{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 315,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015873015873015872,
      "grad_norm": 3.3795859813690186,
      "learning_rate": 6.25e-07,
      "loss": 0.4037,
      "step": 1
    },
    {
      "epoch": 0.031746031746031744,
      "grad_norm": 4.176924228668213,
      "learning_rate": 1.25e-06,
      "loss": 0.6101,
      "step": 2
    },
    {
      "epoch": 0.047619047619047616,
      "grad_norm": 3.2537522315979004,
      "learning_rate": 1.8750000000000003e-06,
      "loss": 0.5335,
      "step": 3
    },
    {
      "epoch": 0.06349206349206349,
      "grad_norm": 3.8701913356781006,
      "learning_rate": 2.5e-06,
      "loss": 0.5156,
      "step": 4
    },
    {
      "epoch": 0.07936507936507936,
      "grad_norm": 3.2381300926208496,
      "learning_rate": 3.125e-06,
      "loss": 0.5435,
      "step": 5
    },
    {
      "epoch": 0.09523809523809523,
      "grad_norm": 2.6569769382476807,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.4349,
      "step": 6
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 2.641036033630371,
      "learning_rate": 4.3750000000000005e-06,
      "loss": 0.4657,
      "step": 7
    },
    {
      "epoch": 0.12698412698412698,
      "grad_norm": 2.500192880630493,
      "learning_rate": 5e-06,
      "loss": 0.387,
      "step": 8
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 2.925560474395752,
      "learning_rate": 5.625e-06,
      "loss": 0.5708,
      "step": 9
    },
    {
      "epoch": 0.15873015873015872,
      "grad_norm": 1.740651249885559,
      "learning_rate": 6.25e-06,
      "loss": 0.4166,
      "step": 10
    },
    {
      "epoch": 0.1746031746031746,
      "grad_norm": 1.8194835186004639,
      "learning_rate": 6.875e-06,
      "loss": 0.4392,
      "step": 11
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 1.2873449325561523,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.3224,
      "step": 12
    },
    {
      "epoch": 0.20634920634920634,
      "grad_norm": 1.4529649019241333,
      "learning_rate": 8.125000000000001e-06,
      "loss": 0.3407,
      "step": 13
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 1.3093940019607544,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.3677,
      "step": 14
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 1.4765881299972534,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.3792,
      "step": 15
    },
    {
      "epoch": 0.25396825396825395,
      "grad_norm": 5.271663665771484,
      "learning_rate": 1e-05,
      "loss": 0.6346,
      "step": 16
    },
    {
      "epoch": 0.2698412698412698,
      "grad_norm": 1.3494993448257446,
      "learning_rate": 9.999724009977419e-06,
      "loss": 0.4433,
      "step": 17
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 1.115939736366272,
      "learning_rate": 9.998896070377873e-06,
      "loss": 0.2864,
      "step": 18
    },
    {
      "epoch": 0.30158730158730157,
      "grad_norm": 1.5331807136535645,
      "learning_rate": 9.99751627260259e-06,
      "loss": 0.3423,
      "step": 19
    },
    {
      "epoch": 0.31746031746031744,
      "grad_norm": 1.5603549480438232,
      "learning_rate": 9.995584768975735e-06,
      "loss": 0.3799,
      "step": 20
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 1.4752039909362793,
      "learning_rate": 9.993101772727602e-06,
      "loss": 0.3565,
      "step": 21
    },
    {
      "epoch": 0.3492063492063492,
      "grad_norm": 9.61251449584961,
      "learning_rate": 9.990067557971068e-06,
      "loss": 0.7244,
      "step": 22
    },
    {
      "epoch": 0.36507936507936506,
      "grad_norm": 1.3895500898361206,
      "learning_rate": 9.986482459671332e-06,
      "loss": 0.3444,
      "step": 23
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 1.231410026550293,
      "learning_rate": 9.982346873608936e-06,
      "loss": 0.3221,
      "step": 24
    },
    {
      "epoch": 0.3968253968253968,
      "grad_norm": 1.9909180402755737,
      "learning_rate": 9.977661256336081e-06,
      "loss": 0.3724,
      "step": 25
    },
    {
      "epoch": 0.4126984126984127,
      "grad_norm": 1.2321062088012695,
      "learning_rate": 9.972426125126208e-06,
      "loss": 0.3391,
      "step": 26
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 5.4099321365356445,
      "learning_rate": 9.966642057916915e-06,
      "loss": 0.5806,
      "step": 27
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 1.0613089799880981,
      "learning_rate": 9.960309693246135e-06,
      "loss": 0.2237,
      "step": 28
    },
    {
      "epoch": 0.4603174603174603,
      "grad_norm": 1.1249053478240967,
      "learning_rate": 9.953429730181653e-06,
      "loss": 0.3618,
      "step": 29
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 1.0634913444519043,
      "learning_rate": 9.94600292824394e-06,
      "loss": 0.3177,
      "step": 30
    },
    {
      "epoch": 0.49206349206349204,
      "grad_norm": 1.1851016283035278,
      "learning_rate": 9.938030107322284e-06,
      "loss": 0.2799,
      "step": 31
    },
    {
      "epoch": 0.5079365079365079,
      "grad_norm": 1.2439275979995728,
      "learning_rate": 9.929512147584297e-06,
      "loss": 0.3415,
      "step": 32
    },
    {
      "epoch": 0.5238095238095238,
      "grad_norm": 1.1060391664505005,
      "learning_rate": 9.920449989378741e-06,
      "loss": 0.2888,
      "step": 33
    },
    {
      "epoch": 0.5396825396825397,
      "grad_norm": 1.4267876148223877,
      "learning_rate": 9.910844633131712e-06,
      "loss": 0.3724,
      "step": 34
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 1.1967415809631348,
      "learning_rate": 9.90069713923621e-06,
      "loss": 0.3362,
      "step": 35
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.9548850059509277,
      "learning_rate": 9.890008627935057e-06,
      "loss": 0.2776,
      "step": 36
    },
    {
      "epoch": 0.5873015873015873,
      "grad_norm": 1.1262502670288086,
      "learning_rate": 9.878780279197246e-06,
      "loss": 0.3215,
      "step": 37
    },
    {
      "epoch": 0.6031746031746031,
      "grad_norm": 0.9589006304740906,
      "learning_rate": 9.867013332587667e-06,
      "loss": 0.2871,
      "step": 38
    },
    {
      "epoch": 0.6190476190476191,
      "grad_norm": 1.2175616025924683,
      "learning_rate": 9.854709087130261e-06,
      "loss": 0.2672,
      "step": 39
    },
    {
      "epoch": 0.6349206349206349,
      "grad_norm": 1.1927433013916016,
      "learning_rate": 9.841868901164621e-06,
      "loss": 0.3167,
      "step": 40
    },
    {
      "epoch": 0.6507936507936508,
      "grad_norm": 1.1496555805206299,
      "learning_rate": 9.828494192196037e-06,
