{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 338,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002962962962962963,
      "grad_norm": 3.02152419090271,
      "learning_rate": 0.0,
      "loss": 0.7166,
      "step": 1
    },
    {
      "epoch": 0.005925925925925926,
      "grad_norm": 3.693032741546631,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.9778,
      "step": 2
    },
    {
      "epoch": 0.008888888888888889,
      "grad_norm": 2.961223840713501,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.7316,
      "step": 3
    },
    {
      "epoch": 0.011851851851851851,
      "grad_norm": 2.767284631729126,
      "learning_rate": 1.2e-05,
      "loss": 0.6977,
      "step": 4
    },
    {
      "epoch": 0.014814814814814815,
      "grad_norm": 3.268404006958008,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.8014,
      "step": 5
    },
    {
      "epoch": 0.017777777777777778,
      "grad_norm": 2.439943790435791,
      "learning_rate": 2e-05,
      "loss": 0.5953,
      "step": 6
    },
    {
      "epoch": 0.02074074074074074,
      "grad_norm": 2.730228900909424,
      "learning_rate": 1.993993993993994e-05,
      "loss": 0.56,
      "step": 7
    },
    {
      "epoch": 0.023703703703703703,
      "grad_norm": 2.1058473587036133,
      "learning_rate": 1.987987987987988e-05,
      "loss": 0.5657,
      "step": 8
    },
    {
      "epoch": 0.02666666666666667,
      "grad_norm": 2.041543483734131,
      "learning_rate": 1.981981981981982e-05,
      "loss": 0.424,
      "step": 9
    },
    {
      "epoch": 0.02962962962962963,
      "grad_norm": 1.9065091609954834,
      "learning_rate": 1.9759759759759763e-05,
      "loss": 0.54,
      "step": 10
    },
    {
      "epoch": 0.03259259259259259,
      "grad_norm": 1.5807809829711914,
      "learning_rate": 1.9699699699699702e-05,
      "loss": 0.473,
      "step": 11
    },
    {
      "epoch": 0.035555555555555556,
      "grad_norm": 1.293297290802002,
      "learning_rate": 1.963963963963964e-05,
      "loss": 0.3949,
      "step": 12
    },
    {
      "epoch": 0.03851851851851852,
      "grad_norm": 1.191159963607788,
      "learning_rate": 1.957957957957958e-05,
      "loss": 0.3329,
      "step": 13
    },
    {
      "epoch": 0.04148148148148148,
      "grad_norm": 0.9701827764511108,
      "learning_rate": 1.951951951951952e-05,
      "loss": 0.307,
      "step": 14
    },
    {
      "epoch": 0.044444444444444446,
      "grad_norm": 1.1163513660430908,
      "learning_rate": 1.9459459459459463e-05,
      "loss": 0.2306,
      "step": 15
    },
    {
      "epoch": 0.047407407407407405,
      "grad_norm": 0.8973691463470459,
      "learning_rate": 1.9399399399399402e-05,
      "loss": 0.3626,
      "step": 16
    },
    {
      "epoch": 0.05037037037037037,
      "grad_norm": 0.7620947957038879,
      "learning_rate": 1.9339339339339342e-05,
      "loss": 0.263,
      "step": 17
    },
    {
      "epoch": 0.05333333333333334,
      "grad_norm": 0.849443793296814,
      "learning_rate": 1.927927927927928e-05,
      "loss": 0.3073,
      "step": 18
    },
    {
      "epoch": 0.056296296296296296,
      "grad_norm": 0.6365483403205872,
      "learning_rate": 1.921921921921922e-05,
      "loss": 0.1658,
      "step": 19
    },
    {
      "epoch": 0.05925925925925926,
      "grad_norm": 0.6874942183494568,
      "learning_rate": 1.915915915915916e-05,
      "loss": 0.2983,
      "step": 20
    },
    {
      "epoch": 0.06222222222222222,
      "grad_norm": 0.7519598603248596,
      "learning_rate": 1.90990990990991e-05,
      "loss": 0.3136,
      "step": 21
    },
    {
      "epoch": 0.06518518518518518,
      "grad_norm": 0.5947675108909607,
      "learning_rate": 1.903903903903904e-05,
      "loss": 0.1854,
      "step": 22
    },
    {
      "epoch": 0.06814814814814815,
      "grad_norm": 0.8981500864028931,
      "learning_rate": 1.8978978978978982e-05,
      "loss": 0.3673,
      "step": 23
    },
    {
      "epoch": 0.07111111111111111,
      "grad_norm": 0.7584127187728882,
      "learning_rate": 1.891891891891892e-05,
      "loss": 0.3001,
      "step": 24
    },
    {
      "epoch": 0.07407407407407407,
      "grad_norm": 0.6769465804100037,
      "learning_rate": 1.885885885885886e-05,
      "loss": 0.2183,
      "step": 25
    },
    {
      "epoch": 0.07703703703703704,
      "grad_norm": 0.5384288430213928,
      "learning_rate": 1.87987987987988e-05,
      "loss": 0.1637,
      "step": 26
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4990951120853424,
      "learning_rate": 1.873873873873874e-05,
      "loss": 0.1335,
      "step": 27
    },
    {
      "epoch": 0.08296296296296296,
      "grad_norm": 0.9515981674194336,
      "learning_rate": 1.8678678678678682e-05,
      "loss": 0.2851,
      "step": 28
    },
    {
      "epoch": 0.08592592592592592,
      "grad_norm": 0.6648741960525513,
      "learning_rate": 1.861861861861862e-05,
      "loss": 0.1999,
      "step": 29
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 0.8580437302589417,
      "learning_rate": 1.855855855855856e-05,
      "loss": 0.1385,
      "step": 30
    },
    {
      "epoch": 0.09185185185185185,
      "grad_norm": 0.7154540419578552,
      "learning_rate": 1.84984984984985e-05,
      "loss": 0.2018,
      "step": 31
    },
    {
      "epoch": 0.09481481481481481,
      "grad_norm": 0.5420019626617432,
      "learning_rate": 1.843843843843844e-05,
      "loss": 0.1262,
      "step": 32
    },
    {
      "epoch": 0.09777777777777778,
      "grad_norm": 0.6761616468429565,
      "learning_rate": 1.8378378378378383e-05,
      "loss": 0.2362,
      "step": 33
    },
    {
      "epoch": 0.10074074074074074,
      "grad_norm": 0.7488560080528259,
      "learning_rate": 1.831831831831832e-05,
      "loss": 0.2885,
      "step": 34
    },
    {
      "epoch": 0.1037037037037037,
      "grad_norm": 0.7649799585342407,
      "learning_rate": 1.8258258258258258e-05,
      "loss": 0.2568,
      "step": 35
    },
    {
      "epoch": 0.10666666666666667,
      "grad_norm": 0.7138866186141968,
      "learning_rate": 1.81981981981982e-05,
      "loss": 0.2512,
      "step": 36
    },
    {
      "epoch": 0.10962962962962963,
      "grad_norm": 0.7375985383987427,
      "learning_rate": 1.813813813813814e-05,
      "loss": 0.269,
      "step": 37
    },
    {
      "epoch": 0.11259259259259259,
      "grad_norm": 0.6407322287559509,
      "learning_rate": 1.807807807807808e-05,
      "loss": 0.2408,
      "step": 38
    },
    {
      "epoch": 0.11555555555555555,
      "grad_norm": 0.7711364030838013,
      "learning_rate": 1.801801801801802e-05,
      "loss": 0.2005,
      "step": 39
    },
    {
      "epoch": 0.11851851851851852,
      "grad_norm": 0.577764093875885,
      "learning_rate": 1.795795795795796e-05,
      "loss": 0.1495,
      "step": 40
    },
    {
      "epoch": 0.12148148148148148,
      "grad_norm": 0.7169051766395569,
      "learning_rate": 1.78978978978979e-05,
      "loss": 0.2329,
      "step": 41
    },
    {
      "epoch": 0.12444444444444444,
      "grad_norm": 0.93023282289505,
      "learning_rate": 1.783783783783784e-05,
      "loss": 0.3046,
      "step": 42
    },
    {
      "epoch": 0.1274074074074074,
