    },
    {
      "epoch": 2.492825951341235,
      "grad_norm": 0.11630082166589754,
      "learning_rate": 1.6622488297215079e-06,
      "loss": 0.0281,
      "num_tokens": 152362651.0,
      "step": 1999
    },
    {
      "epoch": 2.494073611977542,
      "grad_norm": 0.1147252131365444,
      "learning_rate": 1.6590880644787407e-06,
      "loss": 0.0274,
      "num_tokens": 152439284.0,
      "step": 2000
    },
    {
      "epoch": 2.495321272613849,
      "grad_norm": 0.10890875650452751,
      "learning_rate": 1.6559342639702563e-06,
      "loss": 0.0285,
      "num_tokens": 152516175.0,
      "step": 2001
    },
    {
      "epoch": 2.496568933250156,
      "grad_norm": 0.11369281469621417,
      "learning_rate": 1.6527874339148484e-06,
      "loss": 0.0281,
      "num_tokens": 152591844.0,
      "step": 2002
    },
    {
      "epoch": 2.497816593886463,
      "grad_norm": 0.11765443898571652,
      "learning_rate": 1.6496475800186702e-06,
      "loss": 0.0292,
      "num_tokens": 152669310.0,
      "step": 2003
    },
    {
      "epoch": 2.4990642545227697,
      "grad_norm": 0.12392954052131615,
      "learning_rate": 1.6465147079752264e-06,
      "loss": 0.0297,
      "num_tokens": 152745302.0,
      "step": 2004
    },
    {
      "epoch": 2.5003119151590765,
      "grad_norm": 0.10811929633453872,
      "learning_rate": 1.6433888234653614e-06,
      "loss": 0.0279,
      "num_tokens": 152821134.0,
      "step": 2005
    },
    {
      "epoch": 2.5015595757953837,
      "grad_norm": 0.12800440714521488,
      "learning_rate": 1.6402699321572485e-06,
      "loss": 0.0289,
      "num_tokens": 152897262.0,
      "step": 2006
    },
    {
      "epoch": 2.5028072364316905,
      "grad_norm": 0.11428461694118386,
      "learning_rate": 1.6371580397063788e-06,
      "loss": 0.0282,
      "num_tokens": 152973605.0,
      "step": 2007
    },
    {
      "epoch": 2.5040548970679977,
      "grad_norm": 0.10820362542898414,
      "learning_rate": 1.6340531517555563e-06,
      "loss": 0.0283,
      "num_tokens": 153049729.0,
      "step": 2008
    },
    {
      "epoch": 2.5053025577043044,
      "grad_norm": 0.12644172316361868,
      "learning_rate": 1.6309552739348804e-06,
      "loss": 0.0293,
      "num_tokens": 153127475.0,
      "step": 2009
    },
    {
      "epoch": 2.506550218340611,
      "grad_norm": 0.10836425357174682,
      "learning_rate": 1.6278644118617375e-06,
      "loss": 0.0285,
      "num_tokens": 153204165.0,
      "step": 2010
    },
    {
      "epoch": 2.5077978789769184,
      "grad_norm": 0.12135651297840683,
      "learning_rate": 1.6247805711407993e-06,
      "loss": 0.0295,
      "num_tokens": 153280453.0,
      "step": 2011
    },
    {
      "epoch": 2.509045539613225,
      "grad_norm": 0.11582565410724034,
      "learning_rate": 1.6217037573639983e-06,
      "loss": 0.029,
      "num_tokens": 153357256.0,
      "step": 2012
    },
    {
      "epoch": 2.5102932002495324,
      "grad_norm": 0.11923075497477043,
      "learning_rate": 1.6186339761105275e-06,
      "loss": 0.0295,
      "num_tokens": 153434260.0,
      "step": 2013
    },
    {
      "epoch": 2.511540860885839,
      "grad_norm": 0.11545869189350899,
      "learning_rate": 1.6155712329468305e-06,
      "loss": 0.0281,
      "num_tokens": 153512382.0,
      "step": 2014
    },
    {
      "epoch": 2.512788521522146,
      "grad_norm": 0.1106551868324314,
      "learning_rate": 1.6125155334265846e-06,
      "loss": 0.0292,
      "num_tokens": 153587250.0,
      "step": 2015
    },
    {
      "epoch": 2.5140361821584527,
      "grad_norm": 0.11147702747990101,
      "learning_rate": 1.6094668830906959e-06,
      "loss": 0.0285,
      "num_tokens": 153661781.0,
      "step": 2016
    },
    {
      "epoch": 2.51528384279476,
      "grad_norm": 0.13288602885472603,
      "learning_rate": 1.6064252874672904e-06,
      "loss": 0.031,
      "num_tokens": 153737392.0,
      "step": 2017
    },
    {
      "epoch": 2.5165315034310667,
      "grad_norm": 0.11513263509992315,
      "learning_rate": 1.6033907520717008e-06,
      "loss": 0.0284,
      "num_tokens": 153813161.0,
      "step": 2018
    },
    {
      "epoch": 2.517779164067374,
      "grad_norm": 0.1144863171593377,
      "learning_rate": 1.6003632824064553e-06,
      "loss": 0.0288,
      "num_tokens": 153889389.0,
      "step": 2019
    },
    {
      "epoch": 2.5190268247036807,
      "grad_norm": 0.1073831605303728,
      "learning_rate": 1.5973428839612727e-06,
      "loss": 0.0279,
      "num_tokens": 153966333.0,
      "step": 2020
    },
    {
      "epoch": 2.5202744853399874,
      "grad_norm": 0.1205683735174724,
      "learning_rate": 1.5943295622130483e-06,
      "loss": 0.0277,
      "num_tokens": 154042117.0,
      "step": 2021
    },
    {
      "epoch": 2.521522145976294,
      "grad_norm": 0.11709366889958371,
      "learning_rate": 1.5913233226258437e-06,
      "loss": 0.0278,
      "num_tokens": 154117674.0,
      "step": 2022
    },
    {
      "epoch": 2.5227698066126014,
      "grad_norm": 0.12142255989860215,
      "learning_rate": 1.5883241706508823e-06,
      "loss": 0.0296,
      "num_tokens": 154195162.0,
      "step": 2023
    },
    {
      "epoch": 2.524017467248908,
      "grad_norm": 0.11351365164635996,
      "learning_rate": 1.5853321117265317e-06,
      "loss": 0.0283,
      "num_tokens": 154271186.0,
      "step": 2024
    },
    {
      "epoch": 2.5252651278852154,
      "grad_norm": 0.12443820690995414,
      "learning_rate": 1.5823471512782983e-06,
      "loss": 0.0291,
      "num_tokens": 154347579.0,
      "step": 2025
    },
    {
      "epoch": 2.526512788521522,
      "grad_norm": 0.12077454617168958,
      "learning_rate": 1.579369294718819e-06,
      "loss": 0.0279,
      "num_tokens": 154422999.0,
      "step": 2026
    },
    {
      "epoch": 2.527760449157829,
      "grad_norm": 0.11841856753505717,
      "learning_rate": 1.5763985474478483e-06,
      "loss": 0.028,
      "num_tokens": 154498846.0,
      "step": 2027
    },
    {
      "epoch": 2.529008109794136,
      "grad_norm": 0.1191743469311657,
      "learning_rate": 1.5734349148522471e-06,
      "loss": 0.0292,
      "num_tokens": 154576045.0,
      "step": 2028
    },
    {
      "epoch": 2.530255770430443,
      "grad_norm": 0.12040356577829175,
      "learning_rate": 1.5704784023059788e-06,
      "loss": 0.0293,
      "num_tokens": 154652119.0,
      "step": 2029
    },
    {
      "epoch": 2.5315034310667497,
      "grad_norm": 0.1235315800743608,
      "learning_rate": 1.5675290151700937e-06,
      "loss": 0.0299,
      "num_tokens": 154728765.0,
      "step": 2030
    },
    {
      "epoch": 2.532751091703057,
      "grad_norm": 0.11007266293522812,
      "learning_rate": 1.5645867587927208e-06,
      "loss": 0.0276,
      "num_tokens": 154804766.0,
      "step": 2031
    },
    {
      "epoch": 2.5339987523393637,
      "grad_norm": 0.14515242423085778,
      "learning_rate": 1.561651638509062e-06,
      "loss": 0.039,
      "num_tokens": 154881886.0,
      "step": 2032
    },
    {
      "epoch": 2.5352464129756704,
      "grad_norm": 0.11030630179121877,
      "learning_rate": 1.5587236596413773e-06,
      "loss": 0.0288,
      "num_tokens": 154957220.0,
      "step": 2033
    },
    {
      "epoch": 2.5364940736119777,
      "grad_norm": 0.11515001091522285,
      "learning_rate": 1.5558028274989778e-06,
      "loss": 0.0282,
      "num_tokens": 155033151.0,
      "step": 2034
    },
    {
      "epoch": 2.5377417342482844,
      "grad_norm": 0.12846274937409335,
      "learning_rate": 1.5528891473782126e-06,
      "loss": 0.0305,
      "num_tokens": 155110509.0,
      "step": 2035
    },
    {
      "epoch": 2.5389893948845916,
      "grad_norm": 0.13067981208487092,
      "learning_rate": 1.5499826245624674e-06,
      "loss": 0.0302,
      "num_tokens": 155187066.0,
      "step": 2036
    },
    {
      "epoch": 2.5402370555208984,
      "grad_norm": 0.11907261818233296,
      "learning_rate": 1.547083264322145e-06,
      "loss": 0.0292,
      "num_tokens": 155263148.0,
      "step": 2037
    },
    {
      "epoch": 2.541484716157205,
      "grad_norm": 0.12299689014144845,
      "learning_rate": 1.5441910719146616e-06,
      "loss": 0.0288,
      "num_tokens": 155340431.0,
      "step": 2038
    },
    {
      "epoch": 2.542732376793512,
      "grad_norm": 0.13015131026580284,
      "learning_rate": 1.541306052584437e-06,
      "loss": 0.0303,
      "num_tokens": 155416415.0,
      "step": 2039
    },
    {
      "epoch": 2.543980037429819,
      "grad_norm": 0.11049380134309983,
      "learning_rate": 1.5384282115628834e-06,
      "loss": 0.0293,
      "num_tokens": 155494379.0,
      "step": 2040
    },
    {
      "epoch": 2.545227698066126,
      "grad_norm": 0.1162771794153917,
      "learning_rate": 1.5355575540683953e-06,
      "loss": 0.0291,
      "num_tokens": 155569548.0,
      "step": 2041
    },
    {
      "epoch": 2.546475358702433,
      "grad_norm": 0.12174837271925616,
      "learning_rate": 1.5326940853063443e-06,
      "loss": 0.0279,
      "num_tokens": 155647393.0,
      "step": 2042
    },
    {
      "epoch": 2.54772301933874,
      "grad_norm": 0.13480309447667146,
      "learning_rate": 1.5298378104690636e-06,
      "loss": 0.0298,
      "num_tokens": 155724327.0,
      "step": 2043
    },
    {
      "epoch": 2.5489706799750467,
      "grad_norm": 0.11661321788253896,
      "learning_rate": 1.5269887347358414e-06,
      "loss": 0.0283,
      "num_tokens": 155801249.0,
      "step": 2044
    },
    {
      "epoch": 2.5502183406113534,
      "grad_norm": 0.1247113771553184,
      "learning_rate": 1.5241468632729161e-06,
      "loss": 0.0289,
      "num_tokens": 155877398.0,
      "step": 2045
    },
    {
      "epoch": 2.5514660012476607,
      "grad_norm": 0.12248316911861816,
      "learning_rate": 1.5213122012334572e-06,
      "loss": 0.0291,
      "num_tokens": 155953871.0,
      "step": 2046
    },
    {
      "epoch": 2.5527136618839674,
      "grad_norm": 0.11970375576013077,
      "learning_rate": 1.5184847537575647e-06,
      "loss": 0.0298,
      "num_tokens": 156030983.0,
      "step": 2047
    },
    {
      "epoch": 2.5539613225202746,
      "grad_norm": 0.11663205314443578,
      "learning_rate": 1.5156645259722565e-06,
      "loss": 0.0288,
      "num_tokens": 156107454.0,
      "step": 2048
    },
    {
      "epoch": 2.5552089831565814,
      "grad_norm": 0.13006782115149237,
      "learning_rate": 1.5128515229914568e-06,
      "loss": 0.0286,
      "num_tokens": 156183133.0,
      "step": 2049
    },
    {
      "epoch": 2.556456643792888,