      "loss": 0.331,
      "step": 41
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 2.9185266494750977,
      "learning_rate": 9.814586436738998e-06,
      "loss": 0.4792,
      "step": 42
    },
    {
      "epoch": 0.6825396825396826,
      "grad_norm": 4.013699054718018,
      "learning_rate": 9.8001471701542e-06,
      "loss": 0.3316,
      "step": 43
    },
    {
      "epoch": 0.6984126984126984,
      "grad_norm": 1.1543551683425903,
      "learning_rate": 9.785177986479049e-06,
      "loss": 0.2954,
      "step": 44
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.9748827219009399,
      "learning_rate": 9.76968053825168e-06,
      "loss": 0.2267,
      "step": 45
    },
    {
      "epoch": 0.7301587301587301,
      "grad_norm": 1.0915194749832153,
      "learning_rate": 9.753656536328529e-06,
      "loss": 0.3005,
      "step": 46
    },
    {
      "epoch": 0.746031746031746,
      "grad_norm": 2.678406238555908,
      "learning_rate": 9.737107749695456e-06,
      "loss": 0.2828,
      "step": 47
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 0.9532033801078796,
      "learning_rate": 9.72003600527246e-06,
      "loss": 0.2658,
      "step": 48
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 1.1294094324111938,
      "learning_rate": 9.702443187711991e-06,
      "loss": 0.2978,
      "step": 49
    },
    {
      "epoch": 0.7936507936507936,
      "grad_norm": 1.067834734916687,
      "learning_rate": 9.6843312391909e-06,
      "loss": 0.2941,
      "step": 50
    },
    {
      "epoch": 0.8095238095238095,
      "grad_norm": 1.0761404037475586,
      "learning_rate": 9.665702159196014e-06,
      "loss": 0.3056,
      "step": 51
    },
    {
      "epoch": 0.8253968253968254,
      "grad_norm": 4.403201103210449,
      "learning_rate": 9.646558004303419e-06,
      "loss": 0.3728,
      "step": 52
    },
    {
      "epoch": 0.8412698412698413,
      "grad_norm": 1.0028859376907349,
      "learning_rate": 9.62690088795141e-06,
      "loss": 0.2765,
      "step": 53
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 1.0655666589736938,
      "learning_rate": 9.606732980207186e-06,
      "loss": 0.2912,
      "step": 54
    },
    {
      "epoch": 0.873015873015873,
      "grad_norm": 8.566058158874512,
      "learning_rate": 9.586056507527266e-06,
      "loss": 0.5695,
      "step": 55
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 1.487505555152893,
      "learning_rate": 9.564873752511719e-06,
      "loss": 0.2906,
      "step": 56
    },
    {
      "epoch": 0.9047619047619048,
      "grad_norm": 2.458684206008911,
      "learning_rate": 9.543187053652156e-06,
      "loss": 0.4637,
      "step": 57
    },
    {
      "epoch": 0.9206349206349206,
      "grad_norm": 1.1918795108795166,
      "learning_rate": 9.520998805073583e-06,
      "loss": 0.347,
      "step": 58
    },
    {
      "epoch": 0.9365079365079365,
      "grad_norm": 1.9880917072296143,
      "learning_rate": 9.498311456270091e-06,
      "loss": 0.3115,
      "step": 59
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 1.1867977380752563,
      "learning_rate": 9.475127511834438e-06,
      "loss": 0.3172,
      "step": 60
    },
    {
      "epoch": 0.9682539682539683,
      "grad_norm": 1.2388020753860474,
      "learning_rate": 9.451449531181571e-06,
      "loss": 0.252,
      "step": 61
    },
    {
      "epoch": 0.9841269841269841,
      "grad_norm": 9.597868919372559,
      "learning_rate": 9.427280128266049e-06,
      "loss": 0.2668,
      "step": 62
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.9280756711959839,
      "learning_rate": 9.4026219712935e-06,
      "loss": 0.2107,
      "step": 63
    },
    {
      "epoch": 1.0158730158730158,
      "grad_norm": 0.9878024458885193,
      "learning_rate": 9.377477782426041e-06,
      "loss": 0.2008,
      "step": 64
    },
    {
      "epoch": 1.0317460317460316,
      "grad_norm": 2.081068992614746,
      "learning_rate": 9.351850337481774e-06,
      "loss": 0.1998,
      "step": 65
    },
    {
      "epoch": 1.0476190476190477,
      "grad_norm": 0.9242327213287354,
      "learning_rate": 9.325742465628342e-06,
      "loss": 0.1791,
      "step": 66
    },
    {
      "epoch": 1.0634920634920635,
      "grad_norm": 2.085343599319458,
      "learning_rate": 9.299157049070604e-06,
      "loss": 0.1787,
      "step": 67
    },
    {
      "epoch": 1.0793650793650793,
      "grad_norm": 0.815376877784729,
      "learning_rate": 9.272097022732444e-06,
      "loss": 0.1678,
      "step": 68
    },
    {
      "epoch": 1.0952380952380953,
      "grad_norm": 0.8281353116035461,
      "learning_rate": 9.244565373932775e-06,
      "loss": 0.1646,
      "step": 69
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.828342616558075,
      "learning_rate": 9.216565142055745e-06,
      "loss": 0.155,
      "step": 70
    },
    {
      "epoch": 1.126984126984127,
      "grad_norm": 0.77997225522995,
      "learning_rate": 9.188099418215208e-06,
      "loss": 0.1176,
      "step": 71
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 0.9490588903427124,
      "learning_rate": 9.159171344913469e-06,
      "loss": 0.1316,
      "step": 72
    },
    {
      "epoch": 1.1587301587301586,
      "grad_norm": 1.061501145362854,
      "learning_rate": 9.129784115694368e-06,
      "loss": 0.1759,
      "step": 73
    },
    {
      "epoch": 1.1746031746031746,
      "grad_norm": 1.077555537223816,
      "learning_rate": 9.09994097479073e-06,
      "loss": 0.1607,
      "step": 74
    },
    {
      "epoch": 1.1904761904761905,
      "grad_norm": 0.8213773965835571,
      "learning_rate": 9.069645216766207e-06,
      "loss": 0.1123,
      "step": 75
    },
    {
      "epoch": 1.2063492063492063,
      "grad_norm": 0.8540403246879578,
      "learning_rate": 9.038900186151574e-06,
      "loss": 0.1306,
      "step": 76
    },
    {
      "epoch": 1.2222222222222223,
      "grad_norm": 1.0110580921173096,
      "learning_rate": 9.007709277075512e-06,
      "loss": 0.1644,
      "step": 77
    },
    {
      "epoch": 1.2380952380952381,
      "grad_norm": 1.3765249252319336,
      "learning_rate": 8.976075932889896e-06,
      "loss": 0.2027,
      "step": 78
    },
    {
      "epoch": 1.253968253968254,
      "grad_norm": 1.0460267066955566,
      "learning_rate": 8.944003645789678e-06,
      "loss": 0.1635,
      "step": 79
    },
    {
      "epoch": 1.2698412698412698,
      "grad_norm": 0.8925233483314514,
      "learning_rate": 8.911495956427358e-06,
      "loss": 0.1538,
      "step": 80
    },
    {
      "epoch": 1.2857142857142856,
      "grad_norm": 1.1790034770965576,
      "learning_rate": 8.8785564535221e-06,
      "loss": 0.164,