      "grad_norm": 0.7907038927078247,
      "learning_rate": 1.7777777777777777e-05,
      "loss": 0.247,
      "step": 43
    },
    {
      "epoch": 0.13037037037037036,
      "grad_norm": 0.898409366607666,
      "learning_rate": 1.771771771771772e-05,
      "loss": 0.2885,
      "step": 44
    },
    {
      "epoch": 0.13333333333333333,
      "grad_norm": 0.8418571352958679,
      "learning_rate": 1.765765765765766e-05,
      "loss": 0.2865,
      "step": 45
    },
    {
      "epoch": 0.1362962962962963,
      "grad_norm": 0.8857557773590088,
      "learning_rate": 1.7597597597597598e-05,
      "loss": 0.325,
      "step": 46
    },
    {
      "epoch": 0.13925925925925925,
      "grad_norm": 0.7279157638549805,
      "learning_rate": 1.7537537537537538e-05,
      "loss": 0.2676,
      "step": 47
    },
    {
      "epoch": 0.14222222222222222,
      "grad_norm": 0.6223716735839844,
      "learning_rate": 1.7477477477477477e-05,
      "loss": 0.2423,
      "step": 48
    },
    {
      "epoch": 0.1451851851851852,
      "grad_norm": 0.7974292635917664,
      "learning_rate": 1.741741741741742e-05,
      "loss": 0.2321,
      "step": 49
    },
    {
      "epoch": 0.14814814814814814,
      "grad_norm": 0.7039240002632141,
      "learning_rate": 1.735735735735736e-05,
      "loss": 0.1893,
      "step": 50
    },
    {
      "epoch": 0.1511111111111111,
      "grad_norm": 0.5718026161193848,
      "learning_rate": 1.72972972972973e-05,
      "loss": 0.1499,
      "step": 51
    },
    {
      "epoch": 0.15407407407407409,
      "grad_norm": 0.5464204549789429,
      "learning_rate": 1.7237237237237238e-05,
      "loss": 0.1457,
      "step": 52
    },
    {
      "epoch": 0.15703703703703703,
      "grad_norm": 0.7063807845115662,
      "learning_rate": 1.7177177177177177e-05,
      "loss": 0.2184,
      "step": 53
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6784593462944031,
      "learning_rate": 1.711711711711712e-05,
      "loss": 0.1686,
      "step": 54
    },
    {
      "epoch": 0.16296296296296298,
      "grad_norm": 0.6670235991477966,
      "learning_rate": 1.705705705705706e-05,
      "loss": 0.1804,
      "step": 55
    },
    {
      "epoch": 0.16592592592592592,
      "grad_norm": 0.795344352722168,
      "learning_rate": 1.6996996996997e-05,
      "loss": 0.2115,
      "step": 56
    },
    {
      "epoch": 0.1688888888888889,
      "grad_norm": 0.7772676944732666,
      "learning_rate": 1.693693693693694e-05,
      "loss": 0.2022,
      "step": 57
    },
    {
      "epoch": 0.17185185185185184,
      "grad_norm": 0.7115799188613892,
      "learning_rate": 1.6876876876876878e-05,
      "loss": 0.2581,
      "step": 58
    },
    {
      "epoch": 0.1748148148148148,
      "grad_norm": 0.6667796969413757,
      "learning_rate": 1.6816816816816817e-05,
      "loss": 0.1693,
      "step": 59
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 0.6113911271095276,
      "learning_rate": 1.6756756756756757e-05,
      "loss": 0.1183,
      "step": 60
    },
    {
      "epoch": 0.18074074074074073,
      "grad_norm": 0.6023927330970764,
      "learning_rate": 1.6696696696696696e-05,
      "loss": 0.1481,
      "step": 61
    },
    {
      "epoch": 0.1837037037037037,
      "grad_norm": 0.6856749653816223,
      "learning_rate": 1.663663663663664e-05,
      "loss": 0.1615,
      "step": 62
    },
    {
      "epoch": 0.18666666666666668,
      "grad_norm": 0.736131489276886,
      "learning_rate": 1.6576576576576578e-05,
      "loss": 0.2232,
      "step": 63
    },
    {
      "epoch": 0.18962962962962962,
      "grad_norm": 0.7121597528457642,
      "learning_rate": 1.6516516516516518e-05,
      "loss": 0.2168,
      "step": 64
    },
    {
      "epoch": 0.1925925925925926,
      "grad_norm": 0.6730802059173584,
      "learning_rate": 1.6456456456456457e-05,
      "loss": 0.1458,
      "step": 65
    },
    {
      "epoch": 0.19555555555555557,
      "grad_norm": 0.8634228706359863,
      "learning_rate": 1.6396396396396396e-05,
      "loss": 0.2764,
      "step": 66
    },
    {
      "epoch": 0.1985185185185185,
      "grad_norm": 0.8175811171531677,
      "learning_rate": 1.633633633633634e-05,
      "loss": 0.2129,
      "step": 67
    },
    {
      "epoch": 0.20148148148148148,
      "grad_norm": 1.0164388418197632,
      "learning_rate": 1.627627627627628e-05,
      "loss": 0.297,
      "step": 68
    },
    {
      "epoch": 0.20444444444444446,
      "grad_norm": 0.7970155477523804,
      "learning_rate": 1.6216216216216218e-05,
      "loss": 0.1955,
      "step": 69
    },
    {
      "epoch": 0.2074074074074074,
      "grad_norm": 0.6402409672737122,
      "learning_rate": 1.6156156156156157e-05,
      "loss": 0.1665,
      "step": 70
    },
    {
      "epoch": 0.21037037037037037,
      "grad_norm": 0.8887149095535278,
      "learning_rate": 1.6096096096096097e-05,
      "loss": 0.3044,
      "step": 71
    },
    {
      "epoch": 0.21333333333333335,
      "grad_norm": 0.6942129731178284,
      "learning_rate": 1.6036036036036036e-05,
      "loss": 0.1738,
      "step": 72
    },
    {
      "epoch": 0.2162962962962963,
      "grad_norm": 0.7844465374946594,
      "learning_rate": 1.5975975975975976e-05,
      "loss": 0.2142,
      "step": 73
    },
    {
      "epoch": 0.21925925925925926,
      "grad_norm": 0.8093639016151428,
      "learning_rate": 1.591591591591592e-05,
      "loss": 0.2046,
      "step": 74
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 0.6324582695960999,
      "learning_rate": 1.5855855855855858e-05,
      "loss": 0.167,
      "step": 75
    },
    {
      "epoch": 0.22518518518518518,
      "grad_norm": 0.647125780582428,
      "learning_rate": 1.5795795795795797e-05,
      "loss": 0.1265,
      "step": 76
    },
    {
      "epoch": 0.22814814814814816,
      "grad_norm": 0.8480408191680908,
      "learning_rate": 1.5735735735735737e-05,
      "loss": 0.2342,
      "step": 77
    },
    {
      "epoch": 0.2311111111111111,
      "grad_norm": 0.538595974445343,
      "learning_rate": 1.5675675675675676e-05,
      "loss": 0.1349,
      "step": 78
    },
    {
      "epoch": 0.23407407407407407,
      "grad_norm": 0.7440139055252075,
      "learning_rate": 1.5615615615615616e-05,
      "loss": 0.1563,
      "step": 79
    },
    {
      "epoch": 0.23703703703703705,
      "grad_norm": 0.7776991128921509,
      "learning_rate": 1.555555555555556e-05,
      "loss": 0.208,
      "step": 80
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.7442265152931213,
      "learning_rate": 1.5495495495495498e-05,
      "loss": 0.206,
      "step": 81
    },
    {
      "epoch": 0.24296296296296296,
      "grad_norm": 0.9572383165359497,
      "learning_rate": 1.5435435435435437e-05,
      "loss": 0.2597,
      "step": 82
    },
    {
      "epoch": 0.24592592592592594,
      "grad_norm": 0.7419414520263672,
      "learning_rate": 1.5375375375375377e-05,
      "loss": 0.2016,
      "step": 83
    },
    {
      "epoch": 0.24888888888888888,
      "grad_norm": 0.9371577501296997,
      "learning_rate": 1.5315315315315316e-05,
      "loss": 0.2064,
      "step": 84
    },
    {
      "epoch": 0.2518518518518518,
      "grad_norm": 0.6992803812026978,
      "learning_rate": 1.5255255255255257e-05,