      "grad_norm": 0.1205402849568315,
      "learning_rate": 1.5100457499159897e-06,
      "loss": 0.0287,
      "num_tokens": 156258840.0,
      "step": 2050
    },
    {
      "epoch": 2.5577043044291954,
      "grad_norm": 0.11732612306689322,
      "learning_rate": 1.507247211833572e-06,
      "loss": 0.0284,
      "num_tokens": 156334760.0,
      "step": 2051
    },
    {
      "epoch": 2.558951965065502,
      "grad_norm": 0.1096467375696786,
      "learning_rate": 1.5044559138187967e-06,
      "loss": 0.0282,
      "num_tokens": 156411498.0,
      "step": 2052
    },
    {
      "epoch": 2.5601996257018094,
      "grad_norm": 0.10682251478899628,
      "learning_rate": 1.5016718609331315e-06,
      "loss": 0.0277,
      "num_tokens": 156486644.0,
      "step": 2053
    },
    {
      "epoch": 2.561447286338116,
      "grad_norm": 0.1231457023538011,
      "learning_rate": 1.4988950582249061e-06,
      "loss": 0.0279,
      "num_tokens": 156562424.0,
      "step": 2054
    },
    {
      "epoch": 2.562694946974423,
      "grad_norm": 0.11102241679752041,
      "learning_rate": 1.4961255107293044e-06,
      "loss": 0.0284,
      "num_tokens": 156638766.0,
      "step": 2055
    },
    {
      "epoch": 2.5639426076107297,
      "grad_norm": 0.11864437556503948,
      "learning_rate": 1.4933632234683506e-06,
      "loss": 0.0286,
      "num_tokens": 156715589.0,
      "step": 2056
    },
    {
      "epoch": 2.565190268247037,
      "grad_norm": 0.11512676728660194,
      "learning_rate": 1.4906082014509088e-06,
      "loss": 0.0286,
      "num_tokens": 156792091.0,
      "step": 2057
    },
    {
      "epoch": 2.5664379288833437,
      "grad_norm": 0.11182137223682226,
      "learning_rate": 1.4878604496726653e-06,
      "loss": 0.0284,
      "num_tokens": 156867419.0,
      "step": 2058
    },
    {
      "epoch": 2.567685589519651,
      "grad_norm": 0.11138300175009733,
      "learning_rate": 1.4851199731161243e-06,
      "loss": 0.0278,
      "num_tokens": 156942944.0,
      "step": 2059
    },
    {
      "epoch": 2.5689332501559576,
      "grad_norm": 0.11691444098223876,
      "learning_rate": 1.4823867767505981e-06,
      "loss": 0.0283,
      "num_tokens": 157019631.0,
      "step": 2060
    },
    {
      "epoch": 2.5701809107922644,
      "grad_norm": 0.11570369933372947,
      "learning_rate": 1.4796608655322001e-06,
      "loss": 0.0286,
      "num_tokens": 157096265.0,
      "step": 2061
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.12017368506601983,
      "learning_rate": 1.476942244403829e-06,
      "loss": 0.0288,
      "num_tokens": 157173740.0,
      "step": 2062
    },
    {
      "epoch": 2.5726762320648784,
      "grad_norm": 0.11019267524765289,
      "learning_rate": 1.4742309182951663e-06,
      "loss": 0.0288,
      "num_tokens": 157249715.0,
      "step": 2063
    },
    {
      "epoch": 2.573923892701185,
      "grad_norm": 0.10787044892409242,
      "learning_rate": 1.4715268921226677e-06,
      "loss": 0.0278,
      "num_tokens": 157325380.0,
      "step": 2064
    },
    {
      "epoch": 2.5751715533374924,
      "grad_norm": 0.11633281472117153,
      "learning_rate": 1.468830170789548e-06,
      "loss": 0.0287,
      "num_tokens": 157402317.0,
      "step": 2065
    },
    {
      "epoch": 2.576419213973799,
      "grad_norm": 0.11499276278186372,
      "learning_rate": 1.4661407591857795e-06,
      "loss": 0.0284,
      "num_tokens": 157478723.0,
      "step": 2066
    },
    {
      "epoch": 2.577666874610106,
      "grad_norm": 0.11385546589543385,
      "learning_rate": 1.4634586621880786e-06,
      "loss": 0.0291,
      "num_tokens": 157554399.0,
      "step": 2067
    },
    {
      "epoch": 2.578914535246413,
      "grad_norm": 0.11937387845295565,
      "learning_rate": 1.4607838846598959e-06,
      "loss": 0.0289,
      "num_tokens": 157629911.0,
      "step": 2068
    },
    {
      "epoch": 2.58016219588272,
      "grad_norm": 0.11544428870648846,
      "learning_rate": 1.4581164314514127e-06,
      "loss": 0.0287,
      "num_tokens": 157707624.0,
      "step": 2069
    },
    {
      "epoch": 2.581409856519027,
      "grad_norm": 0.12638135944647513,
      "learning_rate": 1.4554563073995284e-06,
      "loss": 0.0289,
      "num_tokens": 157782694.0,
      "step": 2070
    },
    {
      "epoch": 2.582657517155334,
      "grad_norm": 0.12283297777573274,
      "learning_rate": 1.452803517327852e-06,
      "loss": 0.0388,
      "num_tokens": 157859882.0,
      "step": 2071
    },
    {
      "epoch": 2.5839051777916406,
      "grad_norm": 0.11551837073593679,
      "learning_rate": 1.450158066046692e-06,
      "loss": 0.029,
      "num_tokens": 157935097.0,
      "step": 2072
    },
    {
      "epoch": 2.5851528384279474,
      "grad_norm": 0.11356655049306041,
      "learning_rate": 1.4475199583530536e-06,
      "loss": 0.0281,
      "num_tokens": 158010592.0,
      "step": 2073
    },
    {
      "epoch": 2.5864004990642546,
      "grad_norm": 0.1289783745736143,
      "learning_rate": 1.444889199030622e-06,
      "loss": 0.0284,
      "num_tokens": 158087479.0,
      "step": 2074
    },
    {
      "epoch": 2.5876481597005614,
      "grad_norm": 0.11276320187709267,
      "learning_rate": 1.4422657928497572e-06,
      "loss": 0.0283,
      "num_tokens": 158162638.0,
      "step": 2075
    },
    {
      "epoch": 2.5888958203368686,
      "grad_norm": 0.12913303698692233,
      "learning_rate": 1.4396497445674917e-06,
      "loss": 0.0291,
      "num_tokens": 158239511.0,
      "step": 2076
    },
    {
      "epoch": 2.5901434809731754,
      "grad_norm": 0.1161454313827607,
      "learning_rate": 1.4370410589275096e-06,
      "loss": 0.0291,
      "num_tokens": 158316739.0,
      "step": 2077
    },
    {
      "epoch": 2.591391141609482,
      "grad_norm": 0.12022429251512276,
      "learning_rate": 1.4344397406601454e-06,
      "loss": 0.0295,
      "num_tokens": 158392820.0,
      "step": 2078
    },
    {
      "epoch": 2.592638802245789,
      "grad_norm": 0.12182801317509814,
      "learning_rate": 1.4318457944823775e-06,
      "loss": 0.0289,
      "num_tokens": 158469327.0,
      "step": 2079
    },
    {
      "epoch": 2.593886462882096,
      "grad_norm": 0.12531918190640393,
      "learning_rate": 1.4292592250978137e-06,
      "loss": 0.0295,
      "num_tokens": 158544876.0,
      "step": 2080
    },
    {
      "epoch": 2.595134123518403,
      "grad_norm": 0.11731397679998512,
      "learning_rate": 1.4266800371966844e-06,
      "loss": 0.0297,
      "num_tokens": 158621884.0,
      "step": 2081
    },
    {
      "epoch": 2.59638178415471,
      "grad_norm": 0.12030922746281104,
      "learning_rate": 1.424108235455838e-06,
      "loss": 0.0289,
      "num_tokens": 158697140.0,
      "step": 2082
    },
    {
      "epoch": 2.597629444791017,
      "grad_norm": 0.12591593849077176,
      "learning_rate": 1.4215438245387303e-06,
      "loss": 0.03,
      "num_tokens": 158773147.0,
      "step": 2083
    },
    {
      "epoch": 2.5988771054273236,
      "grad_norm": 0.11333367538005146,
      "learning_rate": 1.41898680909541e-06,
      "loss": 0.0289,
      "num_tokens": 158848983.0,
      "step": 2084
    },
    {
      "epoch": 2.600124766063631,
      "grad_norm": 0.12222202988686626,
      "learning_rate": 1.4164371937625222e-06,
      "loss": 0.0298,
      "num_tokens": 158924781.0,
      "step": 2085
    },
    {
      "epoch": 2.6013724266999376,
      "grad_norm": 0.13082834567148519,
      "learning_rate": 1.4138949831632879e-06,
      "loss": 0.0295,
      "num_tokens": 159001962.0,
      "step": 2086
    },
    {
      "epoch": 2.6026200873362444,
      "grad_norm": 0.11153925261065922,
      "learning_rate": 1.4113601819075037e-06,
      "loss": 0.0279,
      "num_tokens": 159076678.0,
      "step": 2087
    },
    {
      "epoch": 2.6038677479725516,
      "grad_norm": 0.12512392901329744,
      "learning_rate": 1.4088327945915315e-06,
      "loss": 0.0298,
      "num_tokens": 159153608.0,
      "step": 2088
    },
    {
      "epoch": 2.6051154086088584,
      "grad_norm": 0.11413584500515658,
      "learning_rate": 1.4063128257982867e-06,
      "loss": 0.0289,
      "num_tokens": 159229793.0,
      "step": 2089
    },
    {
      "epoch": 2.606363069245165,
      "grad_norm": 0.10591067240737526,
      "learning_rate": 1.4038002800972362e-06,
      "loss": 0.0279,
      "num_tokens": 159304561.0,
      "step": 2090
    },
    {
      "epoch": 2.6076107298814724,
      "grad_norm": 0.10787585851556851,
      "learning_rate": 1.401295162044383e-06,
      "loss": 0.0288,
      "num_tokens": 159380713.0,
      "step": 2091
    },
    {
      "epoch": 2.608858390517779,
      "grad_norm": 0.11974792801092993,
      "learning_rate": 1.3987974761822656e-06,
      "loss": 0.0285,
      "num_tokens": 159457696.0,
      "step": 2092
    },
    {
      "epoch": 2.6101060511540863,
      "grad_norm": 0.11054859699617442,
      "learning_rate": 1.3963072270399411e-06,
      "loss": 0.028,
      "num_tokens": 159534657.0,
      "step": 2093
    },
    {
      "epoch": 2.611353711790393,
      "grad_norm": 0.11316028322498743,
      "learning_rate": 1.393824419132986e-06,
      "loss": 0.0281,
      "num_tokens": 159610294.0,
      "step": 2094
    },
    {
      "epoch": 2.6126013724267,
      "grad_norm": 0.11900041193474127,
      "learning_rate": 1.3913490569634796e-06,
      "loss": 0.029,
      "num_tokens": 159686082.0,
      "step": 2095
    },
    {
      "epoch": 2.6138490330630066,
      "grad_norm": 0.1265360706615686,
      "learning_rate": 1.388881145020002e-06,
      "loss": 0.0292,
      "num_tokens": 159762063.0,
      "step": 2096
    },
    {
      "epoch": 2.615096693699314,
      "grad_norm": 0.10932555533496009,
      "learning_rate": 1.3864206877776245e-06,
      "loss": 0.0283,
      "num_tokens": 159838048.0,
      "step": 2097
    },
    {
      "epoch": 2.6163443543356206,
      "grad_norm": 0.12179756460559203,
      "learning_rate": 1.3839676896978997e-06,
      "loss": 0.0292,
      "num_tokens": 159914786.0,
      "step": 2098
    },
    {
      "epoch": 2.617592014971928,
      "grad_norm": 0.11381057818569847,
      "learning_rate": 1.3815221552288541e-06,
      "loss": 0.0286,
      "num_tokens": 159990643.0,
      "step": 2099
    },
    {
      "epoch": 2.6188396756082346,
      "grad_norm": 0.12589741192291856,
      "learning_rate": 1.3790840888049802e-06,
      "loss": 0.0288,
      "num_tokens": 160066206.0,
      "step": 2100
    },
    {
      "epoch": 2.6200873362445414,
      "grad_norm": 0.1314569319576317,