      "step": 81
    },
    {
      "epoch": 1.3015873015873016,
      "grad_norm": 1.1329326629638672,
      "learning_rate": 8.845188773463567e-06,
      "loss": 0.1449,
      "step": 82
    },
    {
      "epoch": 1.3174603174603174,
      "grad_norm": 4.996532440185547,
      "learning_rate": 8.811396599910467e-06,
      "loss": 0.247,
      "step": 83
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.74964839220047,
      "learning_rate": 8.777183663383897e-06,
      "loss": 0.1259,
      "step": 84
    },
    {
      "epoch": 1.3492063492063493,
      "grad_norm": 1.8605432510375977,
      "learning_rate": 8.742553740855507e-06,
      "loss": 0.1935,
      "step": 85
    },
    {
      "epoch": 1.3650793650793651,
      "grad_norm": 0.9615695476531982,
      "learning_rate": 8.707510655330536e-06,
      "loss": 0.1772,
      "step": 86
    },
    {
      "epoch": 1.380952380952381,
      "grad_norm": 1.0028129816055298,
      "learning_rate": 8.672058275425773e-06,
      "loss": 0.159,
      "step": 87
    },
    {
      "epoch": 1.3968253968253967,
      "grad_norm": 0.780433177947998,
      "learning_rate": 8.636200514942466e-06,
      "loss": 0.1374,
      "step": 88
    },
    {
      "epoch": 1.4126984126984126,
      "grad_norm": 2.116007089614868,
      "learning_rate": 8.59994133243427e-06,
      "loss": 0.1951,
      "step": 89
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.8867003917694092,
      "learning_rate": 8.563284730770222e-06,
      "loss": 0.147,
      "step": 90
    },
    {
      "epoch": 1.4444444444444444,
      "grad_norm": 0.8089895248413086,
      "learning_rate": 8.52623475669285e-06,
      "loss": 0.1109,
      "step": 91
    },
    {
      "epoch": 1.4603174603174602,
      "grad_norm": 1.093917727470398,
      "learning_rate": 8.488795500371427e-06,
      "loss": 0.1328,
      "step": 92
    },
    {
      "epoch": 1.4761904761904763,
      "grad_norm": 1.0122655630111694,
      "learning_rate": 8.450971094950433e-06,
      "loss": 0.1998,
      "step": 93
    },
    {
      "epoch": 1.492063492063492,
      "grad_norm": 3.2237043380737305,
      "learning_rate": 8.412765716093273e-06,
      "loss": 0.1949,
      "step": 94
    },
    {
      "epoch": 1.507936507936508,
      "grad_norm": 0.8921442031860352,
      "learning_rate": 8.374183581521288e-06,
      "loss": 0.1659,
      "step": 95
    },
    {
      "epoch": 1.5238095238095237,
      "grad_norm": 0.9087255001068115,
      "learning_rate": 8.335228950548164e-06,
      "loss": 0.1381,
      "step": 96
    },
    {
      "epoch": 1.5396825396825395,
      "grad_norm": 0.8095484972000122,
      "learning_rate": 8.29590612360969e-06,
      "loss": 0.1418,
      "step": 97
    },
    {
      "epoch": 1.5555555555555556,
      "grad_norm": 0.9046949744224548,
      "learning_rate": 8.256219441789023e-06,
      "loss": 0.1362,
      "step": 98
    },
    {
      "epoch": 1.5714285714285714,
      "grad_norm": 1.0229334831237793,
      "learning_rate": 8.216173286337449e-06,
      "loss": 0.1548,
      "step": 99
    },
    {
      "epoch": 1.5873015873015874,
      "grad_norm": 0.7955384254455566,
      "learning_rate": 8.175772078190706e-06,
      "loss": 0.0964,
      "step": 100
    },
    {
      "epoch": 1.6031746031746033,
      "grad_norm": 0.8258505463600159,
      "learning_rate": 8.135020277480933e-06,
      "loss": 0.1436,
      "step": 101
    },
    {
      "epoch": 1.619047619047619,
      "grad_norm": 0.9820398688316345,
      "learning_rate": 8.093922383044293e-06,
      "loss": 0.1875,
      "step": 102
    },
    {
      "epoch": 1.6349206349206349,
      "grad_norm": 0.8524972796440125,
      "learning_rate": 8.052482931924307e-06,
      "loss": 0.1381,
      "step": 103
    },
    {
      "epoch": 1.6507936507936507,
      "grad_norm": 1.515363097190857,
      "learning_rate": 8.010706498870997e-06,
      "loss": 0.1783,
      "step": 104
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.9057204127311707,
      "learning_rate": 7.968597695835845e-06,
      "loss": 0.1491,
      "step": 105
    },
    {
      "epoch": 1.6825396825396826,
      "grad_norm": 0.73180091381073,
      "learning_rate": 7.926161171462647e-06,
      "loss": 0.1216,
      "step": 106
    },
    {
      "epoch": 1.6984126984126984,
      "grad_norm": 1.321354627609253,
      "learning_rate": 7.883401610574338e-06,
      "loss": 0.1496,
      "step": 107
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.9403553009033203,
      "learning_rate": 7.84032373365578e-06,
      "loss": 0.1509,
      "step": 108
    },
    {
      "epoch": 1.7301587301587302,
      "grad_norm": 0.9378424286842346,
      "learning_rate": 7.796932296332666e-06,
      "loss": 0.16,
      "step": 109
    },
    {
      "epoch": 1.746031746031746,
      "grad_norm": 0.8329445123672485,
      "learning_rate": 7.753232088846505e-06,
      "loss": 0.0972,
      "step": 110
    },
    {
      "epoch": 1.7619047619047619,
      "grad_norm": 1.2565405368804932,
      "learning_rate": 7.709227935525796e-06,
      "loss": 0.1837,
      "step": 111
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 0.9211390614509583,
      "learning_rate": 7.664924694253444e-06,
      "loss": 0.1472,
      "step": 112
    },
    {
      "epoch": 1.7936507936507935,
      "grad_norm": 0.8232191801071167,
      "learning_rate": 7.620327255930475e-06,
      "loss": 0.123,
      "step": 113
    },
    {
      "epoch": 1.8095238095238095,
      "grad_norm": 0.8965691328048706,
      "learning_rate": 7.575440543936092e-06,
      "loss": 0.1329,
      "step": 114
    },
    {
      "epoch": 1.8253968253968254,
      "grad_norm": 0.9043343663215637,
      "learning_rate": 7.530269513584158e-06,
      "loss": 0.1268,
      "step": 115
    },
    {
      "epoch": 1.8412698412698414,
      "grad_norm": 0.822960376739502,
      "learning_rate": 7.484819151576148e-06,
      "loss": 0.128,
      "step": 116
    },
    {
      "epoch": 1.8571428571428572,
      "grad_norm": 1.0179648399353027,
      "learning_rate": 7.439094475450638e-06,
      "loss": 0.1732,
      "step": 117
    },
    {
      "epoch": 1.873015873015873,
      "grad_norm": 0.888350784778595,
      "learning_rate": 7.393100533029383e-06,
      "loss": 0.1608,
      "step": 118
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 0.8138338923454285,
      "learning_rate": 7.346842401860069e-06,
      "loss": 0.1534,
      "step": 119
    },
    {
      "epoch": 1.9047619047619047,
      "grad_norm": 1.0096592903137207,
      "learning_rate": 7.300325188655762e-06,
      "loss": 0.1705,
      "step": 120
    },
    {
      "epoch": 1.9206349206349205,
      "grad_norm": 0.8139647841453552,
      "learning_rate": 7.253554028731149e-06,