      "loss": 0.1598,
      "step": 85
    },
    {
      "epoch": 0.2548148148148148,
      "grad_norm": 0.8519673347473145,
      "learning_rate": 1.5195195195195196e-05,
      "loss": 0.2301,
      "step": 86
    },
    {
      "epoch": 0.2577777777777778,
      "grad_norm": 0.8056367039680481,
      "learning_rate": 1.5135135135135138e-05,
      "loss": 0.2347,
      "step": 87
    },
    {
      "epoch": 0.2607407407407407,
      "grad_norm": 0.6643080115318298,
      "learning_rate": 1.5075075075075077e-05,
      "loss": 0.1442,
      "step": 88
    },
    {
      "epoch": 0.2637037037037037,
      "grad_norm": 0.7210400700569153,
      "learning_rate": 1.5015015015015015e-05,
      "loss": 0.1766,
      "step": 89
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 0.5652949810028076,
      "learning_rate": 1.4954954954954957e-05,
      "loss": 0.0852,
      "step": 90
    },
    {
      "epoch": 0.2696296296296296,
      "grad_norm": 0.7074577212333679,
      "learning_rate": 1.4894894894894895e-05,
      "loss": 0.1683,
      "step": 91
    },
    {
      "epoch": 0.2725925925925926,
      "grad_norm": 0.9635120630264282,
      "learning_rate": 1.4834834834834836e-05,
      "loss": 0.251,
      "step": 92
    },
    {
      "epoch": 0.27555555555555555,
      "grad_norm": 0.840675413608551,
      "learning_rate": 1.4774774774774776e-05,
      "loss": 0.1688,
      "step": 93
    },
    {
      "epoch": 0.2785185185185185,
      "grad_norm": 0.9204552173614502,
      "learning_rate": 1.4714714714714715e-05,
      "loss": 0.243,
      "step": 94
    },
    {
      "epoch": 0.2814814814814815,
      "grad_norm": 1.146018147468567,
      "learning_rate": 1.4654654654654656e-05,
      "loss": 0.3438,
      "step": 95
    },
    {
      "epoch": 0.28444444444444444,
      "grad_norm": 0.937786877155304,
      "learning_rate": 1.4594594594594596e-05,
      "loss": 0.2805,
      "step": 96
    },
    {
      "epoch": 0.2874074074074074,
      "grad_norm": 0.920526385307312,
      "learning_rate": 1.4534534534534537e-05,
      "loss": 0.2278,
      "step": 97
    },
    {
      "epoch": 0.2903703703703704,
      "grad_norm": 0.7850074768066406,
      "learning_rate": 1.4474474474474476e-05,
      "loss": 0.188,
      "step": 98
    },
    {
      "epoch": 0.29333333333333333,
      "grad_norm": 0.5933433175086975,
      "learning_rate": 1.4414414414414416e-05,
      "loss": 0.1563,
      "step": 99
    },
    {
      "epoch": 0.2962962962962963,
      "grad_norm": 0.8261187076568604,
      "learning_rate": 1.4354354354354357e-05,
      "loss": 0.2071,
      "step": 100
    },
    {
      "epoch": 0.2992592592592593,
      "grad_norm": 0.8072453737258911,
      "learning_rate": 1.4294294294294296e-05,
      "loss": 0.1701,
      "step": 101
    },
    {
      "epoch": 0.3022222222222222,
      "grad_norm": 0.6039605736732483,
      "learning_rate": 1.4234234234234234e-05,
      "loss": 0.1508,
      "step": 102
    },
    {
      "epoch": 0.30518518518518517,
      "grad_norm": 0.5071831345558167,
      "learning_rate": 1.4174174174174175e-05,
      "loss": 0.1036,
      "step": 103
    },
    {
      "epoch": 0.30814814814814817,
      "grad_norm": 0.8471526503562927,
      "learning_rate": 1.4114114114114114e-05,
      "loss": 0.2336,
      "step": 104
    },
    {
      "epoch": 0.3111111111111111,
      "grad_norm": 0.9204836487770081,
      "learning_rate": 1.4054054054054055e-05,
      "loss": 0.3144,
      "step": 105
    },
    {
      "epoch": 0.31407407407407406,
      "grad_norm": 0.8702786564826965,
      "learning_rate": 1.3993993993993995e-05,
      "loss": 0.236,
      "step": 106
    },
    {
      "epoch": 0.31703703703703706,
      "grad_norm": 0.6849658489227295,
      "learning_rate": 1.3933933933933934e-05,
      "loss": 0.1532,
      "step": 107
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.7833988070487976,
      "learning_rate": 1.3873873873873875e-05,
      "loss": 0.1779,
      "step": 108
    },
    {
      "epoch": 0.32296296296296295,
      "grad_norm": 0.77451092004776,
      "learning_rate": 1.3813813813813815e-05,
      "loss": 0.2169,
      "step": 109
    },
    {
      "epoch": 0.32592592592592595,
      "grad_norm": 0.6471803188323975,
      "learning_rate": 1.3753753753753756e-05,
      "loss": 0.1377,
      "step": 110
    },
    {
      "epoch": 0.3288888888888889,
      "grad_norm": 0.9164314866065979,
      "learning_rate": 1.3693693693693695e-05,
      "loss": 0.2493,
      "step": 111
    },
    {
      "epoch": 0.33185185185185184,
      "grad_norm": 0.86440110206604,
      "learning_rate": 1.3633633633633635e-05,
      "loss": 0.2164,
      "step": 112
    },
    {
      "epoch": 0.3348148148148148,
      "grad_norm": 0.6897555589675903,
      "learning_rate": 1.3573573573573576e-05,
      "loss": 0.1398,
      "step": 113
    },
    {
      "epoch": 0.3377777777777778,
      "grad_norm": 0.9025009870529175,
      "learning_rate": 1.3513513513513515e-05,
      "loss": 0.2363,
      "step": 114
    },
    {
      "epoch": 0.34074074074074073,
      "grad_norm": 1.0257855653762817,
      "learning_rate": 1.3453453453453456e-05,
      "loss": 0.2685,
      "step": 115
    },
    {
      "epoch": 0.3437037037037037,
      "grad_norm": 0.959772527217865,
      "learning_rate": 1.3393393393393394e-05,
      "loss": 0.2709,
      "step": 116
    },
    {
      "epoch": 0.3466666666666667,
      "grad_norm": 0.9299291372299194,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.2323,
      "step": 117
    },
    {
      "epoch": 0.3496296296296296,
      "grad_norm": 0.491132527589798,
      "learning_rate": 1.3273273273273274e-05,
      "loss": 0.1241,
      "step": 118
    },
    {
      "epoch": 0.35259259259259257,
      "grad_norm": 0.7121463418006897,
      "learning_rate": 1.3213213213213214e-05,
      "loss": 0.1565,
      "step": 119
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 0.7814395427703857,
      "learning_rate": 1.3153153153153155e-05,
      "loss": 0.2104,
      "step": 120
    },
    {
      "epoch": 0.3585185185185185,
      "grad_norm": 0.9731724858283997,
      "learning_rate": 1.3093093093093094e-05,
      "loss": 0.235,
      "step": 121
    },
    {
      "epoch": 0.36148148148148146,
      "grad_norm": 0.7757167220115662,
      "learning_rate": 1.3033033033033034e-05,
      "loss": 0.219,
      "step": 122
    },
    {
      "epoch": 0.36444444444444446,
      "grad_norm": 0.7114177942276001,
      "learning_rate": 1.2972972972972975e-05,
      "loss": 0.1382,
      "step": 123
    },
    {
      "epoch": 0.3674074074074074,
      "grad_norm": 0.7851217985153198,
      "learning_rate": 1.2912912912912914e-05,
      "loss": 0.2089,
      "step": 124
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 0.960207462310791,
      "learning_rate": 1.2852852852852854e-05,
      "loss": 0.2355,
      "step": 125
    },
    {
      "epoch": 0.37333333333333335,
      "grad_norm": 0.6681966185569763,
      "learning_rate": 1.2792792792792795e-05,
      "loss": 0.1372,
      "step": 126
    },
    {
      "epoch": 0.3762962962962963,
      "grad_norm": 0.7432051301002502,
      "learning_rate": 1.2732732732732732e-05,
      "loss": 0.1963,
      "step": 127
    },
    {