      "learning_rate": 1.3766534948472307e-06,
      "loss": 0.0292,
      "num_tokens": 160141809.0,
      "step": 2101
    },
    {
      "epoch": 2.621334996880848,
      "grad_norm": 0.11817206295694255,
      "learning_rate": 1.3742303777630057e-06,
      "loss": 0.0296,
      "num_tokens": 160218086.0,
      "step": 2102
    },
    {
      "epoch": 2.6225826575171554,
      "grad_norm": 0.12899261334921272,
      "learning_rate": 1.3718147419461497e-06,
      "loss": 0.0297,
      "num_tokens": 160293689.0,
      "step": 2103
    },
    {
      "epoch": 2.623830318153462,
      "grad_norm": 0.13229897643742755,
      "learning_rate": 1.3694065917769414e-06,
      "loss": 0.0294,
      "num_tokens": 160370524.0,
      "step": 2104
    },
    {
      "epoch": 2.6250779787897693,
      "grad_norm": 0.1150696614738625,
      "learning_rate": 1.367005931622084e-06,
      "loss": 0.0281,
      "num_tokens": 160446520.0,
      "step": 2105
    },
    {
      "epoch": 2.626325639426076,
      "grad_norm": 0.12646685678510525,
      "learning_rate": 1.3646127658346992e-06,
      "loss": 0.0292,
      "num_tokens": 160522619.0,
      "step": 2106
    },
    {
      "epoch": 2.627573300062383,
      "grad_norm": 0.12074932800009718,
      "learning_rate": 1.3622270987543215e-06,
      "loss": 0.0303,
      "num_tokens": 160597550.0,
      "step": 2107
    },
    {
      "epoch": 2.62882096069869,
      "grad_norm": 0.12404470424703581,
      "learning_rate": 1.3598489347068858e-06,
      "loss": 0.03,
      "num_tokens": 160673517.0,
      "step": 2108
    },
    {
      "epoch": 2.630068621334997,
      "grad_norm": 0.12077222330645102,
      "learning_rate": 1.357478278004721e-06,
      "loss": 0.0289,
      "num_tokens": 160749412.0,
      "step": 2109
    },
    {
      "epoch": 2.631316281971304,
      "grad_norm": 0.12197047343054625,
      "learning_rate": 1.3551151329465462e-06,
      "loss": 0.0294,
      "num_tokens": 160827091.0,
      "step": 2110
    },
    {
      "epoch": 2.632563942607611,
      "grad_norm": 0.120305318419666,
      "learning_rate": 1.3527595038174566e-06,
      "loss": 0.0288,
      "num_tokens": 160903819.0,
      "step": 2111
    },
    {
      "epoch": 2.6338116032439176,
      "grad_norm": 0.10870153637376272,
      "learning_rate": 1.35041139488892e-06,
      "loss": 0.0276,
      "num_tokens": 160979263.0,
      "step": 2112
    },
    {
      "epoch": 2.6350592638802244,
      "grad_norm": 0.11925953541163148,
      "learning_rate": 1.3480708104187685e-06,
      "loss": 0.0289,
      "num_tokens": 161055934.0,
      "step": 2113
    },
    {
      "epoch": 2.6363069245165316,
      "grad_norm": 0.11768773201711762,
      "learning_rate": 1.3457377546511882e-06,
      "loss": 0.0285,
      "num_tokens": 161132407.0,
      "step": 2114
    },
    {
      "epoch": 2.6375545851528384,
      "grad_norm": 0.12563149415797625,
      "learning_rate": 1.3434122318167142e-06,
      "loss": 0.0285,
      "num_tokens": 161208444.0,
      "step": 2115
    },
    {
      "epoch": 2.6388022457891456,
      "grad_norm": 0.10760676490787521,
      "learning_rate": 1.3410942461322236e-06,
      "loss": 0.0273,
      "num_tokens": 161284775.0,
      "step": 2116
    },
    {
      "epoch": 2.6400499064254523,
      "grad_norm": 0.12383103239434214,
      "learning_rate": 1.3387838018009239e-06,
      "loss": 0.0296,
      "num_tokens": 161361877.0,
      "step": 2117
    },
    {
      "epoch": 2.641297567061759,
      "grad_norm": 0.12220346777866166,
      "learning_rate": 1.3364809030123477e-06,
      "loss": 0.0279,
      "num_tokens": 161438058.0,
      "step": 2118
    },
    {
      "epoch": 2.642545227698066,
      "grad_norm": 0.11878003052432862,
      "learning_rate": 1.3341855539423499e-06,
      "loss": 0.0298,
      "num_tokens": 161514402.0,
      "step": 2119
    },
    {
      "epoch": 2.643792888334373,
      "grad_norm": 0.1173199590403945,
      "learning_rate": 1.3318977587530907e-06,
      "loss": 0.0287,
      "num_tokens": 161590055.0,
      "step": 2120
    },
    {
      "epoch": 2.64504054897068,
      "grad_norm": 0.1246361351224797,
      "learning_rate": 1.3296175215930326e-06,
      "loss": 0.0287,
      "num_tokens": 161667736.0,
      "step": 2121
    },
    {
      "epoch": 2.646288209606987,
      "grad_norm": 0.11285791824953058,
      "learning_rate": 1.3273448465969376e-06,
      "loss": 0.0279,
      "num_tokens": 161742212.0,
      "step": 2122
    },
    {
      "epoch": 2.647535870243294,
      "grad_norm": 0.11252597487092157,
      "learning_rate": 1.3250797378858507e-06,
      "loss": 0.0286,
      "num_tokens": 161816984.0,
      "step": 2123
    },
    {
      "epoch": 2.6487835308796006,
      "grad_norm": 0.11263456395387882,
      "learning_rate": 1.3228221995670987e-06,
      "loss": 0.0286,
      "num_tokens": 161893337.0,
      "step": 2124
    },
    {
      "epoch": 2.650031191515908,
      "grad_norm": 0.12248242509666539,
      "learning_rate": 1.3205722357342807e-06,
      "loss": 0.0293,
      "num_tokens": 161968441.0,
      "step": 2125
    },
    {
      "epoch": 2.6512788521522146,
      "grad_norm": 0.10926169879500301,
      "learning_rate": 1.3183298504672626e-06,
      "loss": 0.0281,
      "num_tokens": 162044146.0,
      "step": 2126
    },
    {
      "epoch": 2.6525265127885214,
      "grad_norm": 0.11384117282592209,
      "learning_rate": 1.316095047832166e-06,
      "loss": 0.0286,
      "num_tokens": 162120564.0,
      "step": 2127
    },
    {
      "epoch": 2.6537741734248286,
      "grad_norm": 0.11378643996598749,
      "learning_rate": 1.3138678318813618e-06,
      "loss": 0.0281,
      "num_tokens": 162196708.0,
      "step": 2128
    },
    {
      "epoch": 2.6550218340611353,
      "grad_norm": 0.11191194350156596,
      "learning_rate": 1.3116482066534686e-06,
      "loss": 0.0284,
      "num_tokens": 162273273.0,
      "step": 2129
    },
    {
      "epoch": 2.656269494697442,
      "grad_norm": 0.1157012495472187,
      "learning_rate": 1.3094361761733356e-06,
      "loss": 0.0287,
      "num_tokens": 162349729.0,
      "step": 2130
    },
    {
      "epoch": 2.6575171553337493,
      "grad_norm": 0.12725034795133214,
      "learning_rate": 1.3072317444520449e-06,
      "loss": 0.03,
      "num_tokens": 162425187.0,
      "step": 2131
    },
    {
      "epoch": 2.658764815970056,
      "grad_norm": 0.12362640477139697,
      "learning_rate": 1.3050349154868946e-06,
      "loss": 0.0293,
      "num_tokens": 162501517.0,
      "step": 2132
    },
    {
      "epoch": 2.6600124766063633,
      "grad_norm": 0.11027867097576076,
      "learning_rate": 1.3028456932614019e-06,
      "loss": 0.028,
      "num_tokens": 162576306.0,
      "step": 2133
    },
    {
      "epoch": 2.66126013724267,
      "grad_norm": 0.1239816165697189,
      "learning_rate": 1.3006640817452873e-06,
      "loss": 0.0294,
      "num_tokens": 162652236.0,
      "step": 2134
    },
    {
      "epoch": 2.662507797878977,
      "grad_norm": 0.12208914364343404,
      "learning_rate": 1.2984900848944727e-06,
      "loss": 0.0286,
      "num_tokens": 162727601.0,
      "step": 2135
    },
    {
      "epoch": 2.6637554585152836,
      "grad_norm": 0.13101940016702904,
      "learning_rate": 1.2963237066510715e-06,
      "loss": 0.0307,
      "num_tokens": 162803461.0,
      "step": 2136
    },
    {
      "epoch": 2.665003119151591,
      "grad_norm": 0.11919262059257354,
      "learning_rate": 1.2941649509433808e-06,
      "loss": 0.0294,
      "num_tokens": 162879944.0,
      "step": 2137
    },
    {
      "epoch": 2.6662507797878976,
      "grad_norm": 0.11810995954153358,
      "learning_rate": 1.2920138216858791e-06,
      "loss": 0.0291,
      "num_tokens": 162955901.0,
      "step": 2138
    },
    {
      "epoch": 2.667498440424205,
      "grad_norm": 0.14155939684520275,
      "learning_rate": 1.289870322779212e-06,
      "loss": 0.0305,
      "num_tokens": 163032700.0,
      "step": 2139
    },
    {
      "epoch": 2.6687461010605116,
      "grad_norm": 0.11500433107198005,
      "learning_rate": 1.2877344581101922e-06,
      "loss": 0.0276,
      "num_tokens": 163108621.0,
      "step": 2140
    },
    {
      "epoch": 2.6699937616968183,
      "grad_norm": 0.11864582024908449,
      "learning_rate": 1.2856062315517885e-06,
      "loss": 0.0293,
      "num_tokens": 163186508.0,
      "step": 2141
    },
    {
      "epoch": 2.671241422333125,
      "grad_norm": 0.11900223617148727,
      "learning_rate": 1.2834856469631174e-06,
      "loss": 0.0288,
      "num_tokens": 163263228.0,
      "step": 2142
    },
    {
      "epoch": 2.6724890829694323,
      "grad_norm": 0.11641318239763454,
      "learning_rate": 1.28137270818944e-06,
      "loss": 0.0288,
      "num_tokens": 163339630.0,
      "step": 2143
    },
    {
      "epoch": 2.673736743605739,
      "grad_norm": 0.12490145643178303,
      "learning_rate": 1.279267419062155e-06,
      "loss": 0.028,
      "num_tokens": 163416071.0,
      "step": 2144
    },
    {
      "epoch": 2.6749844042420463,
      "grad_norm": 0.11237312172406205,
      "learning_rate": 1.2771697833987852e-06,
      "loss": 0.0287,
      "num_tokens": 163491679.0,
      "step": 2145
    },
    {
      "epoch": 2.676232064878353,
      "grad_norm": 0.12175369762015016,
      "learning_rate": 1.2750798050029782e-06,
      "loss": 0.0299,
      "num_tokens": 163567232.0,
      "step": 2146
    },
    {
      "epoch": 2.67747972551466,
      "grad_norm": 0.12367303304199498,
      "learning_rate": 1.272997487664499e-06,
      "loss": 0.0288,
      "num_tokens": 163645463.0,
      "step": 2147
    },
    {
      "epoch": 2.678727386150967,
      "grad_norm": 0.1404182279844831,
      "learning_rate": 1.2709228351592167e-06,
      "loss": 0.0308,
      "num_tokens": 163722602.0,
      "step": 2148
    },
    {
      "epoch": 2.679975046787274,
      "grad_norm": 0.11022690433972186,
      "learning_rate": 1.2688558512491032e-06,
      "loss": 0.0279,
      "num_tokens": 163798334.0,
      "step": 2149
    },
    {
      "epoch": 2.681222707423581,
      "grad_norm": 0.12539838289586516,
      "learning_rate": 1.2667965396822257e-06,
      "loss": 0.0308,
      "num_tokens": 163875207.0,
      "step": 2150
    },
    {
      "epoch": 2.682470368059888,
      "grad_norm": 0.11620046852726527,
      "learning_rate": 1.2647449041927385e-06,
      "loss": 0.0286,
      "num_tokens": 163950395.0,
      "step": 2151
    },
    {
      "epoch": 2.6837180286961946,
      "grad_norm": 0.11441802870799525,
      "learning_rate": 1.2627009485008754e-06,
      "loss": 0.0288,