      "loss": 0.1453,
      "step": 121
    },
    {
      "epoch": 1.9365079365079365,
      "grad_norm": 1.51543128490448,
      "learning_rate": 7.206534085435626e-06,
      "loss": 0.2008,
      "step": 122
    },
    {
      "epoch": 1.9523809523809523,
      "grad_norm": 0.944955587387085,
      "learning_rate": 7.159270549583278e-06,
      "loss": 0.1793,
      "step": 123
    },
    {
      "epoch": 1.9682539682539684,
      "grad_norm": 0.943600594997406,
      "learning_rate": 7.111768638879834e-06,
      "loss": 0.1783,
      "step": 124
    },
    {
      "epoch": 1.9841269841269842,
      "grad_norm": 0.8988723158836365,
      "learning_rate": 7.064033597346658e-06,
      "loss": 0.1472,
      "step": 125
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8375394940376282,
      "learning_rate": 7.016070694741824e-06,
      "loss": 0.0907,
      "step": 126
    },
    {
      "epoch": 2.015873015873016,
      "grad_norm": 0.5929964780807495,
      "learning_rate": 6.967885225978366e-06,
      "loss": 0.0541,
      "step": 127
    },
    {
      "epoch": 2.0317460317460316,
      "grad_norm": 0.7823465466499329,
      "learning_rate": 6.919482510539723e-06,
      "loss": 0.0553,
      "step": 128
    },
    {
      "epoch": 2.0476190476190474,
      "grad_norm": 0.6394754648208618,
      "learning_rate": 6.870867891892511e-06,
      "loss": 0.0539,
      "step": 129
    },
    {
      "epoch": 2.0634920634920633,
      "grad_norm": 0.6169945001602173,
      "learning_rate": 6.822046736896607e-06,
      "loss": 0.0446,
      "step": 130
    },
    {
      "epoch": 2.0793650793650795,
      "grad_norm": 0.6922060251235962,
      "learning_rate": 6.773024435212678e-06,
      "loss": 0.0474,
      "step": 131
    },
    {
      "epoch": 2.0952380952380953,
      "grad_norm": 0.6317225098609924,
      "learning_rate": 6.723806398707186e-06,
      "loss": 0.0472,
      "step": 132
    },
    {
      "epoch": 2.111111111111111,
      "grad_norm": 0.6741769909858704,
      "learning_rate": 6.674398060854931e-06,
      "loss": 0.042,
      "step": 133
    },
    {
      "epoch": 2.126984126984127,
      "grad_norm": 0.9402458071708679,
      "learning_rate": 6.624804876139227e-06,
      "loss": 0.0647,
      "step": 134
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 0.8518017530441284,
      "learning_rate": 6.57503231944974e-06,
      "loss": 0.0559,
      "step": 135
    },
    {
      "epoch": 2.1587301587301586,
      "grad_norm": 0.9338895678520203,
      "learning_rate": 6.525085885478088e-06,
      "loss": 0.0482,
      "step": 136
    },
    {
      "epoch": 2.1746031746031744,
      "grad_norm": 0.8100844025611877,
      "learning_rate": 6.4749710881112485e-06,
      "loss": 0.0599,
      "step": 137
    },
    {
      "epoch": 2.1904761904761907,
      "grad_norm": 1.0090680122375488,
      "learning_rate": 6.424693459822843e-06,
      "loss": 0.0642,
      "step": 138
    },
    {
      "epoch": 2.2063492063492065,
      "grad_norm": 0.6279119253158569,
      "learning_rate": 6.374258551062377e-06,
      "loss": 0.0514,
      "step": 139
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.6508628129959106,
      "learning_rate": 6.3236719296424985e-06,
      "loss": 0.0464,
      "step": 140
    },
    {
      "epoch": 2.238095238095238,
      "grad_norm": 0.586854100227356,
      "learning_rate": 6.272939180124316e-06,
      "loss": 0.0407,
      "step": 141
    },
    {
      "epoch": 2.253968253968254,
      "grad_norm": 0.6627777814865112,
      "learning_rate": 6.222065903200909e-06,
      "loss": 0.0487,
      "step": 142
    },
    {
      "epoch": 2.2698412698412698,
      "grad_norm": 0.7252594232559204,
      "learning_rate": 6.171057715079012e-06,
      "loss": 0.0568,
      "step": 143
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 0.5209100246429443,
      "learning_rate": 6.119920246859025e-06,
      "loss": 0.0363,
      "step": 144
    },
    {
      "epoch": 2.3015873015873014,
      "grad_norm": 0.5737756490707397,
      "learning_rate": 6.068659143913349e-06,
      "loss": 0.0303,
      "step": 145
    },
    {
      "epoch": 2.317460317460317,
      "grad_norm": 1.3126976490020752,
      "learning_rate": 6.0172800652631706e-06,
      "loss": 0.074,
      "step": 146
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.6515079140663147,
      "learning_rate": 5.965788682953717e-06,
      "loss": 0.047,
      "step": 147
    },
    {
      "epoch": 2.3492063492063493,
      "grad_norm": 0.6837308406829834,
      "learning_rate": 5.914190681428098e-06,
      "loss": 0.0354,
      "step": 148
    },
    {
      "epoch": 2.365079365079365,
      "grad_norm": 1.8536609411239624,
      "learning_rate": 5.862491756899753e-06,
      "loss": 0.0605,
      "step": 149
    },
    {
      "epoch": 2.380952380952381,
      "grad_norm": 0.7980743646621704,
      "learning_rate": 5.8106976167236236e-06,
      "loss": 0.0443,
      "step": 150
    },
    {
      "epoch": 2.3968253968253967,
      "grad_norm": 0.7513721585273743,
      "learning_rate": 5.758813978766077e-06,
      "loss": 0.0554,
      "step": 151
    },
    {
      "epoch": 2.4126984126984126,
      "grad_norm": 0.9976074695587158,
      "learning_rate": 5.706846570773677e-06,
      "loss": 0.0629,
      "step": 152
    },
    {
      "epoch": 2.4285714285714284,
      "grad_norm": 0.6811009049415588,
      "learning_rate": 5.654801129740863e-06,
      "loss": 0.0395,
      "step": 153
    },
    {
      "epoch": 2.4444444444444446,
      "grad_norm": 0.8673996329307556,
      "learning_rate": 5.6026834012766155e-06,
      "loss": 0.0517,
      "step": 154
    },
    {
      "epoch": 2.4603174603174605,
      "grad_norm": 0.6963090300559998,
      "learning_rate": 5.550499138970158e-06,
      "loss": 0.0416,
      "step": 155
    },
    {
      "epoch": 2.4761904761904763,
      "grad_norm": 0.6987242102622986,
      "learning_rate": 5.4982541037557825e-06,
      "loss": 0.0416,
      "step": 156
    },
    {
      "epoch": 2.492063492063492,
      "grad_norm": 1.312720775604248,
      "learning_rate": 5.44595406327687e-06,
      "loss": 0.0499,
      "step": 157
    },
    {
      "epoch": 2.507936507936508,
      "grad_norm": 0.7223400473594666,
      "learning_rate": 5.393604791249158e-06,
      "loss": 0.0617,
      "step": 158
    },
    {
      "epoch": 2.5238095238095237,
      "grad_norm": 0.5593965649604797,
      "learning_rate": 5.341212066823356e-06,
      "loss": 0.0341,
      "step": 159
    },
    {
      "epoch": 2.5396825396825395,
      "grad_norm": 0.80050128698349,
      "learning_rate": 5.288781673947143e-06,
      "loss": 0.0818,
      "step": 160
    },
    {
      "epoch": 2.5555555555555554,