      "epoch": 0.37925925925925924,
      "grad_norm": 0.7255069017410278,
      "learning_rate": 1.2672672672672675e-05,
      "loss": 0.1835,
      "step": 128
    },
    {
      "epoch": 0.38222222222222224,
      "grad_norm": 0.7553523182868958,
      "learning_rate": 1.2612612612612613e-05,
      "loss": 0.1965,
      "step": 129
    },
    {
      "epoch": 0.3851851851851852,
      "grad_norm": 0.6369490027427673,
      "learning_rate": 1.2552552552552552e-05,
      "loss": 0.1363,
      "step": 130
    },
    {
      "epoch": 0.38814814814814813,
      "grad_norm": 1.1354056596755981,
      "learning_rate": 1.2492492492492493e-05,
      "loss": 0.2725,
      "step": 131
    },
    {
      "epoch": 0.39111111111111113,
      "grad_norm": 0.7551750540733337,
      "learning_rate": 1.2432432432432433e-05,
      "loss": 0.1921,
      "step": 132
    },
    {
      "epoch": 0.3940740740740741,
      "grad_norm": 0.8875067830085754,
      "learning_rate": 1.2372372372372374e-05,
      "loss": 0.2196,
      "step": 133
    },
    {
      "epoch": 0.397037037037037,
      "grad_norm": 0.6833935976028442,
      "learning_rate": 1.2312312312312313e-05,
      "loss": 0.1376,
      "step": 134
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.609214186668396,
      "learning_rate": 1.2252252252252253e-05,
      "loss": 0.1464,
      "step": 135
    },
    {
      "epoch": 0.40296296296296297,
      "grad_norm": 0.7427788972854614,
      "learning_rate": 1.2192192192192194e-05,
      "loss": 0.162,
      "step": 136
    },
    {
      "epoch": 0.4059259259259259,
      "grad_norm": 0.8278418183326721,
      "learning_rate": 1.2132132132132133e-05,
      "loss": 0.1863,
      "step": 137
    },
    {
      "epoch": 0.4088888888888889,
      "grad_norm": 0.753386378288269,
      "learning_rate": 1.2072072072072074e-05,
      "loss": 0.1775,
      "step": 138
    },
    {
      "epoch": 0.41185185185185186,
      "grad_norm": 0.7851639986038208,
      "learning_rate": 1.2012012012012014e-05,
      "loss": 0.1874,
      "step": 139
    },
    {
      "epoch": 0.4148148148148148,
      "grad_norm": 0.5876484513282776,
      "learning_rate": 1.1951951951951951e-05,
      "loss": 0.0881,
      "step": 140
    },
    {
      "epoch": 0.4177777777777778,
      "grad_norm": 0.682448148727417,
      "learning_rate": 1.1891891891891894e-05,
      "loss": 0.1583,
      "step": 141
    },
    {
      "epoch": 0.42074074074074075,
      "grad_norm": 0.888619601726532,
      "learning_rate": 1.1831831831831832e-05,
      "loss": 0.2217,
      "step": 142
    },
    {
      "epoch": 0.4237037037037037,
      "grad_norm": 1.0206968784332275,
      "learning_rate": 1.1771771771771771e-05,
      "loss": 0.2924,
      "step": 143
    },
    {
      "epoch": 0.4266666666666667,
      "grad_norm": 0.7869313359260559,
      "learning_rate": 1.1711711711711713e-05,
      "loss": 0.1712,
      "step": 144
    },
    {
      "epoch": 0.42962962962962964,
      "grad_norm": 0.8629177808761597,
      "learning_rate": 1.1651651651651652e-05,
      "loss": 0.2245,
      "step": 145
    },
    {
      "epoch": 0.4325925925925926,
      "grad_norm": 0.75034099817276,
      "learning_rate": 1.1591591591591593e-05,
      "loss": 0.1528,
      "step": 146
    },
    {
      "epoch": 0.43555555555555553,
      "grad_norm": 1.07233726978302,
      "learning_rate": 1.1531531531531532e-05,
      "loss": 0.243,
      "step": 147
    },
    {
      "epoch": 0.43851851851851853,
      "grad_norm": 1.0244115591049194,
      "learning_rate": 1.1471471471471472e-05,
      "loss": 0.2644,
      "step": 148
    },
    {
      "epoch": 0.4414814814814815,
      "grad_norm": 0.795019268989563,
      "learning_rate": 1.1411411411411413e-05,
      "loss": 0.1668,
      "step": 149
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.8064812421798706,
      "learning_rate": 1.1351351351351352e-05,
      "loss": 0.1443,
      "step": 150
    },
    {
      "epoch": 0.4474074074074074,
      "grad_norm": 0.7531503438949585,
      "learning_rate": 1.1291291291291293e-05,
      "loss": 0.1533,
      "step": 151
    },
    {
      "epoch": 0.45037037037037037,
      "grad_norm": 0.7712652087211609,
      "learning_rate": 1.1231231231231233e-05,
      "loss": 0.1872,
      "step": 152
    },
    {
      "epoch": 0.4533333333333333,
      "grad_norm": 1.0477747917175293,
      "learning_rate": 1.117117117117117e-05,
      "loss": 0.3097,
      "step": 153
    },
    {
      "epoch": 0.4562962962962963,
      "grad_norm": 0.9044317603111267,
      "learning_rate": 1.1111111111111113e-05,
      "loss": 0.2246,
      "step": 154
    },
    {
      "epoch": 0.45925925925925926,
      "grad_norm": 0.6190909147262573,
      "learning_rate": 1.1051051051051051e-05,
      "loss": 0.1266,
      "step": 155
    },
    {
      "epoch": 0.4622222222222222,
      "grad_norm": 0.7761991024017334,
      "learning_rate": 1.0990990990990992e-05,
      "loss": 0.1866,
      "step": 156
    },
    {
      "epoch": 0.4651851851851852,
      "grad_norm": 0.7264747619628906,
      "learning_rate": 1.0930930930930932e-05,
      "loss": 0.1651,
      "step": 157
    },
    {
      "epoch": 0.46814814814814815,
      "grad_norm": 0.7001009583473206,
      "learning_rate": 1.0870870870870871e-05,
      "loss": 0.1319,
      "step": 158
    },
    {
      "epoch": 0.4711111111111111,
      "grad_norm": 0.8649933338165283,
      "learning_rate": 1.0810810810810812e-05,
      "loss": 0.2005,
      "step": 159
    },
    {
      "epoch": 0.4740740740740741,
      "grad_norm": 0.6848715543746948,
      "learning_rate": 1.0750750750750751e-05,
      "loss": 0.1443,
      "step": 160
    },
    {
      "epoch": 0.47703703703703704,
      "grad_norm": 0.5689470171928406,
      "learning_rate": 1.0690690690690693e-05,
      "loss": 0.114,
      "step": 161
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8439794778823853,
      "learning_rate": 1.0630630630630632e-05,
      "loss": 0.2272,
      "step": 162
    },
    {
      "epoch": 0.482962962962963,
      "grad_norm": 0.5890029072761536,
      "learning_rate": 1.0570570570570571e-05,
      "loss": 0.1435,
      "step": 163
    },
    {
      "epoch": 0.48592592592592593,
      "grad_norm": 0.7650848627090454,
      "learning_rate": 1.0510510510510512e-05,
      "loss": 0.1893,
      "step": 164
    },
    {
      "epoch": 0.4888888888888889,
      "grad_norm": 0.7879796624183655,
      "learning_rate": 1.0450450450450452e-05,
      "loss": 0.157,
      "step": 165
    },
    {
      "epoch": 0.4918518518518519,
      "grad_norm": 1.069579839706421,
      "learning_rate": 1.039039039039039e-05,
      "loss": 0.2411,
      "step": 166
    },
    {
      "epoch": 0.4948148148148148,
      "grad_norm": 0.8042727112770081,
      "learning_rate": 1.033033033033033e-05,
      "loss": 0.1624,
      "step": 167
    },
    {
      "epoch": 0.49777777777777776,
      "grad_norm": 0.9903252720832825,
      "learning_rate": 1.027027027027027e-05,
      "loss": 0.2373,
      "step": 168
    },
    {
      "epoch": 0.5007407407407407,
      "grad_norm": 0.7909887433052063,
      "learning_rate": 1.0210210210210211e-05,
      "loss": 0.1945,
      "step": 169
    },
    {
      "epoch": 0.5037037037037037,