      "num_tokens": 164025874.0,
      "step": 2152
    },
    {
      "epoch": 2.6849656893325013,
      "grad_norm": 0.12686677130328586,
      "learning_rate": 1.2606646763129476e-06,
      "loss": 0.0283,
      "num_tokens": 164102560.0,
      "step": 2153
    },
    {
      "epoch": 2.6862133499688086,
      "grad_norm": 0.11642503259858736,
      "learning_rate": 1.2586360913213315e-06,
      "loss": 0.0286,
      "num_tokens": 164178621.0,
      "step": 2154
    },
    {
      "epoch": 2.6874610106051153,
      "grad_norm": 0.123454048742059,
      "learning_rate": 1.256615197204465e-06,
      "loss": 0.0295,
      "num_tokens": 164257244.0,
      "step": 2155
    },
    {
      "epoch": 2.6887086712414225,
      "grad_norm": 0.12546748989789816,
      "learning_rate": 1.2546019976268403e-06,
      "loss": 0.03,
      "num_tokens": 164333838.0,
      "step": 2156
    },
    {
      "epoch": 2.6899563318777293,
      "grad_norm": 0.11383031726636761,
      "learning_rate": 1.2525964962389961e-06,
      "loss": 0.0281,
      "num_tokens": 164410088.0,
      "step": 2157
    },
    {
      "epoch": 2.691203992514036,
      "grad_norm": 0.11824387324543177,
      "learning_rate": 1.250598696677512e-06,
      "loss": 0.029,
      "num_tokens": 164487052.0,
      "step": 2158
    },
    {
      "epoch": 2.692451653150343,
      "grad_norm": 0.10958624171174712,
      "learning_rate": 1.2486086025650045e-06,
      "loss": 0.0276,
      "num_tokens": 164563947.0,
      "step": 2159
    },
    {
      "epoch": 2.69369931378665,
      "grad_norm": 0.1261007581520669,
      "learning_rate": 1.246626217510114e-06,
      "loss": 0.0292,
      "num_tokens": 164641006.0,
      "step": 2160
    },
    {
      "epoch": 2.694946974422957,
      "grad_norm": 0.11602850342843693,
      "learning_rate": 1.244651545107503e-06,
      "loss": 0.0285,
      "num_tokens": 164716872.0,
      "step": 2161
    },
    {
      "epoch": 2.696194635059264,
      "grad_norm": 0.12345956113398639,
      "learning_rate": 1.2426845889378516e-06,
      "loss": 0.0284,
      "num_tokens": 164792532.0,
      "step": 2162
    },
    {
      "epoch": 2.697442295695571,
      "grad_norm": 0.1134742454092433,
      "learning_rate": 1.2407253525678453e-06,
      "loss": 0.0288,
      "num_tokens": 164868337.0,
      "step": 2163
    },
    {
      "epoch": 2.6986899563318776,
      "grad_norm": 0.11240379817612962,
      "learning_rate": 1.2387738395501714e-06,
      "loss": 0.0286,
      "num_tokens": 164944340.0,
      "step": 2164
    },
    {
      "epoch": 2.699937616968185,
      "grad_norm": 0.11581388826993168,
      "learning_rate": 1.236830053423512e-06,
      "loss": 0.0283,
      "num_tokens": 165023931.0,
      "step": 2165
    },
    {
      "epoch": 2.7011852776044916,
      "grad_norm": 0.11453442673809472,
      "learning_rate": 1.2348939977125412e-06,
      "loss": 0.0279,
      "num_tokens": 165103314.0,
      "step": 2166
    },
    {
      "epoch": 2.7024329382407988,
      "grad_norm": 0.11155255093180133,
      "learning_rate": 1.2329656759279108e-06,
      "loss": 0.0287,
      "num_tokens": 165179114.0,
      "step": 2167
    },
    {
      "epoch": 2.7036805988771055,
      "grad_norm": 0.11330932351293212,
      "learning_rate": 1.2310450915662516e-06,
      "loss": 0.0279,
      "num_tokens": 165254021.0,
      "step": 2168
    },
    {
      "epoch": 2.7049282595134123,
      "grad_norm": 0.12186227732333973,
      "learning_rate": 1.229132248110165e-06,
      "loss": 0.0292,
      "num_tokens": 165330407.0,
      "step": 2169
    },
    {
      "epoch": 2.706175920149719,
      "grad_norm": 0.11674335927535405,
      "learning_rate": 1.2272271490282134e-06,
      "loss": 0.0287,
      "num_tokens": 165406621.0,
      "step": 2170
    },
    {
      "epoch": 2.7074235807860263,
      "grad_norm": 0.11965473757061497,
      "learning_rate": 1.2253297977749163e-06,
      "loss": 0.0276,
      "num_tokens": 165482134.0,
      "step": 2171
    },
    {
      "epoch": 2.708671241422333,
      "grad_norm": 0.11983758170547645,
      "learning_rate": 1.2234401977907468e-06,
      "loss": 0.0286,
      "num_tokens": 165557987.0,
      "step": 2172
    },
    {
      "epoch": 2.7099189020586403,
      "grad_norm": 0.11471324375894294,
      "learning_rate": 1.2215583525021203e-06,
      "loss": 0.0281,
      "num_tokens": 165634677.0,
      "step": 2173
    },
    {
      "epoch": 2.711166562694947,
      "grad_norm": 0.11355220218054589,
      "learning_rate": 1.2196842653213896e-06,
      "loss": 0.028,
      "num_tokens": 165710052.0,
      "step": 2174
    },
    {
      "epoch": 2.712414223331254,
      "grad_norm": 0.11767285515107533,
      "learning_rate": 1.2178179396468428e-06,
      "loss": 0.0278,
      "num_tokens": 165786084.0,
      "step": 2175
    },
    {
      "epoch": 2.7136618839675606,
      "grad_norm": 0.12160385618810454,
      "learning_rate": 1.215959378862692e-06,
      "loss": 0.0288,
      "num_tokens": 165863041.0,
      "step": 2176
    },
    {
      "epoch": 2.714909544603868,
      "grad_norm": 0.11461675756327239,
      "learning_rate": 1.2141085863390696e-06,
      "loss": 0.0286,
      "num_tokens": 165938910.0,
      "step": 2177
    },
    {
      "epoch": 2.7161572052401746,
      "grad_norm": 0.11525113632750776,
      "learning_rate": 1.2122655654320225e-06,
      "loss": 0.0285,
      "num_tokens": 166014302.0,
      "step": 2178
    },
    {
      "epoch": 2.717404865876482,
      "grad_norm": 0.11491608215364695,
      "learning_rate": 1.210430319483504e-06,
      "loss": 0.0277,
      "num_tokens": 166089644.0,
      "step": 2179
    },
    {
      "epoch": 2.7186525265127885,
      "grad_norm": 0.11524119513899224,
      "learning_rate": 1.2086028518213694e-06,
      "loss": 0.0282,
      "num_tokens": 166166298.0,
      "step": 2180
    },
    {
      "epoch": 2.7199001871490953,
      "grad_norm": 0.1286038266919602,
      "learning_rate": 1.206783165759371e-06,
      "loss": 0.03,
      "num_tokens": 166242302.0,
      "step": 2181
    },
    {
      "epoch": 2.7211478477854025,
      "grad_norm": 0.1285603705106346,
      "learning_rate": 1.204971264597148e-06,
      "loss": 0.0287,
      "num_tokens": 166320367.0,
      "step": 2182
    },
    {
      "epoch": 2.7223955084217093,
      "grad_norm": 0.12149048372392342,
      "learning_rate": 1.2031671516202263e-06,
      "loss": 0.0288,
      "num_tokens": 166396980.0,
      "step": 2183
    },
    {
      "epoch": 2.723643169058016,
      "grad_norm": 0.11689989628810289,
      "learning_rate": 1.2013708301000082e-06,
      "loss": 0.0282,
      "num_tokens": 166472334.0,
      "step": 2184
    },
    {
      "epoch": 2.7248908296943233,
      "grad_norm": 0.1296343978344365,
      "learning_rate": 1.199582303293767e-06,
      "loss": 0.03,
      "num_tokens": 166548633.0,
      "step": 2185
    },
    {
      "epoch": 2.72613849033063,
      "grad_norm": 0.11201979473864716,
      "learning_rate": 1.1978015744446417e-06,
      "loss": 0.0276,
      "num_tokens": 166624211.0,
      "step": 2186
    },
    {
      "epoch": 2.727386150966937,
      "grad_norm": 0.11959368808418909,
      "learning_rate": 1.1960286467816331e-06,
      "loss": 0.0294,
      "num_tokens": 166701131.0,
      "step": 2187
    },
    {
      "epoch": 2.728633811603244,
      "grad_norm": 0.1150404963886038,
      "learning_rate": 1.1942635235195949e-06,
      "loss": 0.028,
      "num_tokens": 166780144.0,
      "step": 2188
    },
    {
      "epoch": 2.729881472239551,
      "grad_norm": 0.11825528418985107,
      "learning_rate": 1.1925062078592279e-06,
      "loss": 0.0296,
      "num_tokens": 166856545.0,
      "step": 2189
    },
    {
      "epoch": 2.731129132875858,
      "grad_norm": 0.11280737921120523,
      "learning_rate": 1.190756702987077e-06,
      "loss": 0.0279,
      "num_tokens": 166932621.0,
      "step": 2190
    },
    {
      "epoch": 2.732376793512165,
      "grad_norm": 0.119452906940112,
      "learning_rate": 1.1890150120755244e-06,
      "loss": 0.0293,
      "num_tokens": 167009151.0,
      "step": 2191
    },
    {
      "epoch": 2.7336244541484715,
      "grad_norm": 0.14615475941269104,
      "learning_rate": 1.1872811382827811e-06,
      "loss": 0.03,
      "num_tokens": 167085945.0,
      "step": 2192
    },
    {
      "epoch": 2.7348721147847783,
      "grad_norm": 0.12157097414483191,
      "learning_rate": 1.1855550847528849e-06,
      "loss": 0.0278,
      "num_tokens": 167162551.0,
      "step": 2193
    },
    {
      "epoch": 2.7361197754210855,
      "grad_norm": 0.11817467596856256,
      "learning_rate": 1.1838368546156924e-06,
      "loss": 0.0295,
      "num_tokens": 167238923.0,
      "step": 2194
    },
    {
      "epoch": 2.7373674360573923,
      "grad_norm": 0.11504894992061995,
      "learning_rate": 1.182126450986874e-06,
      "loss": 0.0282,
      "num_tokens": 167314732.0,
      "step": 2195
    },
    {
      "epoch": 2.7386150966936995,
      "grad_norm": 0.12495964859507203,
      "learning_rate": 1.1804238769679077e-06,
      "loss": 0.0299,
      "num_tokens": 167391511.0,
      "step": 2196
    },
    {
      "epoch": 2.7398627573300063,
      "grad_norm": 0.11246368198397648,
      "learning_rate": 1.178729135646077e-06,
      "loss": 0.0278,
      "num_tokens": 167467831.0,
      "step": 2197
    },
    {
      "epoch": 2.741110417966313,
      "grad_norm": 0.12287539336021341,
      "learning_rate": 1.1770422300944586e-06,
      "loss": 0.028,
      "num_tokens": 167543169.0,
      "step": 2198
    },
    {
      "epoch": 2.74235807860262,
      "grad_norm": 0.1196178502547976,
      "learning_rate": 1.1753631633719217e-06,
      "loss": 0.0284,
      "num_tokens": 167618234.0,
      "step": 2199
    },
    {
      "epoch": 2.743605739238927,
      "grad_norm": 0.12728052422070063,
      "learning_rate": 1.1736919385231236e-06,
      "loss": 0.0298,
      "num_tokens": 167694674.0,
      "step": 2200
    },
    {
      "epoch": 2.744853399875234,
      "grad_norm": 0.11082580587324581,
      "learning_rate": 1.1720285585784983e-06,
      "loss": 0.0284,
      "num_tokens": 167770195.0,
      "step": 2201
    },
    {
      "epoch": 2.746101060511541,
      "grad_norm": 0.1126882445851649,
      "learning_rate": 1.1703730265542569e-06,
      "loss": 0.0287,
      "num_tokens": 167847001.0,
      "step": 2202
    },
    {
      "epoch": 2.747348721147848,
      "grad_norm": 0.11484684620419089,
      "learning_rate": 1.16872534545238e-06,
      "loss": 0.0282,
      "num_tokens": 167923273.0,
      "step": 2203
    },
    {