      "grad_norm": 0.9423879981040955,
      "learning_rate": 5.2363194007266435e-06,
      "loss": 0.0533,
      "step": 161
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.5583639740943909,
      "learning_rate": 5.183831038787449e-06,
      "loss": 0.0308,
      "step": 162
    },
    {
      "epoch": 2.5873015873015874,
      "grad_norm": 0.9041386842727661,
      "learning_rate": 5.131322382635236e-06,
      "loss": 0.0393,
      "step": 163
    },
    {
      "epoch": 2.6031746031746033,
      "grad_norm": 0.7514568567276001,
      "learning_rate": 5.078799229016083e-06,
      "loss": 0.0379,
      "step": 164
    },
    {
      "epoch": 2.619047619047619,
      "grad_norm": 0.6620778441429138,
      "learning_rate": 5.0262673762765316e-06,
      "loss": 0.042,
      "step": 165
    },
    {
      "epoch": 2.634920634920635,
      "grad_norm": 0.7500037550926208,
      "learning_rate": 4.973732623723471e-06,
      "loss": 0.0494,
      "step": 166
    },
    {
      "epoch": 2.6507936507936507,
      "grad_norm": 0.6531999707221985,
      "learning_rate": 4.921200770983919e-06,
      "loss": 0.0364,
      "step": 167
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 1.0730338096618652,
      "learning_rate": 4.8686776173647655e-06,
      "loss": 0.0554,
      "step": 168
    },
    {
      "epoch": 2.682539682539683,
      "grad_norm": 0.6723483204841614,
      "learning_rate": 4.816168961212553e-06,
      "loss": 0.0532,
      "step": 169
    },
    {
      "epoch": 2.6984126984126986,
      "grad_norm": 0.7021626234054565,
      "learning_rate": 4.763680599273357e-06,
      "loss": 0.0611,
      "step": 170
    },
    {
      "epoch": 2.7142857142857144,
      "grad_norm": 0.7522814273834229,
      "learning_rate": 4.711218326052859e-06,
      "loss": 0.0426,
      "step": 171
    },
    {
      "epoch": 2.7301587301587302,
      "grad_norm": 0.6586665511131287,
      "learning_rate": 4.6587879331766465e-06,
      "loss": 0.0368,
      "step": 172
    },
    {
      "epoch": 2.746031746031746,
      "grad_norm": 1.2219913005828857,
      "learning_rate": 4.606395208750844e-06,
      "loss": 0.0852,
      "step": 173
    },
    {
      "epoch": 2.761904761904762,
      "grad_norm": 0.6554266810417175,
      "learning_rate": 4.554045936723132e-06,
      "loss": 0.0516,
      "step": 174
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 0.5427403450012207,
      "learning_rate": 4.501745896244219e-06,
      "loss": 0.0388,
      "step": 175
    },
    {
      "epoch": 2.7936507936507935,
      "grad_norm": 0.5815466046333313,
      "learning_rate": 4.4495008610298435e-06,
      "loss": 0.0452,
      "step": 176
    },
    {
      "epoch": 2.8095238095238093,
      "grad_norm": 0.957595944404602,
      "learning_rate": 4.397316598723385e-06,
      "loss": 0.0584,
      "step": 177
    },
    {
      "epoch": 2.825396825396825,
      "grad_norm": 0.6840035915374756,
      "learning_rate": 4.345198870259139e-06,
      "loss": 0.0503,
      "step": 178
    },
    {
      "epoch": 2.8412698412698414,
      "grad_norm": 0.7481380105018616,
      "learning_rate": 4.2931534292263265e-06,
      "loss": 0.0561,
      "step": 179
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.5073666572570801,
      "learning_rate": 4.241186021233925e-06,
      "loss": 0.0339,
      "step": 180
    },
    {
      "epoch": 2.873015873015873,
      "grad_norm": 0.6581488251686096,
      "learning_rate": 4.189302383276378e-06,
      "loss": 0.0629,
      "step": 181
    },
    {
      "epoch": 2.888888888888889,
      "grad_norm": 0.623148500919342,
      "learning_rate": 4.137508243100249e-06,
      "loss": 0.0521,
      "step": 182
    },
    {
      "epoch": 2.9047619047619047,
      "grad_norm": 0.621598482131958,
      "learning_rate": 4.085809318571905e-06,
      "loss": 0.0458,
      "step": 183
    },
    {
      "epoch": 2.9206349206349205,
      "grad_norm": 0.6772786974906921,
      "learning_rate": 4.034211317046285e-06,
      "loss": 0.044,
      "step": 184
    },
    {
      "epoch": 2.9365079365079367,
      "grad_norm": 1.262326955795288,
      "learning_rate": 3.982719934736832e-06,
      "loss": 0.046,
      "step": 185
    },
    {
      "epoch": 2.9523809523809526,
      "grad_norm": 0.6345784068107605,
      "learning_rate": 3.931340856086652e-06,
      "loss": 0.042,
      "step": 186
    },
    {
      "epoch": 2.9682539682539684,
      "grad_norm": 0.5799025297164917,
      "learning_rate": 3.880079753140978e-06,
      "loss": 0.0467,
      "step": 187
    },
    {
      "epoch": 2.984126984126984,
      "grad_norm": 0.6458066701889038,
      "learning_rate": 3.82894228492099e-06,
      "loss": 0.047,
      "step": 188
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.5344202518463135,
      "learning_rate": 3.777934096799094e-06,
      "loss": 0.0418,
      "step": 189
    },
    {
      "epoch": 3.015873015873016,
      "grad_norm": 0.3208453953266144,
      "learning_rate": 3.7270608198756852e-06,
      "loss": 0.0124,
      "step": 190
    },
    {
      "epoch": 3.0317460317460316,
      "grad_norm": 0.48605024814605713,
      "learning_rate": 3.676328070357503e-06,
      "loss": 0.0332,
      "step": 191
    },
    {
      "epoch": 3.0476190476190474,
      "grad_norm": 0.5282842516899109,
      "learning_rate": 3.6257414489376217e-06,
      "loss": 0.0155,
      "step": 192
    },
    {
      "epoch": 3.0634920634920633,
      "grad_norm": 0.3682861030101776,
      "learning_rate": 3.5753065401771577e-06,
      "loss": 0.017,
      "step": 193
    },
    {
      "epoch": 3.0793650793650795,
      "grad_norm": 0.4284667372703552,
      "learning_rate": 3.5250289118887515e-06,
      "loss": 0.0107,
      "step": 194
    },
    {
      "epoch": 3.0952380952380953,
      "grad_norm": 0.36055704951286316,
      "learning_rate": 3.4749141145219118e-06,
      "loss": 0.0116,
      "step": 195
    },
    {
      "epoch": 3.111111111111111,
      "grad_norm": 0.42703211307525635,
      "learning_rate": 3.424967680550261e-06,
      "loss": 0.0089,
      "step": 196
    },
    {
      "epoch": 3.126984126984127,
      "grad_norm": 0.4422130882740021,
      "learning_rate": 3.3751951238607745e-06,
      "loss": 0.015,
      "step": 197
    },
    {
      "epoch": 3.142857142857143,
      "grad_norm": 1.8617546558380127,
      "learning_rate": 3.3256019391450696e-06,
      "loss": 0.044,
      "step": 198
    },
    {
      "epoch": 3.1587301587301586,
      "grad_norm": 0.4572526216506958,
      "learning_rate": 3.2761936012928147e-06,
      "loss": 0.0092,
      "step": 199
    },
    {
      "epoch": 3.1746031746031744,
      "grad_norm": 0.606041431427002,
      "learning_rate": 3.226975564787322e-06,