      "grad_norm": 0.9773135185241699,
      "learning_rate": 1.015015015015015e-05,
      "loss": 0.2429,
      "step": 170
    },
    {
      "epoch": 0.5066666666666667,
      "grad_norm": 0.9995744824409485,
      "learning_rate": 1.009009009009009e-05,
      "loss": 0.2393,
      "step": 171
    },
    {
      "epoch": 0.5096296296296297,
      "grad_norm": 0.6852418780326843,
      "learning_rate": 1.0030030030030031e-05,
      "loss": 0.1126,
      "step": 172
    },
    {
      "epoch": 0.5125925925925926,
      "grad_norm": 0.688604474067688,
      "learning_rate": 9.96996996996997e-06,
      "loss": 0.1703,
      "step": 173
    },
    {
      "epoch": 0.5155555555555555,
      "grad_norm": 0.7257514595985413,
      "learning_rate": 9.90990990990991e-06,
      "loss": 0.1413,
      "step": 174
    },
    {
      "epoch": 0.5185185185185185,
      "grad_norm": 0.9704393148422241,
      "learning_rate": 9.849849849849851e-06,
      "loss": 0.2468,
      "step": 175
    },
    {
      "epoch": 0.5214814814814814,
      "grad_norm": 0.9529659152030945,
      "learning_rate": 9.78978978978979e-06,
      "loss": 0.1996,
      "step": 176
    },
    {
      "epoch": 0.5244444444444445,
      "grad_norm": 0.7243568897247314,
      "learning_rate": 9.729729729729732e-06,
      "loss": 0.1494,
      "step": 177
    },
    {
      "epoch": 0.5274074074074074,
      "grad_norm": 0.7829004526138306,
      "learning_rate": 9.669669669669671e-06,
      "loss": 0.1708,
      "step": 178
    },
    {
      "epoch": 0.5303703703703704,
      "grad_norm": 0.8569172620773315,
      "learning_rate": 9.60960960960961e-06,
      "loss": 0.1831,
      "step": 179
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 0.7898672223091125,
      "learning_rate": 9.54954954954955e-06,
      "loss": 0.2213,
      "step": 180
    },
    {
      "epoch": 0.5362962962962963,
      "grad_norm": 0.9253923296928406,
      "learning_rate": 9.489489489489491e-06,
      "loss": 0.2417,
      "step": 181
    },
    {
      "epoch": 0.5392592592592592,
      "grad_norm": 0.8531503081321716,
      "learning_rate": 9.42942942942943e-06,
      "loss": 0.2049,
      "step": 182
    },
    {
      "epoch": 0.5422222222222223,
      "grad_norm": 0.8846073150634766,
      "learning_rate": 9.36936936936937e-06,
      "loss": 0.2037,
      "step": 183
    },
    {
      "epoch": 0.5451851851851852,
      "grad_norm": 0.866161584854126,
      "learning_rate": 9.30930930930931e-06,
      "loss": 0.2555,
      "step": 184
    },
    {
      "epoch": 0.5481481481481482,
      "grad_norm": 0.8531607985496521,
      "learning_rate": 9.24924924924925e-06,
      "loss": 0.2237,
      "step": 185
    },
    {
      "epoch": 0.5511111111111111,
      "grad_norm": 0.9555264115333557,
      "learning_rate": 9.189189189189191e-06,
      "loss": 0.2451,
      "step": 186
    },
    {
      "epoch": 0.554074074074074,
      "grad_norm": 0.8263760805130005,
      "learning_rate": 9.129129129129129e-06,
      "loss": 0.213,
      "step": 187
    },
    {
      "epoch": 0.557037037037037,
      "grad_norm": 0.9021344184875488,
      "learning_rate": 9.06906906906907e-06,
      "loss": 0.1881,
      "step": 188
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.662638247013092,
      "learning_rate": 9.00900900900901e-06,
      "loss": 0.1603,
      "step": 189
    },
    {
      "epoch": 0.562962962962963,
      "grad_norm": 0.6390892267227173,
      "learning_rate": 8.94894894894895e-06,
      "loss": 0.1305,
      "step": 190
    },
    {
      "epoch": 0.5659259259259259,
      "grad_norm": 1.066746711730957,
      "learning_rate": 8.888888888888888e-06,
      "loss": 0.2671,
      "step": 191
    },
    {
      "epoch": 0.5688888888888889,
      "grad_norm": 1.1001776456832886,
      "learning_rate": 8.82882882882883e-06,
      "loss": 0.2553,
      "step": 192
    },
    {
      "epoch": 0.5718518518518518,
      "grad_norm": 0.7368589043617249,
      "learning_rate": 8.768768768768769e-06,
      "loss": 0.1293,
      "step": 193
    },
    {
      "epoch": 0.5748148148148148,
      "grad_norm": 0.5754815936088562,
      "learning_rate": 8.70870870870871e-06,
      "loss": 0.1206,
      "step": 194
    },
    {
      "epoch": 0.5777777777777777,
      "grad_norm": 0.4183901846408844,
      "learning_rate": 8.64864864864865e-06,
      "loss": 0.0499,
      "step": 195
    },
    {
      "epoch": 0.5807407407407408,
      "grad_norm": 0.8268377184867859,
      "learning_rate": 8.588588588588589e-06,
      "loss": 0.222,
      "step": 196
    },
    {
      "epoch": 0.5837037037037037,
      "grad_norm": 0.7017174363136292,
      "learning_rate": 8.52852852852853e-06,
      "loss": 0.1491,
      "step": 197
    },
    {
      "epoch": 0.5866666666666667,
      "grad_norm": 0.7820301651954651,
      "learning_rate": 8.46846846846847e-06,
      "loss": 0.2033,
      "step": 198
    },
    {
      "epoch": 0.5896296296296296,
      "grad_norm": 0.7056408524513245,
      "learning_rate": 8.408408408408409e-06,
      "loss": 0.1905,
      "step": 199
    },
    {
      "epoch": 0.5925925925925926,
      "grad_norm": 0.8303424715995789,
      "learning_rate": 8.348348348348348e-06,
      "loss": 0.1687,
      "step": 200
    },
    {
      "epoch": 0.5955555555555555,
      "grad_norm": 0.6162233352661133,
      "learning_rate": 8.288288288288289e-06,
      "loss": 0.1413,
      "step": 201
    },
    {
      "epoch": 0.5985185185185186,
      "grad_norm": 0.6666263937950134,
      "learning_rate": 8.228228228228229e-06,
      "loss": 0.1677,
      "step": 202
    },
    {
      "epoch": 0.6014814814814815,
      "grad_norm": 0.8014916777610779,
      "learning_rate": 8.16816816816817e-06,
      "loss": 0.1505,
      "step": 203
    },
    {
      "epoch": 0.6044444444444445,
      "grad_norm": 0.575874388217926,
      "learning_rate": 8.108108108108109e-06,
      "loss": 0.0924,
      "step": 204
    },
    {
      "epoch": 0.6074074074074074,
      "grad_norm": 0.6968632936477661,
      "learning_rate": 8.048048048048048e-06,
      "loss": 0.1663,
      "step": 205
    },
    {
      "epoch": 0.6103703703703703,
      "grad_norm": 0.8517196774482727,
      "learning_rate": 7.987987987987988e-06,
      "loss": 0.193,
      "step": 206
    },
    {
      "epoch": 0.6133333333333333,
      "grad_norm": 1.008732557296753,
      "learning_rate": 7.927927927927929e-06,
      "loss": 0.2663,
      "step": 207
    },
    {
      "epoch": 0.6162962962962963,
      "grad_norm": 0.8401888012886047,
      "learning_rate": 7.867867867867868e-06,
      "loss": 0.2105,
      "step": 208
    },
    {
      "epoch": 0.6192592592592593,
      "grad_norm": 0.8538157939910889,
      "learning_rate": 7.807807807807808e-06,
      "loss": 0.2152,
      "step": 209
    },
    {
      "epoch": 0.6222222222222222,
      "grad_norm": 0.8950092196464539,
      "learning_rate": 7.747747747747749e-06,
      "loss": 0.216,
      "step": 210
    },
    {
      "epoch": 0.6251851851851852,
      "grad_norm": 0.9501015543937683,
      "learning_rate": 7.687687687687688e-06,
      "loss": 0.1953,
      "step": 211
    },
    {
      "epoch": 0.6281481481481481,
      "grad_norm": 0.7668682932853699,
      "learning_rate": 7.6276276276276285e-06,