      "epoch": 2.7485963817841546,
      "grad_norm": 0.12201620108822543,
      "learning_rate": 1.1670855182606106e-06,
      "loss": 0.0284,
      "num_tokens": 168000228.0,
      "step": 2204
    },
    {
      "epoch": 2.7498440424204618,
      "grad_norm": 0.11672097206864245,
      "learning_rate": 1.1654535479524511e-06,
      "loss": 0.0295,
      "num_tokens": 168076189.0,
      "step": 2205
    },
    {
      "epoch": 2.7510917030567685,
      "grad_norm": 0.10967114609254536,
      "learning_rate": 1.163829437487158e-06,
      "loss": 0.0277,
      "num_tokens": 168154549.0,
      "step": 2206
    },
    {
      "epoch": 2.7523393636930757,
      "grad_norm": 0.12032675683422693,
      "learning_rate": 1.162213189809734e-06,
      "loss": 0.0296,
      "num_tokens": 168230773.0,
      "step": 2207
    },
    {
      "epoch": 2.7535870243293825,
      "grad_norm": 0.12109856736796651,
      "learning_rate": 1.1606048078509235e-06,
      "loss": 0.0291,
      "num_tokens": 168306819.0,
      "step": 2208
    },
    {
      "epoch": 2.7548346849656893,
      "grad_norm": 0.12014982496104747,
      "learning_rate": 1.1590042945272108e-06,
      "loss": 0.0291,
      "num_tokens": 168383185.0,
      "step": 2209
    },
    {
      "epoch": 2.756082345601996,
      "grad_norm": 0.11366355466444636,
      "learning_rate": 1.1574116527408093e-06,
      "loss": 0.0286,
      "num_tokens": 168459637.0,
      "step": 2210
    },
    {
      "epoch": 2.7573300062383033,
      "grad_norm": 0.12086871113763324,
      "learning_rate": 1.1558268853796597e-06,
      "loss": 0.0298,
      "num_tokens": 168536455.0,
      "step": 2211
    },
    {
      "epoch": 2.75857766687461,
      "grad_norm": 0.1208261526681414,
      "learning_rate": 1.1542499953174257e-06,
      "loss": 0.0294,
      "num_tokens": 168612474.0,
      "step": 2212
    },
    {
      "epoch": 2.7598253275109172,
      "grad_norm": 0.12495039793683214,
      "learning_rate": 1.1526809854134844e-06,
      "loss": 0.0301,
      "num_tokens": 168688888.0,
      "step": 2213
    },
    {
      "epoch": 2.761072988147224,
      "grad_norm": 0.11023696555976253,
      "learning_rate": 1.151119858512925e-06,
      "loss": 0.0277,
      "num_tokens": 168764636.0,
      "step": 2214
    },
    {
      "epoch": 2.762320648783531,
      "grad_norm": 0.12130654220412528,
      "learning_rate": 1.149566617446543e-06,
      "loss": 0.0279,
      "num_tokens": 168840715.0,
      "step": 2215
    },
    {
      "epoch": 2.7635683094198376,
      "grad_norm": 0.11126959655043199,
      "learning_rate": 1.1480212650308337e-06,
      "loss": 0.0288,
      "num_tokens": 168916777.0,
      "step": 2216
    },
    {
      "epoch": 2.7648159700561448,
      "grad_norm": 0.12573107467674868,
      "learning_rate": 1.1464838040679876e-06,
      "loss": 0.0291,
      "num_tokens": 168992599.0,
      "step": 2217
    },
    {
      "epoch": 2.7660636306924515,
      "grad_norm": 0.11635105667815067,
      "learning_rate": 1.1449542373458867e-06,
      "loss": 0.029,
      "num_tokens": 169069076.0,
      "step": 2218
    },
    {
      "epoch": 2.7673112913287587,
      "grad_norm": 0.11224018904577808,
      "learning_rate": 1.1434325676380983e-06,
      "loss": 0.0284,
      "num_tokens": 169144728.0,
      "step": 2219
    },
    {
      "epoch": 2.7685589519650655,
      "grad_norm": 0.11548677961325926,
      "learning_rate": 1.141918797703868e-06,
      "loss": 0.0277,
      "num_tokens": 169219973.0,
      "step": 2220
    },
    {
      "epoch": 2.7698066126013723,
      "grad_norm": 0.12508992711477832,
      "learning_rate": 1.1404129302881193e-06,
      "loss": 0.029,
      "num_tokens": 169296490.0,
      "step": 2221
    },
    {
      "epoch": 2.7710542732376795,
      "grad_norm": 0.12294871269999798,
      "learning_rate": 1.1389149681214456e-06,
      "loss": 0.0286,
      "num_tokens": 169372081.0,
      "step": 2222
    },
    {
      "epoch": 2.7723019338739863,
      "grad_norm": 0.1240743594501783,
      "learning_rate": 1.1374249139201035e-06,
      "loss": 0.0293,
      "num_tokens": 169448879.0,
      "step": 2223
    },
    {
      "epoch": 2.773549594510293,
      "grad_norm": 0.1189641989925224,
      "learning_rate": 1.135942770386013e-06,
      "loss": 0.0276,
      "num_tokens": 169525370.0,
      "step": 2224
    },
    {
      "epoch": 2.7747972551466002,
      "grad_norm": 0.12031710218780445,
      "learning_rate": 1.1344685402067475e-06,
      "loss": 0.0299,
      "num_tokens": 169601681.0,
      "step": 2225
    },
    {
      "epoch": 2.776044915782907,
      "grad_norm": 0.1145940404852384,
      "learning_rate": 1.1330022260555321e-06,
      "loss": 0.028,
      "num_tokens": 169677327.0,
      "step": 2226
    },
    {
      "epoch": 2.777292576419214,
      "grad_norm": 0.12036981134601163,
      "learning_rate": 1.1315438305912377e-06,
      "loss": 0.0287,
      "num_tokens": 169752650.0,
      "step": 2227
    },
    {
      "epoch": 2.778540237055521,
      "grad_norm": 0.1232602059747911,
      "learning_rate": 1.1300933564583764e-06,
      "loss": 0.028,
      "num_tokens": 169828354.0,
      "step": 2228
    },
    {
      "epoch": 2.7797878976918278,
      "grad_norm": 0.1160036562852578,
      "learning_rate": 1.1286508062870952e-06,
      "loss": 0.0284,
      "num_tokens": 169903693.0,
      "step": 2229
    },
    {
      "epoch": 2.781035558328135,
      "grad_norm": 0.11503613474247552,
      "learning_rate": 1.1272161826931745e-06,
      "loss": 0.0285,
      "num_tokens": 169978944.0,
      "step": 2230
    },
    {
      "epoch": 2.7822832189644418,
      "grad_norm": 0.11815421947653913,
      "learning_rate": 1.1257894882780206e-06,
      "loss": 0.0284,
      "num_tokens": 170054946.0,
      "step": 2231
    },
    {
      "epoch": 2.7835308796007485,
      "grad_norm": 0.118767676640114,
      "learning_rate": 1.1243707256286606e-06,
      "loss": 0.0284,
      "num_tokens": 170131746.0,
      "step": 2232
    },
    {
      "epoch": 2.7847785402370553,
      "grad_norm": 0.11476662209958129,
      "learning_rate": 1.1229598973177407e-06,
      "loss": 0.0286,
      "num_tokens": 170207963.0,
      "step": 2233
    },
    {
      "epoch": 2.7860262008733625,
      "grad_norm": 0.11806306817521028,
      "learning_rate": 1.1215570059035199e-06,
      "loss": 0.0293,
      "num_tokens": 170283230.0,
      "step": 2234
    },
    {
      "epoch": 2.7872738615096693,
      "grad_norm": 0.10882246659086171,
      "learning_rate": 1.1201620539298636e-06,
      "loss": 0.0276,
      "num_tokens": 170359684.0,
      "step": 2235
    },
    {
      "epoch": 2.7885215221459765,
      "grad_norm": 0.11492211796238391,
      "learning_rate": 1.1187750439262405e-06,
      "loss": 0.028,
      "num_tokens": 170435000.0,
      "step": 2236
    },
    {
      "epoch": 2.7897691827822833,
      "grad_norm": 0.1282203899930841,
      "learning_rate": 1.1173959784077207e-06,
      "loss": 0.0302,
      "num_tokens": 170511425.0,
      "step": 2237
    },
    {
      "epoch": 2.79101684341859,
      "grad_norm": 0.11459020963724369,
      "learning_rate": 1.1160248598749652e-06,
      "loss": 0.0287,
      "num_tokens": 170588124.0,
      "step": 2238
    },
    {
      "epoch": 2.7922645040548972,
      "grad_norm": 0.118219825937956,
      "learning_rate": 1.114661690814227e-06,
      "loss": 0.0286,
      "num_tokens": 170664662.0,
      "step": 2239
    },
    {
      "epoch": 2.793512164691204,
      "grad_norm": 0.12474399789247946,
      "learning_rate": 1.1133064736973443e-06,
      "loss": 0.0286,
      "num_tokens": 170741642.0,
      "step": 2240
    },
    {
      "epoch": 2.7947598253275108,
      "grad_norm": 0.11978611303131881,
      "learning_rate": 1.1119592109817346e-06,
      "loss": 0.0291,
      "num_tokens": 170817653.0,
      "step": 2241
    },
    {
      "epoch": 2.796007485963818,
      "grad_norm": 0.11840215504882184,
      "learning_rate": 1.1106199051103922e-06,
      "loss": 0.0288,
      "num_tokens": 170892935.0,
      "step": 2242
    },
    {
      "epoch": 2.7972551466001248,
      "grad_norm": 0.11623909659922183,
      "learning_rate": 1.109288558511884e-06,
      "loss": 0.0291,
      "num_tokens": 170969000.0,
      "step": 2243
    },
    {
      "epoch": 2.7985028072364315,
      "grad_norm": 0.11698666849697116,
      "learning_rate": 1.1079651736003441e-06,
      "loss": 0.0288,
      "num_tokens": 171044723.0,
      "step": 2244
    },
    {
      "epoch": 2.7997504678727387,
      "grad_norm": 0.13606371430590922,
      "learning_rate": 1.106649752775468e-06,
      "loss": 0.0293,
      "num_tokens": 171121538.0,
      "step": 2245
    },
    {
      "epoch": 2.8009981285090455,
      "grad_norm": 0.12048429868469147,
      "learning_rate": 1.1053422984225127e-06,
      "loss": 0.0281,
      "num_tokens": 171198077.0,
      "step": 2246
    },
    {
      "epoch": 2.8022457891453527,
      "grad_norm": 0.12470269426113269,
      "learning_rate": 1.1040428129122873e-06,
      "loss": 0.0279,
      "num_tokens": 171277362.0,
      "step": 2247
    },
    {
      "epoch": 2.8034934497816595,
      "grad_norm": 0.19130209142145127,
      "learning_rate": 1.102751298601152e-06,
      "loss": 0.0296,
      "num_tokens": 171354521.0,
      "step": 2248
    },
    {
      "epoch": 2.8047411104179663,
      "grad_norm": 0.12456455451583687,
      "learning_rate": 1.1014677578310128e-06,
      "loss": 0.029,
      "num_tokens": 171432360.0,
      "step": 2249
    },
    {
      "epoch": 2.805988771054273,
      "grad_norm": 0.11699408718174314,
      "learning_rate": 1.1001921929293172e-06,
      "loss": 0.0283,
      "num_tokens": 171507814.0,
      "step": 2250
    },
    {
      "epoch": 2.8072364316905802,
      "grad_norm": 0.12226521217229692,
      "learning_rate": 1.0989246062090495e-06,
      "loss": 0.0286,
      "num_tokens": 171585169.0,
      "step": 2251
    },
    {
      "epoch": 2.808484092326887,
      "grad_norm": 0.11991178229904237,
      "learning_rate": 1.0976649999687282e-06,
      "loss": 0.0296,
      "num_tokens": 171661276.0,
      "step": 2252
    },
    {
      "epoch": 2.809731752963194,
      "grad_norm": 0.10974299356259266,
      "learning_rate": 1.096413376492399e-06,
      "loss": 0.0281,
      "num_tokens": 171736965.0,
      "step": 2253
    },
    {
      "epoch": 2.810979413599501,
      "grad_norm": 0.12052441591094999,
      "learning_rate": 1.0951697380496343e-06,
      "loss": 0.0287,
      "num_tokens": 171812257.0,
      "step": 2254
    },
    {
      "epoch": 2.8122270742358078,