      "loss": 0.013,
      "step": 200
    },
    {
      "epoch": 3.1904761904761907,
      "grad_norm": 0.6772161722183228,
      "learning_rate": 3.177953263103394e-06,
      "loss": 0.0233,
      "step": 201
    },
    {
      "epoch": 3.2063492063492065,
      "grad_norm": 0.4099721908569336,
      "learning_rate": 3.1291321081074887e-06,
      "loss": 0.0074,
      "step": 202
    },
    {
      "epoch": 3.2222222222222223,
      "grad_norm": 0.34896424412727356,
      "learning_rate": 3.0805174894602775e-06,
      "loss": 0.0081,
      "step": 203
    },
    {
      "epoch": 3.238095238095238,
      "grad_norm": 0.5158839821815491,
      "learning_rate": 3.032114774021636e-06,
      "loss": 0.0174,
      "step": 204
    },
    {
      "epoch": 3.253968253968254,
      "grad_norm": 0.6538414359092712,
      "learning_rate": 2.9839293052581767e-06,
      "loss": 0.0188,
      "step": 205
    },
    {
      "epoch": 3.2698412698412698,
      "grad_norm": 0.4532378911972046,
      "learning_rate": 2.9359664026533443e-06,
      "loss": 0.0123,
      "step": 206
    },
    {
      "epoch": 3.2857142857142856,
      "grad_norm": 0.4588467478752136,
      "learning_rate": 2.8882313611201684e-06,
      "loss": 0.0154,
      "step": 207
    },
    {
      "epoch": 3.3015873015873014,
      "grad_norm": 0.6380066871643066,
      "learning_rate": 2.8407294504167238e-06,
      "loss": 0.0168,
      "step": 208
    },
    {
      "epoch": 3.317460317460317,
      "grad_norm": 0.3832227289676666,
      "learning_rate": 2.793465914564375e-06,
      "loss": 0.0118,
      "step": 209
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.4849812090396881,
      "learning_rate": 2.7464459712688517e-06,
      "loss": 0.0156,
      "step": 210
    },
    {
      "epoch": 3.3492063492063493,
      "grad_norm": 0.5382815003395081,
      "learning_rate": 2.6996748113442397e-06,
      "loss": 0.0162,
      "step": 211
    },
    {
      "epoch": 3.365079365079365,
      "grad_norm": 0.4217240810394287,
      "learning_rate": 2.653157598139932e-06,
      "loss": 0.0144,
      "step": 212
    },
    {
      "epoch": 3.380952380952381,
      "grad_norm": 0.3669329881668091,
      "learning_rate": 2.6068994669706184e-06,
      "loss": 0.0068,
      "step": 213
    },
    {
      "epoch": 3.3968253968253967,
      "grad_norm": 0.3910522162914276,
      "learning_rate": 2.560905524549364e-06,
      "loss": 0.0134,
      "step": 214
    },
    {
      "epoch": 3.4126984126984126,
      "grad_norm": 0.3835786283016205,
      "learning_rate": 2.515180848423853e-06,
      "loss": 0.009,
      "step": 215
    },
    {
      "epoch": 3.4285714285714284,
      "grad_norm": 0.39538145065307617,
      "learning_rate": 2.469730486415842e-06,
      "loss": 0.0154,
      "step": 216
    },
    {
      "epoch": 3.4444444444444446,
      "grad_norm": 0.48559364676475525,
      "learning_rate": 2.4245594560639086e-06,
      "loss": 0.0159,
      "step": 217
    },
    {
      "epoch": 3.4603174603174605,
      "grad_norm": 0.6111780405044556,
      "learning_rate": 2.379672744069527e-06,
      "loss": 0.0216,
      "step": 218
    },
    {
      "epoch": 3.4761904761904763,
      "grad_norm": 0.48673927783966064,
      "learning_rate": 2.335075305746558e-06,
      "loss": 0.0124,
      "step": 219
    },
    {
      "epoch": 3.492063492063492,
      "grad_norm": 0.4308861792087555,
      "learning_rate": 2.2907720644742064e-06,
      "loss": 0.0112,
      "step": 220
    },
    {
      "epoch": 3.507936507936508,
      "grad_norm": 0.39692750573158264,
      "learning_rate": 2.2467679111534963e-06,
      "loss": 0.0162,
      "step": 221
    },
    {
      "epoch": 3.5238095238095237,
      "grad_norm": 0.4090782701969147,
      "learning_rate": 2.2030677036673345e-06,
      "loss": 0.0214,
      "step": 222
    },
    {
      "epoch": 3.5396825396825395,
      "grad_norm": 0.4135613441467285,
      "learning_rate": 2.159676266344222e-06,
      "loss": 0.0157,
      "step": 223
    },
    {
      "epoch": 3.5555555555555554,
      "grad_norm": 0.8922034502029419,
      "learning_rate": 2.1165983894256647e-06,
      "loss": 0.0355,
      "step": 224
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 0.25958430767059326,
      "learning_rate": 2.0738388285373532e-06,
      "loss": 0.0057,
      "step": 225
    },
    {
      "epoch": 3.5873015873015874,
      "grad_norm": 0.3267192244529724,
      "learning_rate": 2.0314023041641567e-06,
      "loss": 0.0089,
      "step": 226
    },
    {
      "epoch": 3.6031746031746033,
      "grad_norm": 0.4624428451061249,
      "learning_rate": 1.9892935011290037e-06,
      "loss": 0.0159,
      "step": 227
    },
    {
      "epoch": 3.619047619047619,
      "grad_norm": 0.2781091630458832,
      "learning_rate": 1.947517068075694e-06,
      "loss": 0.0083,
      "step": 228
    },
    {
      "epoch": 3.634920634920635,
      "grad_norm": 0.31553471088409424,
      "learning_rate": 1.9060776169557083e-06,
      "loss": 0.0137,
      "step": 229
    },
    {
      "epoch": 3.6507936507936507,
      "grad_norm": 0.5683518052101135,
      "learning_rate": 1.864979722519068e-06,
      "loss": 0.0298,
      "step": 230
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.3412047028541565,
      "learning_rate": 1.8242279218092968e-06,
      "loss": 0.0137,
      "step": 231
    },
    {
      "epoch": 3.682539682539683,
      "grad_norm": 0.5104885101318359,
      "learning_rate": 1.7838267136625536e-06,
      "loss": 0.0114,
      "step": 232
    },
    {
      "epoch": 3.6984126984126986,
      "grad_norm": 0.4502926170825958,
      "learning_rate": 1.743780558210979e-06,
      "loss": 0.013,
      "step": 233
    },
    {
      "epoch": 3.7142857142857144,
      "grad_norm": 0.48987045884132385,
      "learning_rate": 1.704093876390312e-06,
      "loss": 0.0217,
      "step": 234
    },
    {
      "epoch": 3.7301587301587302,
      "grad_norm": 0.3420720398426056,
      "learning_rate": 1.664771049451837e-06,
      "loss": 0.0089,
      "step": 235
    },
    {
      "epoch": 3.746031746031746,
      "grad_norm": 0.3107622563838959,
      "learning_rate": 1.6258164184787123e-06,
      "loss": 0.0085,
      "step": 236
    },
    {
      "epoch": 3.761904761904762,
      "grad_norm": 0.5032781958580017,
      "learning_rate": 1.5872342839067305e-06,
      "loss": 0.0075,
      "step": 237
    },
    {
      "epoch": 3.7777777777777777,
      "grad_norm": 0.3874753415584564,
      "learning_rate": 1.5490289050495678e-06,
      "loss": 0.0108,
      "step": 238
    },
    {
      "epoch": 3.7936507936507935,
      "grad_norm": 0.30904141068458557,
      "learning_rate": 1.511204499628574e-06,