      "loss": 0.187,
      "step": 212
    },
    {
      "epoch": 0.6311111111111111,
      "grad_norm": 0.9458869099617004,
      "learning_rate": 7.567567567567569e-06,
      "loss": 0.2163,
      "step": 213
    },
    {
      "epoch": 0.6340740740740741,
      "grad_norm": 0.7746089696884155,
      "learning_rate": 7.507507507507507e-06,
      "loss": 0.1789,
      "step": 214
    },
    {
      "epoch": 0.6370370370370371,
      "grad_norm": 0.7921304106712341,
      "learning_rate": 7.447447447447448e-06,
      "loss": 0.1929,
      "step": 215
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9870914816856384,
      "learning_rate": 7.387387387387388e-06,
      "loss": 0.2484,
      "step": 216
    },
    {
      "epoch": 0.642962962962963,
      "grad_norm": 0.501486599445343,
      "learning_rate": 7.327327327327328e-06,
      "loss": 0.0788,
      "step": 217
    },
    {
      "epoch": 0.6459259259259259,
      "grad_norm": 0.6069294214248657,
      "learning_rate": 7.267267267267268e-06,
      "loss": 0.1039,
      "step": 218
    },
    {
      "epoch": 0.6488888888888888,
      "grad_norm": 0.8053973317146301,
      "learning_rate": 7.207207207207208e-06,
      "loss": 0.1514,
      "step": 219
    },
    {
      "epoch": 0.6518518518518519,
      "grad_norm": 0.7112203240394592,
      "learning_rate": 7.147147147147148e-06,
      "loss": 0.1389,
      "step": 220
    },
    {
      "epoch": 0.6548148148148148,
      "grad_norm": 0.6717841029167175,
      "learning_rate": 7.087087087087087e-06,
      "loss": 0.167,
      "step": 221
    },
    {
      "epoch": 0.6577777777777778,
      "grad_norm": 0.8067883849143982,
      "learning_rate": 7.027027027027028e-06,
      "loss": 0.1897,
      "step": 222
    },
    {
      "epoch": 0.6607407407407407,
      "grad_norm": 0.7958944439888,
      "learning_rate": 6.966966966966967e-06,
      "loss": 0.1792,
      "step": 223
    },
    {
      "epoch": 0.6637037037037037,
      "grad_norm": 0.6707459092140198,
      "learning_rate": 6.906906906906907e-06,
      "loss": 0.1416,
      "step": 224
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.8315305113792419,
      "learning_rate": 6.846846846846848e-06,
      "loss": 0.1786,
      "step": 225
    },
    {
      "epoch": 0.6696296296296296,
      "grad_norm": 0.8573684096336365,
      "learning_rate": 6.786786786786788e-06,
      "loss": 0.1734,
      "step": 226
    },
    {
      "epoch": 0.6725925925925926,
      "grad_norm": 0.6853092312812805,
      "learning_rate": 6.726726726726728e-06,
      "loss": 0.1314,
      "step": 227
    },
    {
      "epoch": 0.6755555555555556,
      "grad_norm": 0.8630515336990356,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.1932,
      "step": 228
    },
    {
      "epoch": 0.6785185185185185,
      "grad_norm": 0.6693196892738342,
      "learning_rate": 6.606606606606607e-06,
      "loss": 0.1231,
      "step": 229
    },
    {
      "epoch": 0.6814814814814815,
      "grad_norm": 0.6657727360725403,
      "learning_rate": 6.546546546546547e-06,
      "loss": 0.1423,
      "step": 230
    },
    {
      "epoch": 0.6844444444444444,
      "grad_norm": 0.8997541069984436,
      "learning_rate": 6.486486486486487e-06,
      "loss": 0.2338,
      "step": 231
    },
    {
      "epoch": 0.6874074074074074,
      "grad_norm": 0.7045704126358032,
      "learning_rate": 6.426426426426427e-06,
      "loss": 0.1729,
      "step": 232
    },
    {
      "epoch": 0.6903703703703704,
      "grad_norm": 0.6608391404151917,
      "learning_rate": 6.366366366366366e-06,
      "loss": 0.1309,
      "step": 233
    },
    {
      "epoch": 0.6933333333333334,
      "grad_norm": 1.0434973239898682,
      "learning_rate": 6.3063063063063065e-06,
      "loss": 0.211,
      "step": 234
    },
    {
      "epoch": 0.6962962962962963,
      "grad_norm": 0.8890491127967834,
      "learning_rate": 6.246246246246247e-06,
      "loss": 0.2194,
      "step": 235
    },
    {
      "epoch": 0.6992592592592592,
      "grad_norm": 0.7882158756256104,
      "learning_rate": 6.186186186186187e-06,
      "loss": 0.2079,
      "step": 236
    },
    {
      "epoch": 0.7022222222222222,
      "grad_norm": 0.8074647188186646,
      "learning_rate": 6.126126126126126e-06,
      "loss": 0.1924,
      "step": 237
    },
    {
      "epoch": 0.7051851851851851,
      "grad_norm": 0.7704930901527405,
      "learning_rate": 6.066066066066067e-06,
      "loss": 0.1749,
      "step": 238
    },
    {
      "epoch": 0.7081481481481482,
      "grad_norm": 0.7319555282592773,
      "learning_rate": 6.006006006006007e-06,
      "loss": 0.133,
      "step": 239
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 0.7055200934410095,
      "learning_rate": 5.945945945945947e-06,
      "loss": 0.1493,
      "step": 240
    },
    {
      "epoch": 0.7140740740740741,
      "grad_norm": 0.5785249471664429,
      "learning_rate": 5.885885885885886e-06,
      "loss": 0.1256,
      "step": 241
    },
    {
      "epoch": 0.717037037037037,
      "grad_norm": 0.8348391652107239,
      "learning_rate": 5.825825825825826e-06,
      "loss": 0.1676,
      "step": 242
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9011757373809814,
      "learning_rate": 5.765765765765766e-06,
      "loss": 0.2155,
      "step": 243
    },
    {
      "epoch": 0.7229629629629629,
      "grad_norm": 0.8325709104537964,
      "learning_rate": 5.7057057057057065e-06,
      "loss": 0.2246,
      "step": 244
    },
    {
      "epoch": 0.725925925925926,
      "grad_norm": 0.9126022458076477,
      "learning_rate": 5.645645645645647e-06,
      "loss": 0.2313,
      "step": 245
    },
    {
      "epoch": 0.7288888888888889,
      "grad_norm": 0.8830057382583618,
      "learning_rate": 5.585585585585585e-06,
      "loss": 0.1802,
      "step": 246
    },
    {
      "epoch": 0.7318518518518519,
      "grad_norm": 0.6011220812797546,
      "learning_rate": 5.5255255255255255e-06,
      "loss": 0.1163,
      "step": 247
    },
    {
      "epoch": 0.7348148148148148,
      "grad_norm": 0.6792129874229431,
      "learning_rate": 5.465465465465466e-06,
      "loss": 0.1569,
      "step": 248
    },
    {
      "epoch": 0.7377777777777778,
      "grad_norm": 0.6678133606910706,
      "learning_rate": 5.405405405405406e-06,
      "loss": 0.156,
      "step": 249
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 0.6404563188552856,
      "learning_rate": 5.345345345345346e-06,
      "loss": 0.1426,
      "step": 250
    },
    {
      "epoch": 0.7437037037037038,
      "grad_norm": 0.6993813514709473,
      "learning_rate": 5.285285285285286e-06,
      "loss": 0.1377,
      "step": 251
    },
    {
      "epoch": 0.7466666666666667,
      "grad_norm": 1.0929012298583984,
      "learning_rate": 5.225225225225226e-06,
      "loss": 0.2673,
      "step": 252
    },
    {
      "epoch": 0.7496296296296296,
      "grad_norm": 0.7563413381576538,
      "learning_rate": 5.165165165165165e-06,
      "loss": 0.1417,
      "step": 253
    },
    {
      "epoch": 0.7525925925925926,
      "grad_norm": 0.657166063785553,
      "learning_rate": 5.105105105105106e-06,
      "loss": 0.1578,
      "step": 254
    },
    {
      "epoch": 0.7555555555555555,