      "grad_norm": 0.11085207515729092,
      "learning_rate": 1.093934086895526e-06,
      "loss": 0.0271,
      "num_tokens": 171888869.0,
      "step": 2255
    },
    {
      "epoch": 2.8134747348721145,
      "grad_norm": 0.13794568182203568,
      "learning_rate": 1.0927064252706845e-06,
      "loss": 0.0291,
      "num_tokens": 171965014.0,
      "step": 2256
    },
    {
      "epoch": 2.8147223955084217,
      "grad_norm": 0.11756094006014182,
      "learning_rate": 1.0914867554012297e-06,
      "loss": 0.0291,
      "num_tokens": 172042712.0,
      "step": 2257
    },
    {
      "epoch": 2.8159700561447285,
      "grad_norm": 0.11949250572717551,
      "learning_rate": 1.090275079498793e-06,
      "loss": 0.0283,
      "num_tokens": 172117766.0,
      "step": 2258
    },
    {
      "epoch": 2.8172177167810357,
      "grad_norm": 0.12671245281109003,
      "learning_rate": 1.0890713997605085e-06,
      "loss": 0.0299,
      "num_tokens": 172194141.0,
      "step": 2259
    },
    {
      "epoch": 2.8184653774173425,
      "grad_norm": 0.11890427787757048,
      "learning_rate": 1.0878757183690112e-06,
      "loss": 0.0289,
      "num_tokens": 172271023.0,
      "step": 2260
    },
    {
      "epoch": 2.8197130380536493,
      "grad_norm": 0.11355197416439035,
      "learning_rate": 1.086688037492433e-06,
      "loss": 0.028,
      "num_tokens": 172346761.0,
      "step": 2261
    },
    {
      "epoch": 2.8209606986899565,
      "grad_norm": 0.1273777473876187,
      "learning_rate": 1.0855083592843985e-06,
      "loss": 0.0305,
      "num_tokens": 172422932.0,
      "step": 2262
    },
    {
      "epoch": 2.8222083593262632,
      "grad_norm": 0.12440355227618752,
      "learning_rate": 1.0843366858840209e-06,
      "loss": 0.029,
      "num_tokens": 172500459.0,
      "step": 2263
    },
    {
      "epoch": 2.8234560199625705,
      "grad_norm": 0.13976356807487542,
      "learning_rate": 1.0831730194158982e-06,
      "loss": 0.0308,
      "num_tokens": 172578112.0,
      "step": 2264
    },
    {
      "epoch": 2.824703680598877,
      "grad_norm": 0.11599206471109613,
      "learning_rate": 1.0820173619901093e-06,
      "loss": 0.0277,
      "num_tokens": 172653995.0,
      "step": 2265
    },
    {
      "epoch": 2.825951341235184,
      "grad_norm": 0.11559659458722597,
      "learning_rate": 1.08086971570221e-06,
      "loss": 0.028,
      "num_tokens": 172730223.0,
      "step": 2266
    },
    {
      "epoch": 2.8271990018714908,
      "grad_norm": 0.1195433989503934,
      "learning_rate": 1.0797300826332307e-06,
      "loss": 0.028,
      "num_tokens": 172805019.0,
      "step": 2267
    },
    {
      "epoch": 2.828446662507798,
      "grad_norm": 0.11876185697375152,
      "learning_rate": 1.07859846484967e-06,
      "loss": 0.028,
      "num_tokens": 172880699.0,
      "step": 2268
    },
    {
      "epoch": 2.8296943231441047,
      "grad_norm": 0.11684909395782095,
      "learning_rate": 1.0774748644034936e-06,
      "loss": 0.0291,
      "num_tokens": 172959158.0,
      "step": 2269
    },
    {
      "epoch": 2.830941983780412,
      "grad_norm": 0.10958538812294302,
      "learning_rate": 1.0763592833321277e-06,
      "loss": 0.0289,
      "num_tokens": 173034913.0,
      "step": 2270
    },
    {
      "epoch": 2.8321896444167187,
      "grad_norm": 0.11373588402175755,
      "learning_rate": 1.0752517236584595e-06,
      "loss": 0.0289,
      "num_tokens": 173110805.0,
      "step": 2271
    },
    {
      "epoch": 2.8334373050530255,
      "grad_norm": 0.12280476964744709,
      "learning_rate": 1.0741521873908283e-06,
      "loss": 0.0286,
      "num_tokens": 173186522.0,
      "step": 2272
    },
    {
      "epoch": 2.8346849656893323,
      "grad_norm": 0.12463575409519533,
      "learning_rate": 1.0730606765230257e-06,
      "loss": 0.0294,
      "num_tokens": 173264138.0,
      "step": 2273
    },
    {
      "epoch": 2.8359326263256395,
      "grad_norm": 0.11778463862290611,
      "learning_rate": 1.0719771930342913e-06,
      "loss": 0.0285,
      "num_tokens": 173340444.0,
      "step": 2274
    },
    {
      "epoch": 2.8371802869619462,
      "grad_norm": 0.12702722233670585,
      "learning_rate": 1.0709017388893075e-06,
      "loss": 0.0294,
      "num_tokens": 173416777.0,
      "step": 2275
    },
    {
      "epoch": 2.8384279475982535,
      "grad_norm": 0.11507472807146055,
      "learning_rate": 1.0698343160381987e-06,
      "loss": 0.0289,
      "num_tokens": 173493542.0,
      "step": 2276
    },
    {
      "epoch": 2.8396756082345602,
      "grad_norm": 0.13647975622257214,
      "learning_rate": 1.0687749264165248e-06,
      "loss": 0.0304,
      "num_tokens": 173571061.0,
      "step": 2277
    },
    {
      "epoch": 2.840923268870867,
      "grad_norm": 0.11531884608838848,
      "learning_rate": 1.067723571945279e-06,
      "loss": 0.0286,
      "num_tokens": 173650004.0,
      "step": 2278
    },
    {
      "epoch": 2.842170929507174,
      "grad_norm": 0.1243551197922493,
      "learning_rate": 1.0666802545308847e-06,
      "loss": 0.0285,
      "num_tokens": 173725813.0,
      "step": 2279
    },
    {
      "epoch": 2.843418590143481,
      "grad_norm": 0.1050977663187282,
      "learning_rate": 1.065644976065193e-06,
      "loss": 0.0285,
      "num_tokens": 173801065.0,
      "step": 2280
    },
    {
      "epoch": 2.8446662507797877,
      "grad_norm": 0.1104574422188529,
      "learning_rate": 1.0646177384254747e-06,
      "loss": 0.028,
      "num_tokens": 173875892.0,
      "step": 2281
    },
    {
      "epoch": 2.845913911416095,
      "grad_norm": 0.12767157960439646,
      "learning_rate": 1.063598543474423e-06,
      "loss": 0.0292,
      "num_tokens": 173952726.0,
      "step": 2282
    },
    {
      "epoch": 2.8471615720524017,
      "grad_norm": 0.11758328425535548,
      "learning_rate": 1.062587393060147e-06,
      "loss": 0.0298,
      "num_tokens": 174031294.0,
      "step": 2283
    },
    {
      "epoch": 2.8484092326887085,
      "grad_norm": 0.1193197581441782,
      "learning_rate": 1.0615842890161675e-06,
      "loss": 0.0298,
      "num_tokens": 174107939.0,
      "step": 2284
    },
    {
      "epoch": 2.8496568933250157,
      "grad_norm": 0.11396787069116965,
      "learning_rate": 1.0605892331614158e-06,
      "loss": 0.0286,
      "num_tokens": 174183146.0,
      "step": 2285
    },
    {
      "epoch": 2.8509045539613225,
      "grad_norm": 0.1108030173333656,
      "learning_rate": 1.0596022273002282e-06,
      "loss": 0.0278,
      "num_tokens": 174259415.0,
      "step": 2286
    },
    {
      "epoch": 2.8521522145976297,
      "grad_norm": 0.11734545931287618,
      "learning_rate": 1.0586232732223446e-06,
      "loss": 0.0291,
      "num_tokens": 174336612.0,
      "step": 2287
    },
    {
      "epoch": 2.8533998752339365,
      "grad_norm": 0.12867876195037828,
      "learning_rate": 1.0576523727029053e-06,
      "loss": 0.0289,
      "num_tokens": 174414524.0,
      "step": 2288
    },
    {
      "epoch": 2.8546475358702432,
      "grad_norm": 0.11748175909555085,
      "learning_rate": 1.0566895275024458e-06,
      "loss": 0.0283,
      "num_tokens": 174489815.0,
      "step": 2289
    },
    {
      "epoch": 2.85589519650655,
      "grad_norm": 0.11978284345118138,
      "learning_rate": 1.0557347393668966e-06,
      "loss": 0.029,
      "num_tokens": 174564919.0,
      "step": 2290
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.12034753501508996,
      "learning_rate": 1.0547880100275755e-06,
      "loss": 0.0289,
      "num_tokens": 174640337.0,
      "step": 2291
    },
    {
      "epoch": 2.858390517779164,
      "grad_norm": 0.11619971380015172,
      "learning_rate": 1.0538493412011901e-06,
      "loss": 0.0275,
      "num_tokens": 174716258.0,
      "step": 2292
    },
    {
      "epoch": 2.859638178415471,
      "grad_norm": 0.11708515456949027,
      "learning_rate": 1.0529187345898304e-06,
      "loss": 0.0282,
      "num_tokens": 174792952.0,
      "step": 2293
    },
    {
      "epoch": 2.860885839051778,
      "grad_norm": 0.1167407498714762,
      "learning_rate": 1.0519961918809675e-06,
      "loss": 0.0284,
      "num_tokens": 174869657.0,
      "step": 2294
    },
    {
      "epoch": 2.8621334996880847,
      "grad_norm": 0.11555197503669004,
      "learning_rate": 1.05108171474745e-06,
      "loss": 0.0279,
      "num_tokens": 174944457.0,
      "step": 2295
    },
    {
      "epoch": 2.8633811603243915,
      "grad_norm": 0.12153004275913126,
      "learning_rate": 1.050175304847502e-06,
      "loss": 0.0285,
      "num_tokens": 175021106.0,
      "step": 2296
    },
    {
      "epoch": 2.8646288209606987,
      "grad_norm": 0.11834069413546869,
      "learning_rate": 1.0492769638247177e-06,
      "loss": 0.0287,
      "num_tokens": 175097270.0,
      "step": 2297
    },
    {
      "epoch": 2.8658764815970055,
      "grad_norm": 0.11656588674507007,
      "learning_rate": 1.0483866933080611e-06,
      "loss": 0.0279,
      "num_tokens": 175172911.0,
      "step": 2298
    },
    {
      "epoch": 2.8671241422333127,
      "grad_norm": 0.1085038466941198,
      "learning_rate": 1.0475044949118624e-06,
      "loss": 0.0282,
      "num_tokens": 175248640.0,
      "step": 2299
    },
    {
      "epoch": 2.8683718028696195,
      "grad_norm": 0.12037669922139323,
      "learning_rate": 1.0466303702358139e-06,
      "loss": 0.0296,
      "num_tokens": 175325430.0,
      "step": 2300
    },
    {
      "epoch": 2.8696194635059262,
      "grad_norm": 0.13236897091504393,
      "learning_rate": 1.0457643208649665e-06,
      "loss": 0.0284,
      "num_tokens": 175401029.0,
      "step": 2301
    },
    {
      "epoch": 2.8708671241422334,
      "grad_norm": 0.11858987566645689,
      "learning_rate": 1.044906348369731e-06,
      "loss": 0.0294,
      "num_tokens": 175476957.0,
      "step": 2302
    },
    {
      "epoch": 2.87211478477854,
      "grad_norm": 0.1227951804625425,
      "learning_rate": 1.0440564543058703e-06,
      "loss": 0.0288,
      "num_tokens": 175552076.0,
      "step": 2303
    },
    {
      "epoch": 2.8733624454148474,
      "grad_norm": 0.12523536720297093,
      "learning_rate": 1.0432146402144986e-06,
      "loss": 0.0301,
      "num_tokens": 175628963.0,
      "step": 2304
    },
    {
      "epoch": 2.874610106051154,
      "grad_norm": 0.11875115342024792,
      "learning_rate": 1.0423809076220805e-06,
      "loss": 0.0284,
      "num_tokens": 175704868.0,
      "step": 2305
    },
    {
      "epoch": 2.875857766687461,
      "grad_norm": 0.11671628776248034,
      "learning_rate": 1.041555258040425e-06,