      "loss": 0.0086,
      "step": 239
    },
    {
      "epoch": 3.8095238095238093,
      "grad_norm": 0.39750581979751587,
      "learning_rate": 1.4737652433071515e-06,
      "loss": 0.0135,
      "step": 240
    },
    {
      "epoch": 3.825396825396825,
      "grad_norm": 0.5290718674659729,
      "learning_rate": 1.4367152692297799e-06,
      "loss": 0.0187,
      "step": 241
    },
    {
      "epoch": 3.8412698412698414,
      "grad_norm": 0.4600606858730316,
      "learning_rate": 1.4000586675657312e-06,
      "loss": 0.0111,
      "step": 242
    },
    {
      "epoch": 3.857142857142857,
      "grad_norm": 0.461280882358551,
      "learning_rate": 1.3637994850575342e-06,
      "loss": 0.0147,
      "step": 243
    },
    {
      "epoch": 3.873015873015873,
      "grad_norm": 0.5423941016197205,
      "learning_rate": 1.3279417245742288e-06,
      "loss": 0.0236,
      "step": 244
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 0.39030736684799194,
      "learning_rate": 1.2924893446694648e-06,
      "loss": 0.006,
      "step": 245
    },
    {
      "epoch": 3.9047619047619047,
      "grad_norm": 0.5754550099372864,
      "learning_rate": 1.257446259144494e-06,
      "loss": 0.0286,
      "step": 246
    },
    {
      "epoch": 3.9206349206349205,
      "grad_norm": 0.35015997290611267,
      "learning_rate": 1.222816336616104e-06,
      "loss": 0.0143,
      "step": 247
    },
    {
      "epoch": 3.9365079365079367,
      "grad_norm": 0.4731229245662689,
      "learning_rate": 1.1886034000895341e-06,
      "loss": 0.0177,
      "step": 248
    },
    {
      "epoch": 3.9523809523809526,
      "grad_norm": 0.3958592116832733,
      "learning_rate": 1.1548112265364336e-06,
      "loss": 0.0132,
      "step": 249
    },
    {
      "epoch": 3.9682539682539684,
      "grad_norm": 0.4273618161678314,
      "learning_rate": 1.1214435464779006e-06,
      "loss": 0.0206,
      "step": 250
    },
    {
      "epoch": 3.984126984126984,
      "grad_norm": 0.429865300655365,
      "learning_rate": 1.088504043572643e-06,
      "loss": 0.0131,
      "step": 251
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.28878334164619446,
      "learning_rate": 1.055996354210323e-06,
      "loss": 0.0079,
      "step": 252
    },
    {
      "epoch": 4.015873015873016,
      "grad_norm": 0.12734107673168182,
      "learning_rate": 1.0239240671101065e-06,
      "loss": 0.0027,
      "step": 253
    },
    {
      "epoch": 4.031746031746032,
      "grad_norm": 0.24097059667110443,
      "learning_rate": 9.922907229244905e-07,
      "loss": 0.0063,
      "step": 254
    },
    {
      "epoch": 4.0476190476190474,
      "grad_norm": 0.18659502267837524,
      "learning_rate": 9.610998138484262e-07,
      "loss": 0.0032,
      "step": 255
    },
    {
      "epoch": 4.063492063492063,
      "grad_norm": 0.16850638389587402,
      "learning_rate": 9.303547832337934e-07,
      "loss": 0.0036,
      "step": 256
    },
    {
      "epoch": 4.079365079365079,
      "grad_norm": 0.13990715146064758,
      "learning_rate": 9.000590252092701e-07,
      "loss": 0.0034,
      "step": 257
    },
    {
      "epoch": 4.095238095238095,
      "grad_norm": 0.16486185789108276,
      "learning_rate": 8.702158843056319e-07,
      "loss": 0.0029,
      "step": 258
    },
    {
      "epoch": 4.111111111111111,
      "grad_norm": 0.13398580253124237,
      "learning_rate": 8.408286550865319e-07,
      "loss": 0.0041,
      "step": 259
    },
    {
      "epoch": 4.1269841269841265,
      "grad_norm": 0.14831550419330597,
      "learning_rate": 8.119005817847924e-07,
      "loss": 0.0039,
      "step": 260
    },
    {
      "epoch": 4.142857142857143,
      "grad_norm": 0.1601114124059677,
      "learning_rate": 7.834348579442552e-07,
      "loss": 0.0034,
      "step": 261
    },
    {
      "epoch": 4.158730158730159,
      "grad_norm": 0.2705823481082916,
      "learning_rate": 7.554346260672263e-07,
      "loss": 0.0054,
      "step": 262
    },
    {
      "epoch": 4.174603174603175,
      "grad_norm": 0.17291611433029175,
      "learning_rate": 7.279029772675572e-07,
      "loss": 0.0025,
      "step": 263
    },
    {
      "epoch": 4.190476190476191,
      "grad_norm": 0.11752714961767197,
      "learning_rate": 7.008429509293979e-07,
      "loss": 0.002,
      "step": 264
    },
    {
      "epoch": 4.2063492063492065,
      "grad_norm": 0.2471189945936203,
      "learning_rate": 6.742575343716584e-07,
      "loss": 0.0053,
      "step": 265
    },
    {
      "epoch": 4.222222222222222,
      "grad_norm": 0.2652353048324585,
      "learning_rate": 6.481496625182271e-07,
      "loss": 0.0035,
      "step": 266
    },
    {
      "epoch": 4.238095238095238,
      "grad_norm": 0.5520457029342651,
      "learning_rate": 6.225222175739598e-07,
      "loss": 0.0094,
      "step": 267
    },
    {
      "epoch": 4.253968253968254,
      "grad_norm": 0.1428154557943344,
      "learning_rate": 5.973780287065007e-07,
      "loss": 0.002,
      "step": 268
    },
    {
      "epoch": 4.26984126984127,
      "grad_norm": 0.38164809346199036,
      "learning_rate": 5.727198717339511e-07,
      "loss": 0.0185,
      "step": 269
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.2093685418367386,
      "learning_rate": 5.485504688184307e-07,
      "loss": 0.0016,
      "step": 270
    },
    {
      "epoch": 4.301587301587301,
      "grad_norm": 0.15652765333652496,
      "learning_rate": 5.24872488165562e-07,
      "loss": 0.0029,
      "step": 271
    },
    {
      "epoch": 4.317460317460317,
      "grad_norm": 0.17758126556873322,
      "learning_rate": 5.016885437299113e-07,
      "loss": 0.003,
      "step": 272
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.18992979824543,
      "learning_rate": 4.790011949264173e-07,
      "loss": 0.0064,
      "step": 273
    },
    {
      "epoch": 4.349206349206349,
      "grad_norm": 0.10369639843702316,
      "learning_rate": 4.5681294634784437e-07,
      "loss": 0.0014,
      "step": 274
    },
    {
      "epoch": 4.365079365079365,
      "grad_norm": 0.09521545469760895,
      "learning_rate": 4.3512624748828225e-07,
      "loss": 0.0012,
      "step": 275
    },
    {
      "epoch": 4.380952380952381,
      "grad_norm": 0.310982346534729,
      "learning_rate": 4.139434924727359e-07,
      "loss": 0.0059,
      "step": 276
    },
    {
      "epoch": 4.396825396825397,
      "grad_norm": 0.3128473162651062,
      "learning_rate": 3.9326701979281623e-07,
      "loss": 0.0059,
      "step": 277
    },
    {
      "epoch": 4.412698412698413,
      "grad_norm": 0.33223262429237366,
      "learning_rate": 3.7309911204858997e-07,
      "loss": 0.0031,
      "step": 278
    },
    {