      "grad_norm": 0.5090486407279968,
      "learning_rate": 5.045045045045045e-06,
      "loss": 0.0761,
      "step": 255
    },
    {
      "epoch": 0.7585185185185185,
      "grad_norm": 0.9304921627044678,
      "learning_rate": 4.984984984984985e-06,
      "loss": 0.2237,
      "step": 256
    },
    {
      "epoch": 0.7614814814814815,
      "grad_norm": 0.8906775116920471,
      "learning_rate": 4.9249249249249255e-06,
      "loss": 0.1846,
      "step": 257
    },
    {
      "epoch": 0.7644444444444445,
      "grad_norm": 0.7848439812660217,
      "learning_rate": 4.864864864864866e-06,
      "loss": 0.1771,
      "step": 258
    },
    {
      "epoch": 0.7674074074074074,
      "grad_norm": 0.8842347860336304,
      "learning_rate": 4.804804804804805e-06,
      "loss": 0.2247,
      "step": 259
    },
    {
      "epoch": 0.7703703703703704,
      "grad_norm": 0.80910724401474,
      "learning_rate": 4.7447447447447454e-06,
      "loss": 0.2321,
      "step": 260
    },
    {
      "epoch": 0.7733333333333333,
      "grad_norm": 0.7453343272209167,
      "learning_rate": 4.684684684684685e-06,
      "loss": 0.1313,
      "step": 261
    },
    {
      "epoch": 0.7762962962962963,
      "grad_norm": 1.0571560859680176,
      "learning_rate": 4.624624624624625e-06,
      "loss": 0.2783,
      "step": 262
    },
    {
      "epoch": 0.7792592592592592,
      "grad_norm": 1.0411971807479858,
      "learning_rate": 4.5645645645645645e-06,
      "loss": 0.2901,
      "step": 263
    },
    {
      "epoch": 0.7822222222222223,
      "grad_norm": 0.7956094145774841,
      "learning_rate": 4.504504504504505e-06,
      "loss": 0.1771,
      "step": 264
    },
    {
      "epoch": 0.7851851851851852,
      "grad_norm": 0.8621307611465454,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.2128,
      "step": 265
    },
    {
      "epoch": 0.7881481481481482,
      "grad_norm": 0.6987168192863464,
      "learning_rate": 4.384384384384384e-06,
      "loss": 0.1244,
      "step": 266
    },
    {
      "epoch": 0.7911111111111111,
      "grad_norm": 0.8145896196365356,
      "learning_rate": 4.324324324324325e-06,
      "loss": 0.2093,
      "step": 267
    },
    {
      "epoch": 0.794074074074074,
      "grad_norm": 0.7333759069442749,
      "learning_rate": 4.264264264264265e-06,
      "loss": 0.1668,
      "step": 268
    },
    {
      "epoch": 0.797037037037037,
      "grad_norm": 0.8501369953155518,
      "learning_rate": 4.204204204204204e-06,
      "loss": 0.2391,
      "step": 269
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.7657726407051086,
      "learning_rate": 4.1441441441441446e-06,
      "loss": 0.1954,
      "step": 270
    },
    {
      "epoch": 0.802962962962963,
      "grad_norm": 0.7661365270614624,
      "learning_rate": 4.084084084084085e-06,
      "loss": 0.1933,
      "step": 271
    },
    {
      "epoch": 0.8059259259259259,
      "grad_norm": 1.0544160604476929,
      "learning_rate": 4.024024024024024e-06,
      "loss": 0.2213,
      "step": 272
    },
    {
      "epoch": 0.8088888888888889,
      "grad_norm": 0.8172379732131958,
      "learning_rate": 3.9639639639639645e-06,
      "loss": 0.2103,
      "step": 273
    },
    {
      "epoch": 0.8118518518518518,
      "grad_norm": 0.678930938243866,
      "learning_rate": 3.903903903903904e-06,
      "loss": 0.1306,
      "step": 274
    },
    {
      "epoch": 0.8148148148148148,
      "grad_norm": 1.108206033706665,
      "learning_rate": 3.843843843843844e-06,
      "loss": 0.2736,
      "step": 275
    },
    {
      "epoch": 0.8177777777777778,
      "grad_norm": 0.7077149748802185,
      "learning_rate": 3.7837837837837844e-06,
      "loss": 0.1244,
      "step": 276
    },
    {
      "epoch": 0.8207407407407408,
      "grad_norm": 0.7655659914016724,
      "learning_rate": 3.723723723723724e-06,
      "loss": 0.1531,
      "step": 277
    },
    {
      "epoch": 0.8237037037037037,
      "grad_norm": 0.7861891984939575,
      "learning_rate": 3.663663663663664e-06,
      "loss": 0.1813,
      "step": 278
    },
    {
      "epoch": 0.8266666666666667,
      "grad_norm": 0.598543107509613,
      "learning_rate": 3.603603603603604e-06,
      "loss": 0.1331,
      "step": 279
    },
    {
      "epoch": 0.8296296296296296,
      "grad_norm": 0.9774670004844666,
      "learning_rate": 3.5435435435435437e-06,
      "loss": 0.2694,
      "step": 280
    },
    {
      "epoch": 0.8325925925925926,
      "grad_norm": 0.8227265477180481,
      "learning_rate": 3.4834834834834835e-06,
      "loss": 0.182,
      "step": 281
    },
    {
      "epoch": 0.8355555555555556,
      "grad_norm": 0.7918034791946411,
      "learning_rate": 3.423423423423424e-06,
      "loss": 0.246,
      "step": 282
    },
    {
      "epoch": 0.8385185185185186,
      "grad_norm": 0.7987902164459229,
      "learning_rate": 3.363363363363364e-06,
      "loss": 0.1529,
      "step": 283
    },
    {
      "epoch": 0.8414814814814815,
      "grad_norm": 1.144445776939392,
      "learning_rate": 3.3033033033033035e-06,
      "loss": 0.2491,
      "step": 284
    },
    {
      "epoch": 0.8444444444444444,
      "grad_norm": 0.6797105669975281,
      "learning_rate": 3.2432432432432437e-06,
      "loss": 0.157,
      "step": 285
    },
    {
      "epoch": 0.8474074074074074,
      "grad_norm": 0.9168769121170044,
      "learning_rate": 3.183183183183183e-06,
      "loss": 0.212,
      "step": 286
    },
    {
      "epoch": 0.8503703703703703,
      "grad_norm": 0.5935839414596558,
      "learning_rate": 3.1231231231231234e-06,
      "loss": 0.0953,
      "step": 287
    },
    {
      "epoch": 0.8533333333333334,
      "grad_norm": 0.5830032229423523,
      "learning_rate": 3.063063063063063e-06,
      "loss": 0.1169,
      "step": 288
    },
    {
      "epoch": 0.8562962962962963,
      "grad_norm": 0.6495141983032227,
      "learning_rate": 3.0030030030030034e-06,
      "loss": 0.1531,
      "step": 289
    },
    {
      "epoch": 0.8592592592592593,
      "grad_norm": 0.9052508473396301,
      "learning_rate": 2.942942942942943e-06,
      "loss": 0.2584,
      "step": 290
    },
    {
      "epoch": 0.8622222222222222,
      "grad_norm": 0.4937392473220825,
      "learning_rate": 2.882882882882883e-06,
      "loss": 0.0978,
      "step": 291
    },
    {
      "epoch": 0.8651851851851852,
      "grad_norm": 0.8803480267524719,
      "learning_rate": 2.8228228228228234e-06,
      "loss": 0.21,
      "step": 292
    },
    {
      "epoch": 0.8681481481481481,
      "grad_norm": 0.7183660864830017,
      "learning_rate": 2.7627627627627628e-06,
      "loss": 0.1728,
      "step": 293
    },
    {
      "epoch": 0.8711111111111111,
      "grad_norm": 0.706757128238678,
      "learning_rate": 2.702702702702703e-06,
      "loss": 0.1599,
      "step": 294
    },
    {
      "epoch": 0.8740740740740741,
      "grad_norm": 0.5740911960601807,
      "learning_rate": 2.642642642642643e-06,
      "loss": 0.094,
      "step": 295
    },
    {
      "epoch": 0.8770370370370371,
      "grad_norm": 0.9149448275566101,
      "learning_rate": 2.5825825825825827e-06,
      "loss": 0.2414,
      "step": 296
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.6597499847412109,
      "learning_rate": 2.5225225225225225e-06,