      "loss": 0.0287,
      "num_tokens": 175780283.0,
      "step": 2306
    },
    {
      "epoch": 2.8771054273237677,
      "grad_norm": 0.12368134811625374,
      "learning_rate": 1.0407376929666833e-06,
      "loss": 0.0284,
      "num_tokens": 175856347.0,
      "step": 2307
    },
    {
      "epoch": 2.878353087960075,
      "grad_norm": 0.1162472062590892,
      "learning_rate": 1.0399282138833488e-06,
      "loss": 0.0288,
      "num_tokens": 175932649.0,
      "step": 2308
    },
    {
      "epoch": 2.8796007485963817,
      "grad_norm": 0.11556916437765882,
      "learning_rate": 1.039126822258252e-06,
      "loss": 0.0277,
      "num_tokens": 176007896.0,
      "step": 2309
    },
    {
      "epoch": 2.880848409232689,
      "grad_norm": 0.1119609539627241,
      "learning_rate": 1.0383335195445573e-06,
      "loss": 0.0287,
      "num_tokens": 176083475.0,
      "step": 2310
    },
    {
      "epoch": 2.8820960698689957,
      "grad_norm": 0.1387995361889247,
      "learning_rate": 1.0375483071807626e-06,
      "loss": 0.0304,
      "num_tokens": 176161678.0,
      "step": 2311
    },
    {
      "epoch": 2.8833437305053025,
      "grad_norm": 0.12502477938746268,
      "learning_rate": 1.036771186590696e-06,
      "loss": 0.0283,
      "num_tokens": 176238838.0,
      "step": 2312
    },
    {
      "epoch": 2.8845913911416092,
      "grad_norm": 0.1240241231270145,
      "learning_rate": 1.0360021591835108e-06,
      "loss": 0.0289,
      "num_tokens": 176315044.0,
      "step": 2313
    },
    {
      "epoch": 2.8858390517779164,
      "grad_norm": 0.11504412374999518,
      "learning_rate": 1.0352412263536868e-06,
      "loss": 0.0282,
      "num_tokens": 176392990.0,
      "step": 2314
    },
    {
      "epoch": 2.887086712414223,
      "grad_norm": 0.1218440945357655,
      "learning_rate": 1.0344883894810257e-06,
      "loss": 0.0284,
      "num_tokens": 176468026.0,
      "step": 2315
    },
    {
      "epoch": 2.8883343730505304,
      "grad_norm": 0.11561022011101908,
      "learning_rate": 1.033743649930647e-06,
      "loss": 0.0279,
      "num_tokens": 176543576.0,
      "step": 2316
    },
    {
      "epoch": 2.889582033686837,
      "grad_norm": 0.11506182188310878,
      "learning_rate": 1.03300700905299e-06,
      "loss": 0.0285,
      "num_tokens": 176618675.0,
      "step": 2317
    },
    {
      "epoch": 2.890829694323144,
      "grad_norm": 0.10953695860741473,
      "learning_rate": 1.0322784681838062e-06,
      "loss": 0.0278,
      "num_tokens": 176695934.0,
      "step": 2318
    },
    {
      "epoch": 2.892077354959451,
      "grad_norm": 0.11379689055717356,
      "learning_rate": 1.0315580286441616e-06,
      "loss": 0.0279,
      "num_tokens": 176773662.0,
      "step": 2319
    },
    {
      "epoch": 2.893325015595758,
      "grad_norm": 0.12620578142759756,
      "learning_rate": 1.0308456917404294e-06,
      "loss": 0.0287,
      "num_tokens": 176850476.0,
      "step": 2320
    },
    {
      "epoch": 2.8945726762320647,
      "grad_norm": 0.12447892729114969,
      "learning_rate": 1.0301414587642926e-06,
      "loss": 0.0287,
      "num_tokens": 176927212.0,
      "step": 2321
    },
    {
      "epoch": 2.895820336868372,
      "grad_norm": 0.11779830546114492,
      "learning_rate": 1.029445330992738e-06,
      "loss": 0.0287,
      "num_tokens": 177003895.0,
      "step": 2322
    },
    {
      "epoch": 2.8970679975046787,
      "grad_norm": 0.11577768724484899,
      "learning_rate": 1.0287573096880566e-06,
      "loss": 0.0283,
      "num_tokens": 177079543.0,
      "step": 2323
    },
    {
      "epoch": 2.8983156581409855,
      "grad_norm": 0.12451751141007653,
      "learning_rate": 1.028077396097838e-06,
      "loss": 0.0294,
      "num_tokens": 177157185.0,
      "step": 2324
    },
    {
      "epoch": 2.8995633187772927,
      "grad_norm": 0.11716404656090428,
      "learning_rate": 1.0274055914549708e-06,
      "loss": 0.0291,
      "num_tokens": 177233249.0,
      "step": 2325
    },
    {
      "epoch": 2.9008109794135994,
      "grad_norm": 0.11057577294351982,
      "learning_rate": 1.0267418969776405e-06,
      "loss": 0.0269,
      "num_tokens": 177308957.0,
      "step": 2326
    },
    {
      "epoch": 2.9020586400499067,
      "grad_norm": 0.1264787802133383,
      "learning_rate": 1.0260863138693264e-06,
      "loss": 0.0293,
      "num_tokens": 177385062.0,
      "step": 2327
    },
    {
      "epoch": 2.9033063006862134,
      "grad_norm": 0.11358632489611671,
      "learning_rate": 1.0254388433187975e-06,
      "loss": 0.0282,
      "num_tokens": 177460873.0,
      "step": 2328
    },
    {
      "epoch": 2.90455396132252,
      "grad_norm": 0.11835170817675868,
      "learning_rate": 1.0247994865001147e-06,
      "loss": 0.0285,
      "num_tokens": 177535848.0,
      "step": 2329
    },
    {
      "epoch": 2.905801621958827,
      "grad_norm": 0.1116111761311637,
      "learning_rate": 1.0241682445726246e-06,
      "loss": 0.0283,
      "num_tokens": 177610552.0,
      "step": 2330
    },
    {
      "epoch": 2.907049282595134,
      "grad_norm": 0.11670378910446504,
      "learning_rate": 1.0235451186809596e-06,
      "loss": 0.0278,
      "num_tokens": 177685467.0,
      "step": 2331
    },
    {
      "epoch": 2.908296943231441,
      "grad_norm": 0.1148751254783169,
      "learning_rate": 1.0229301099550352e-06,
      "loss": 0.0279,
      "num_tokens": 177761525.0,
      "step": 2332
    },
    {
      "epoch": 2.909544603867748,
      "grad_norm": 0.11532511319179199,
      "learning_rate": 1.0223232195100485e-06,
      "loss": 0.0281,
      "num_tokens": 177837853.0,
      "step": 2333
    },
    {
      "epoch": 2.910792264504055,
      "grad_norm": 0.12020968010672094,
      "learning_rate": 1.0217244484464758e-06,
      "loss": 0.0278,
      "num_tokens": 177915043.0,
      "step": 2334
    },
    {
      "epoch": 2.9120399251403617,
      "grad_norm": 0.11806265091105844,
      "learning_rate": 1.0211337978500687e-06,
      "loss": 0.0286,
      "num_tokens": 177990872.0,
      "step": 2335
    },
    {
      "epoch": 2.913287585776669,
      "grad_norm": 0.11329008165553046,
      "learning_rate": 1.0205512687918558e-06,
      "loss": 0.029,
      "num_tokens": 178066454.0,
      "step": 2336
    },
    {
      "epoch": 2.9145352464129757,
      "grad_norm": 0.12620894909003305,
      "learning_rate": 1.0199768623281388e-06,
      "loss": 0.0298,
      "num_tokens": 178142963.0,
      "step": 2337
    },
    {
      "epoch": 2.9157829070492824,
      "grad_norm": 0.12495079283176895,
      "learning_rate": 1.0194105795004896e-06,
      "loss": 0.0288,
      "num_tokens": 178217723.0,
      "step": 2338
    },
    {
      "epoch": 2.9170305676855897,
      "grad_norm": 0.11862596150519701,
      "learning_rate": 1.0188524213357507e-06,
      "loss": 0.0282,
      "num_tokens": 178293553.0,
      "step": 2339
    },
    {
      "epoch": 2.9182782283218964,
      "grad_norm": 0.1341753117881708,
      "learning_rate": 1.0183023888460312e-06,
      "loss": 0.0295,
      "num_tokens": 178372300.0,
      "step": 2340
    },
    {
      "epoch": 2.919525888958203,
      "grad_norm": 0.12185364553046377,
      "learning_rate": 1.017760483028706e-06,
      "loss": 0.0284,
      "num_tokens": 178448395.0,
      "step": 2341
    },
    {
      "epoch": 2.9207735495945104,
      "grad_norm": 0.11320330905305766,
      "learning_rate": 1.017226704866415e-06,
      "loss": 0.0283,
      "num_tokens": 178524884.0,
      "step": 2342
    },
    {
      "epoch": 2.922021210230817,
      "grad_norm": 0.11032447769007697,
      "learning_rate": 1.0167010553270588e-06,
      "loss": 0.0273,
      "num_tokens": 178601402.0,
      "step": 2343
    },
    {
      "epoch": 2.9232688708671244,
      "grad_norm": 0.11664488161577292,
      "learning_rate": 1.016183535363799e-06,
      "loss": 0.0278,
      "num_tokens": 178677354.0,
      "step": 2344
    },
    {
      "epoch": 2.924516531503431,
      "grad_norm": 0.12448061115121994,
      "learning_rate": 1.0156741459150556e-06,
      "loss": 0.0288,
      "num_tokens": 178752504.0,
      "step": 2345
    },
    {
      "epoch": 2.925764192139738,
      "grad_norm": 0.12087959836076952,
      "learning_rate": 1.0151728879045057e-06,
      "loss": 0.0282,
      "num_tokens": 178828590.0,
      "step": 2346
    },
    {
      "epoch": 2.9270118527760447,
      "grad_norm": 0.12213029669986336,
      "learning_rate": 1.0146797622410813e-06,
      "loss": 0.028,
      "num_tokens": 178905502.0,
      "step": 2347
    },
    {
      "epoch": 2.928259513412352,
      "grad_norm": 0.11036904360530286,
      "learning_rate": 1.0141947698189684e-06,
      "loss": 0.0283,
      "num_tokens": 178980958.0,
      "step": 2348
    },
    {
      "epoch": 2.9295071740486587,
      "grad_norm": 0.11503129305804824,
      "learning_rate": 1.0137179115176055e-06,
      "loss": 0.0279,
      "num_tokens": 179056800.0,
      "step": 2349
    },
    {
      "epoch": 2.930754834684966,
      "grad_norm": 0.11773797518401417,
      "learning_rate": 1.0132491882016805e-06,
      "loss": 0.0284,
      "num_tokens": 179132364.0,
      "step": 2350
    },
    {
      "epoch": 2.9320024953212727,
      "grad_norm": 0.11841160723334339,
      "learning_rate": 1.0127886007211298e-06,
      "loss": 0.0301,
      "num_tokens": 179208373.0,
      "step": 2351
    },
    {
      "epoch": 2.9332501559575794,
      "grad_norm": 0.11954849849392611,
      "learning_rate": 1.0123361499111383e-06,
      "loss": 0.0292,
      "num_tokens": 179285645.0,
      "step": 2352
    },
    {
      "epoch": 2.934497816593886,
      "grad_norm": 0.11647024615398077,
      "learning_rate": 1.011891836592136e-06,
      "loss": 0.0282,
      "num_tokens": 179361853.0,
      "step": 2353
    },
    {
      "epoch": 2.9357454772301934,
      "grad_norm": 0.11857643419166924,
      "learning_rate": 1.0114556615697971e-06,
      "loss": 0.0289,
      "num_tokens": 179436883.0,
      "step": 2354
    },
    {
      "epoch": 2.9369931378665,
      "grad_norm": 0.12316371816786122,
      "learning_rate": 1.0110276256350393e-06,
      "loss": 0.0293,
      "num_tokens": 179513523.0,
      "step": 2355
    },
    {
      "epoch": 2.9382407985028074,
      "grad_norm": 0.11002325656322508,
      "learning_rate": 1.010607729564021e-06,
      "loss": 0.0271,
      "num_tokens": 179588664.0,
      "step": 2356
    },
    {
      "epoch": 2.939488459139114,
      "grad_norm": 0.13014534677930756,
      "learning_rate": 1.0101959741181396e-06,
      "loss": 0.0296,