      "epoch": 4.428571428571429,
      "grad_norm": 0.13230553269386292,
      "learning_rate": 3.534419956965823e-07,
      "loss": 0.0024,
      "step": 279
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 0.20840996503829956,
      "learning_rate": 3.3429784080398765e-07,
      "loss": 0.005,
      "step": 280
    },
    {
      "epoch": 4.4603174603174605,
      "grad_norm": 0.21250075101852417,
      "learning_rate": 3.1566876080910193e-07,
      "loss": 0.0042,
      "step": 281
    },
    {
      "epoch": 4.476190476190476,
      "grad_norm": 0.14549191296100616,
      "learning_rate": 2.9755681228800904e-07,
      "loss": 0.0023,
      "step": 282
    },
    {
      "epoch": 4.492063492063492,
      "grad_norm": 0.27009841799736023,
      "learning_rate": 2.799639947275412e-07,
      "loss": 0.0037,
      "step": 283
    },
    {
      "epoch": 4.507936507936508,
      "grad_norm": 0.18486513197422028,
      "learning_rate": 2.6289225030454556e-07,
      "loss": 0.0025,
      "step": 284
    },
    {
      "epoch": 4.523809523809524,
      "grad_norm": 0.2685333490371704,
      "learning_rate": 2.4634346367147233e-07,
      "loss": 0.0038,
      "step": 285
    },
    {
      "epoch": 4.5396825396825395,
      "grad_norm": 0.22350570559501648,
      "learning_rate": 2.303194617483212e-07,
      "loss": 0.0052,
      "step": 286
    },
    {
      "epoch": 4.555555555555555,
      "grad_norm": 0.1270400732755661,
      "learning_rate": 2.1482201352095277e-07,
      "loss": 0.0037,
      "step": 287
    },
    {
      "epoch": 4.571428571428571,
      "grad_norm": 0.14703014492988586,
      "learning_rate": 1.998528298458019e-07,
      "loss": 0.0026,
      "step": 288
    },
    {
      "epoch": 4.587301587301587,
      "grad_norm": 0.05500807240605354,
      "learning_rate": 1.8541356326100436e-07,
      "loss": 0.0006,
      "step": 289
    },
    {
      "epoch": 4.603174603174603,
      "grad_norm": 0.23130561411380768,
      "learning_rate": 1.7150580780396385e-07,
      "loss": 0.0041,
      "step": 290
    },
    {
      "epoch": 4.619047619047619,
      "grad_norm": 0.14613750576972961,
      "learning_rate": 1.5813109883537792e-07,
      "loss": 0.0022,
      "step": 291
    },
    {
      "epoch": 4.634920634920634,
      "grad_norm": 0.19623324275016785,
      "learning_rate": 1.4529091286973994e-07,
      "loss": 0.0029,
      "step": 292
    },
    {
      "epoch": 4.650793650793651,
      "grad_norm": 0.2580835819244385,
      "learning_rate": 1.3298666741233424e-07,
      "loss": 0.0047,
      "step": 293
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.19105015695095062,
      "learning_rate": 1.2121972080275378e-07,
      "loss": 0.0041,
      "step": 294
    },
    {
      "epoch": 4.682539682539683,
      "grad_norm": 0.18912175297737122,
      "learning_rate": 1.0999137206494315e-07,
      "loss": 0.0045,
      "step": 295
    },
    {
      "epoch": 4.698412698412699,
      "grad_norm": 0.2492668479681015,
      "learning_rate": 9.93028607637908e-08,
      "loss": 0.0067,
      "step": 296
    },
    {
      "epoch": 4.714285714285714,
      "grad_norm": 0.1554499715566635,
      "learning_rate": 8.915536686828764e-08,
      "loss": 0.0085,
      "step": 297
    },
    {
      "epoch": 4.73015873015873,
      "grad_norm": 0.16742238402366638,
      "learning_rate": 7.955001062125989e-08,
      "loss": 0.0036,
      "step": 298
    },
    {
      "epoch": 4.746031746031746,
      "grad_norm": 0.1783057302236557,
      "learning_rate": 7.048785241570321e-08,
      "loss": 0.0028,
      "step": 299
    },
    {
      "epoch": 4.761904761904762,
      "grad_norm": 0.1566859483718872,
      "learning_rate": 6.19698926777168e-08,
      "loss": 0.0033,
      "step": 300
    },
    {
      "epoch": 4.777777777777778,
      "grad_norm": 0.1521420031785965,
      "learning_rate": 5.399707175606117e-08,
      "loss": 0.0016,
      "step": 301
    },
    {
      "epoch": 4.7936507936507935,
      "grad_norm": 0.17157761752605438,
      "learning_rate": 4.657026981834623e-08,
      "loss": 0.002,
      "step": 302
    },
    {
      "epoch": 4.809523809523809,
      "grad_norm": 0.1927778720855713,
      "learning_rate": 3.9690306753866204e-08,
      "loss": 0.0029,
      "step": 303
    },
    {
      "epoch": 4.825396825396825,
      "grad_norm": 0.15908092260360718,
      "learning_rate": 3.3357942083085404e-08,
      "loss": 0.0016,
      "step": 304
    },
    {
      "epoch": 4.841269841269841,
      "grad_norm": 0.175292506814003,
      "learning_rate": 2.7573874873791372e-08,
      "loss": 0.0043,
      "step": 305
    },
    {
      "epoch": 4.857142857142857,
      "grad_norm": 0.121407650411129,
      "learning_rate": 2.233874366391997e-08,
      "loss": 0.001,
      "step": 306
    },
    {
      "epoch": 4.8730158730158735,
      "grad_norm": 0.2875802218914032,
      "learning_rate": 1.7653126391063425e-08,
      "loss": 0.005,
      "step": 307
    },
    {
      "epoch": 4.888888888888889,
      "grad_norm": 0.18822738528251648,
      "learning_rate": 1.3517540328669143e-08,
      "loss": 0.0025,
      "step": 308
    },
    {
      "epoch": 4.904761904761905,
      "grad_norm": 0.15360276401042938,
      "learning_rate": 9.93244202893262e-09,
      "loss": 0.0023,
      "step": 309
    },
    {
      "epoch": 4.920634920634921,
      "grad_norm": 0.29353615641593933,
      "learning_rate": 6.898227272398306e-09,
      "loss": 0.0065,
      "step": 310
    },
    {
      "epoch": 4.936507936507937,
      "grad_norm": 0.5181974172592163,
      "learning_rate": 4.415231024265665e-09,
      "loss": 0.0108,
      "step": 311
    },
    {
      "epoch": 4.9523809523809526,
      "grad_norm": 0.15942662954330444,
      "learning_rate": 2.4837273974115393e-09,
      "loss": 0.0026,
      "step": 312
    },
    {
      "epoch": 4.968253968253968,
      "grad_norm": 0.1265093982219696,
      "learning_rate": 1.1039296221276863e-09,
      "loss": 0.0116,
      "step": 313
    },
    {
      "epoch": 4.984126984126984,
      "grad_norm": 0.29910922050476074,
      "learning_rate": 2.7599002258127395e-10,
      "loss": 0.0022,
      "step": 314
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.09251413494348526,
      "learning_rate": 0.0,
      "loss": 0.0015,
      "step": 315
    },
    {
      "epoch": 5.0,
      "step": 315,
      "total_flos": 32588508020736.0,
      "train_loss": 0.11869781007546754,
      "train_runtime": 5355.6377,
      "train_samples_per_second": 0.934,
      "train_steps_per_second": 0.059
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 315,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 32588508020736.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}