      "loss": 0.1116,
      "step": 297
    },
    {
      "epoch": 0.882962962962963,
      "grad_norm": 0.9077011942863464,
      "learning_rate": 2.4624624624624628e-06,
      "loss": 0.2162,
      "step": 298
    },
    {
      "epoch": 0.8859259259259259,
      "grad_norm": 0.6825413703918457,
      "learning_rate": 2.4024024024024026e-06,
      "loss": 0.157,
      "step": 299
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.7451571226119995,
      "learning_rate": 2.3423423423423424e-06,
      "loss": 0.1212,
      "step": 300
    },
    {
      "epoch": 0.8918518518518519,
      "grad_norm": 0.6453779935836792,
      "learning_rate": 2.2822822822822822e-06,
      "loss": 0.1262,
      "step": 301
    },
    {
      "epoch": 0.8948148148148148,
      "grad_norm": 0.9824874997138977,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.2523,
      "step": 302
    },
    {
      "epoch": 0.8977777777777778,
      "grad_norm": 0.8559462428092957,
      "learning_rate": 2.1621621621621623e-06,
      "loss": 0.1863,
      "step": 303
    },
    {
      "epoch": 0.9007407407407407,
      "grad_norm": 0.8423551321029663,
      "learning_rate": 2.102102102102102e-06,
      "loss": 0.193,
      "step": 304
    },
    {
      "epoch": 0.9037037037037037,
      "grad_norm": 0.7797116041183472,
      "learning_rate": 2.0420420420420424e-06,
      "loss": 0.1908,
      "step": 305
    },
    {
      "epoch": 0.9066666666666666,
      "grad_norm": 0.48622554540634155,
      "learning_rate": 1.9819819819819822e-06,
      "loss": 0.0445,
      "step": 306
    },
    {
      "epoch": 0.9096296296296297,
      "grad_norm": 0.7244918346405029,
      "learning_rate": 1.921921921921922e-06,
      "loss": 0.1502,
      "step": 307
    },
    {
      "epoch": 0.9125925925925926,
      "grad_norm": 0.9780031442642212,
      "learning_rate": 1.861861861861862e-06,
      "loss": 0.246,
      "step": 308
    },
    {
      "epoch": 0.9155555555555556,
      "grad_norm": 0.7556838393211365,
      "learning_rate": 1.801801801801802e-06,
      "loss": 0.1603,
      "step": 309
    },
    {
      "epoch": 0.9185185185185185,
      "grad_norm": 0.5545371770858765,
      "learning_rate": 1.7417417417417418e-06,
      "loss": 0.1052,
      "step": 310
    },
    {
      "epoch": 0.9214814814814815,
      "grad_norm": 0.9142343997955322,
      "learning_rate": 1.681681681681682e-06,
      "loss": 0.1953,
      "step": 311
    },
    {
      "epoch": 0.9244444444444444,
      "grad_norm": 0.6124354004859924,
      "learning_rate": 1.6216216216216219e-06,
      "loss": 0.1231,
      "step": 312
    },
    {
      "epoch": 0.9274074074074075,
      "grad_norm": 0.8703904747962952,
      "learning_rate": 1.5615615615615617e-06,
      "loss": 0.1495,
      "step": 313
    },
    {
      "epoch": 0.9303703703703704,
      "grad_norm": 1.0496271848678589,
      "learning_rate": 1.5015015015015017e-06,
      "loss": 0.3053,
      "step": 314
    },
    {
      "epoch": 0.9333333333333333,
      "grad_norm": 0.8166453242301941,
      "learning_rate": 1.4414414414414416e-06,
      "loss": 0.1677,
      "step": 315
    },
    {
      "epoch": 0.9362962962962963,
      "grad_norm": 0.8786421418190002,
      "learning_rate": 1.3813813813813814e-06,
      "loss": 0.2057,
      "step": 316
    },
    {
      "epoch": 0.9392592592592592,
      "grad_norm": 0.8979144096374512,
      "learning_rate": 1.3213213213213214e-06,
      "loss": 0.2082,
      "step": 317
    },
    {
      "epoch": 0.9422222222222222,
      "grad_norm": 0.7222133874893188,
      "learning_rate": 1.2612612612612613e-06,
      "loss": 0.1288,
      "step": 318
    },
    {
      "epoch": 0.9451851851851852,
      "grad_norm": 0.9162706732749939,
      "learning_rate": 1.2012012012012013e-06,
      "loss": 0.209,
      "step": 319
    },
    {
      "epoch": 0.9481481481481482,
      "grad_norm": 0.7953757047653198,
      "learning_rate": 1.1411411411411411e-06,
      "loss": 0.1788,
      "step": 320
    },
    {
      "epoch": 0.9511111111111111,
      "grad_norm": 0.9482904076576233,
      "learning_rate": 1.0810810810810812e-06,
      "loss": 0.2045,
      "step": 321
    },
    {
      "epoch": 0.9540740740740741,
      "grad_norm": 0.7969358563423157,
      "learning_rate": 1.0210210210210212e-06,
      "loss": 0.186,
      "step": 322
    },
    {
      "epoch": 0.957037037037037,
      "grad_norm": 0.7884560823440552,
      "learning_rate": 9.60960960960961e-07,
      "loss": 0.1892,
      "step": 323
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.0385804176330566,
      "learning_rate": 9.00900900900901e-07,
      "loss": 0.236,
      "step": 324
    },
    {
      "epoch": 0.9629629629629629,
      "grad_norm": 1.1920503377914429,
      "learning_rate": 8.40840840840841e-07,
      "loss": 0.2931,
      "step": 325
    },
    {
      "epoch": 0.965925925925926,
      "grad_norm": 0.7292988300323486,
      "learning_rate": 7.807807807807808e-07,
      "loss": 0.1497,
      "step": 326
    },
    {
      "epoch": 0.9688888888888889,
      "grad_norm": 0.8104760646820068,
      "learning_rate": 7.207207207207208e-07,
      "loss": 0.2147,
      "step": 327
    },
    {
      "epoch": 0.9718518518518519,
      "grad_norm": 1.0047756433486938,
      "learning_rate": 6.606606606606607e-07,
      "loss": 0.2137,
      "step": 328
    },
    {
      "epoch": 0.9748148148148148,
      "grad_norm": 0.7823283076286316,
      "learning_rate": 6.006006006006006e-07,
      "loss": 0.1514,
      "step": 329
    },
    {
      "epoch": 0.9777777777777777,
      "grad_norm": 0.7249939441680908,
      "learning_rate": 5.405405405405406e-07,
      "loss": 0.1722,
      "step": 330
    },
    {
      "epoch": 0.9807407407407407,
      "grad_norm": 0.5971034169197083,
      "learning_rate": 4.804804804804805e-07,
      "loss": 0.0831,
      "step": 331
    },
    {
      "epoch": 0.9837037037037037,
      "grad_norm": 1.0094950199127197,
      "learning_rate": 4.204204204204205e-07,
      "loss": 0.2318,
      "step": 332
    },
    {
      "epoch": 0.9866666666666667,
      "grad_norm": 0.8682098388671875,
      "learning_rate": 3.603603603603604e-07,
      "loss": 0.1707,
      "step": 333
    },
    {
      "epoch": 0.9896296296296296,
      "grad_norm": 0.8257949352264404,
      "learning_rate": 3.003003003003003e-07,
      "loss": 0.1442,
      "step": 334
    },
    {
      "epoch": 0.9925925925925926,
      "grad_norm": 0.9115303754806519,
      "learning_rate": 2.4024024024024026e-07,
      "loss": 0.1683,
      "step": 335
    },
    {
      "epoch": 0.9955555555555555,
      "grad_norm": 0.47800108790397644,
      "learning_rate": 1.801801801801802e-07,
      "loss": 0.0912,
      "step": 336
    },
    {
      "epoch": 0.9985185185185185,
      "grad_norm": 0.9469006657600403,
      "learning_rate": 1.2012012012012013e-07,
      "loss": 0.2259,
      "step": 337
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.164496660232544,
      "learning_rate": 6.006006006006006e-08,
      "loss": 0.2045,
      "step": 338
    }
  ],
  "logging_steps": 1,
  "max_steps": 338,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.517478373873357e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}