      "num_tokens": 179665065.0,
      "step": 2357
    },
    {
      "epoch": 2.940736119775421,
      "grad_norm": 0.12006219056776087,
      "learning_rate": 1.0097923600440335e-06,
      "loss": 0.0287,
      "num_tokens": 179740516.0,
      "step": 2358
    },
    {
      "epoch": 2.941983780411728,
      "grad_norm": 0.11782985541856984,
      "learning_rate": 1.0093968880735762e-06,
      "loss": 0.0287,
      "num_tokens": 179816567.0,
      "step": 2359
    },
    {
      "epoch": 2.943231441048035,
      "grad_norm": 0.12636100120057928,
      "learning_rate": 1.009009558923878e-06,
      "loss": 0.028,
      "num_tokens": 179892466.0,
      "step": 2360
    },
    {
      "epoch": 2.944479101684342,
      "grad_norm": 0.12246295728147075,
      "learning_rate": 1.0086303732972843e-06,
      "loss": 0.029,
      "num_tokens": 179969033.0,
      "step": 2361
    },
    {
      "epoch": 2.945726762320649,
      "grad_norm": 0.12183181729491675,
      "learning_rate": 1.0082593318813728e-06,
      "loss": 0.0292,
      "num_tokens": 180044689.0,
      "step": 2362
    },
    {
      "epoch": 2.9469744229569557,
      "grad_norm": 0.11666644572098828,
      "learning_rate": 1.0078964353489536e-06,
      "loss": 0.0296,
      "num_tokens": 180121893.0,
      "step": 2363
    },
    {
      "epoch": 2.9482220835932624,
      "grad_norm": 0.12254191085037336,
      "learning_rate": 1.0075416843580687e-06,
      "loss": 0.0283,
      "num_tokens": 180197860.0,
      "step": 2364
    },
    {
      "epoch": 2.9494697442295696,
      "grad_norm": 0.12166198014054956,
      "learning_rate": 1.0071950795519873e-06,
      "loss": 0.0288,
      "num_tokens": 180273997.0,
      "step": 2365
    },
    {
      "epoch": 2.9507174048658764,
      "grad_norm": 0.12047516412080388,
      "learning_rate": 1.00685662155921e-06,
      "loss": 0.0281,
      "num_tokens": 180349865.0,
      "step": 2366
    },
    {
      "epoch": 2.9519650655021836,
      "grad_norm": 0.12922197730204377,
      "learning_rate": 1.0065263109934633e-06,
      "loss": 0.028,
      "num_tokens": 180425678.0,
      "step": 2367
    },
    {
      "epoch": 2.9532127261384904,
      "grad_norm": 0.11396094054724218,
      "learning_rate": 1.0062041484536994e-06,
      "loss": 0.0283,
      "num_tokens": 180501674.0,
      "step": 2368
    },
    {
      "epoch": 2.954460386774797,
      "grad_norm": 0.12194886207173511,
      "learning_rate": 1.0058901345240967e-06,
      "loss": 0.028,
      "num_tokens": 180578582.0,
      "step": 2369
    },
    {
      "epoch": 2.955708047411104,
      "grad_norm": 0.11173158238574944,
      "learning_rate": 1.0055842697740576e-06,
      "loss": 0.0274,
      "num_tokens": 180652999.0,
      "step": 2370
    },
    {
      "epoch": 2.956955708047411,
      "grad_norm": 0.11706697991931227,
      "learning_rate": 1.0052865547582074e-06,
      "loss": 0.0293,
      "num_tokens": 180728953.0,
      "step": 2371
    },
    {
      "epoch": 2.958203368683718,
      "grad_norm": 0.1255232811981919,
      "learning_rate": 1.004996990016393e-06,
      "loss": 0.0292,
      "num_tokens": 180805626.0,
      "step": 2372
    },
    {
      "epoch": 2.959451029320025,
      "grad_norm": 0.11392151489402788,
      "learning_rate": 1.0047155760736828e-06,
      "loss": 0.0277,
      "num_tokens": 180881984.0,
      "step": 2373
    },
    {
      "epoch": 2.960698689956332,
      "grad_norm": 0.13404422054020804,
      "learning_rate": 1.004442313440366e-06,
      "loss": 0.03,
      "num_tokens": 180957570.0,
      "step": 2374
    },
    {
      "epoch": 2.9619463505926387,
      "grad_norm": 0.11808665816651781,
      "learning_rate": 1.0041772026119493e-06,
      "loss": 0.0279,
      "num_tokens": 181033179.0,
      "step": 2375
    },
    {
      "epoch": 2.963194011228946,
      "grad_norm": 0.1241735141067898,
      "learning_rate": 1.0039202440691598e-06,
      "loss": 0.0308,
      "num_tokens": 181110216.0,
      "step": 2376
    },
    {
      "epoch": 2.9644416718652526,
      "grad_norm": 0.11937831230038913,
      "learning_rate": 1.0036714382779405e-06,
      "loss": 0.0291,
      "num_tokens": 181187152.0,
      "step": 2377
    },
    {
      "epoch": 2.9656893325015594,
      "grad_norm": 0.11692757515863732,
      "learning_rate": 1.0034307856894511e-06,
      "loss": 0.0291,
      "num_tokens": 181263915.0,
      "step": 2378
    },
    {
      "epoch": 2.9669369931378666,
      "grad_norm": 0.12227927398147019,
      "learning_rate": 1.0031982867400683e-06,
      "loss": 0.029,
      "num_tokens": 181340275.0,
      "step": 2379
    },
    {
      "epoch": 2.9681846537741734,
      "grad_norm": 0.11978900526732714,
      "learning_rate": 1.0029739418513825e-06,
      "loss": 0.0291,
      "num_tokens": 181415540.0,
      "step": 2380
    },
    {
      "epoch": 2.96943231441048,
      "grad_norm": 0.13883180958152286,
      "learning_rate": 1.0027577514301988e-06,
      "loss": 0.0291,
      "num_tokens": 181491477.0,
      "step": 2381
    },
    {
      "epoch": 2.9706799750467874,
      "grad_norm": 0.12341835004033076,
      "learning_rate": 1.002549715868536e-06,
      "loss": 0.0286,
      "num_tokens": 181566773.0,
      "step": 2382
    },
    {
      "epoch": 2.971927635683094,
      "grad_norm": 0.11368537266706243,
      "learning_rate": 1.0023498355436255e-06,
      "loss": 0.028,
      "num_tokens": 181642670.0,
      "step": 2383
    },
    {
      "epoch": 2.9731752963194014,
      "grad_norm": 0.11944682523487593,
      "learning_rate": 1.0021581108179105e-06,
      "loss": 0.0294,
      "num_tokens": 181719527.0,
      "step": 2384
    },
    {
      "epoch": 2.974422956955708,
      "grad_norm": 0.12346774894332574,
      "learning_rate": 1.0019745420390455e-06,
      "loss": 0.0289,
      "num_tokens": 181796043.0,
      "step": 2385
    },
    {
      "epoch": 2.975670617592015,
      "grad_norm": 0.12541437580408807,
      "learning_rate": 1.001799129539897e-06,
      "loss": 0.029,
      "num_tokens": 181872950.0,
      "step": 2386
    },
    {
      "epoch": 2.9769182782283217,
      "grad_norm": 0.12878850331430744,
      "learning_rate": 1.0016318736385406e-06,
      "loss": 0.0293,
      "num_tokens": 181950529.0,
      "step": 2387
    },
    {
      "epoch": 2.978165938864629,
      "grad_norm": 0.12597349481901815,
      "learning_rate": 1.0014727746382615e-06,
      "loss": 0.0303,
      "num_tokens": 182026703.0,
      "step": 2388
    },
    {
      "epoch": 2.9794135995009356,
      "grad_norm": 0.12172216490214947,
      "learning_rate": 1.0013218328275544e-06,
      "loss": 0.0298,
      "num_tokens": 182103106.0,
      "step": 2389
    },
    {
      "epoch": 2.980661260137243,
      "grad_norm": 0.11138447238599976,
      "learning_rate": 1.0011790484801231e-06,
      "loss": 0.0279,
      "num_tokens": 182178585.0,
      "step": 2390
    },
    {
      "epoch": 2.9819089207735496,
      "grad_norm": 0.10906966830793707,
      "learning_rate": 1.0010444218548777e-06,
      "loss": 0.0275,
      "num_tokens": 182253254.0,
      "step": 2391
    },
    {
      "epoch": 2.9831565814098564,
      "grad_norm": 0.12012535808427474,
      "learning_rate": 1.0009179531959374e-06,
      "loss": 0.0281,
      "num_tokens": 182329628.0,
      "step": 2392
    },
    {
      "epoch": 2.984404242046163,
      "grad_norm": 0.11447485418209809,
      "learning_rate": 1.0007996427326282e-06,
      "loss": 0.0283,
      "num_tokens": 182404805.0,
      "step": 2393
    },
    {
      "epoch": 2.9856519026824704,
      "grad_norm": 0.1180082136584759,
      "learning_rate": 1.0006894906794828e-06,
      "loss": 0.0282,
      "num_tokens": 182480786.0,
      "step": 2394
    },
    {
      "epoch": 2.986899563318777,
      "grad_norm": 0.11790449657258598,
      "learning_rate": 1.0005874972362403e-06,
      "loss": 0.0287,
      "num_tokens": 182557495.0,
      "step": 2395
    },
    {
      "epoch": 2.9881472239550844,
      "grad_norm": 0.11528577093578535,
      "learning_rate": 1.000493662587845e-06,
      "loss": 0.0279,
      "num_tokens": 182633656.0,
      "step": 2396
    },
    {
      "epoch": 2.989394884591391,
      "grad_norm": 0.12016212997807887,
      "learning_rate": 1.0004079869044482e-06,
      "loss": 0.0286,
      "num_tokens": 182709398.0,
      "step": 2397
    },
    {
      "epoch": 2.990642545227698,
      "grad_norm": 0.11852576824631264,
      "learning_rate": 1.0003304703414053e-06,
      "loss": 0.0302,
      "num_tokens": 182785596.0,
      "step": 2398
    },
    {
      "epoch": 2.991890205864005,
      "grad_norm": 0.12423687799838075,
      "learning_rate": 1.0002611130392772e-06,
      "loss": 0.0296,
      "num_tokens": 182863057.0,
      "step": 2399
    },
    {
      "epoch": 2.993137866500312,
      "grad_norm": 0.13300645679134993,
      "learning_rate": 1.0001999151238303e-06,
      "loss": 0.0283,
      "num_tokens": 182939191.0,
      "step": 2400
    },
    {
      "epoch": 2.994385527136619,
      "grad_norm": 0.12445363603171215,
      "learning_rate": 1.0001468767060341e-06,
      "loss": 0.0297,
      "num_tokens": 183014674.0,
      "step": 2401
    },
    {
      "epoch": 2.995633187772926,
      "grad_norm": 0.12331291687142039,
      "learning_rate": 1.000101997882064e-06,
      "loss": 0.0279,
      "num_tokens": 183090765.0,
      "step": 2402
    },
    {
      "epoch": 2.9968808484092326,
      "grad_norm": 0.11985250982047893,
      "learning_rate": 1.0000652787332984e-06,
      "loss": 0.0295,
      "num_tokens": 183167359.0,
      "step": 2403
    },
    {
      "epoch": 2.9981285090455394,
      "grad_norm": 0.13275772055558846,
      "learning_rate": 1.0000367193263206e-06,
      "loss": 0.0294,
      "num_tokens": 183243243.0,
      "step": 2404
    },
    {
      "epoch": 2.9993761696818466,
      "grad_norm": 0.23449268816415472,
      "learning_rate": 1.000016319712917e-06,
      "loss": 0.0481,
      "num_tokens": 183319605.0,
      "step": 2405
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.23449268816415472,
      "learning_rate": 1.0000040799300788e-06,
      "loss": 0.0262,
      "num_tokens": 183357177.0,
      "step": 2406
    },
    {
      "epoch": 3.0,
      "step": 2406,
      "total_flos": 3.555842322968822e+17,
      "train_loss": 0.05951405684243042,
      "train_runtime": 5065.1458,
      "train_samples_per_second": 60.733,
      "train_steps_per_second": 0.475
    }
  ],
  "logging_steps": 1,
  "max_steps": 2406,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.555842322968822e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}