{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9936034115138592,
  "eval_steps": 500,
  "global_step": 234,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01279317697228145,
      "grad_norm": 6.832852840423584,
      "learning_rate": 4.1666666666666667e-07,
      "loss": 1.0868,
      "step": 1
    },
    {
      "epoch": 0.0255863539445629,
      "grad_norm": 6.904089450836182,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.1044,
      "step": 2
    },
    {
      "epoch": 0.03837953091684435,
      "grad_norm": 6.991506576538086,
      "learning_rate": 1.25e-06,
      "loss": 1.1401,
      "step": 3
    },
    {
      "epoch": 0.0511727078891258,
      "grad_norm": 6.811994552612305,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.1168,
      "step": 4
    },
    {
      "epoch": 0.06396588486140725,
      "grad_norm": 6.493495464324951,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 1.0934,
      "step": 5
    },
    {
      "epoch": 0.0767590618336887,
      "grad_norm": 5.279635429382324,
      "learning_rate": 2.5e-06,
      "loss": 1.081,
      "step": 6
    },
    {
      "epoch": 0.08955223880597014,
      "grad_norm": 4.791995525360107,
      "learning_rate": 2.916666666666667e-06,
      "loss": 1.0356,
      "step": 7
    },
    {
      "epoch": 0.1023454157782516,
      "grad_norm": 2.9714953899383545,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.9912,
      "step": 8
    },
    {
      "epoch": 0.11513859275053305,
      "grad_norm": 2.6682045459747314,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.947,
      "step": 9
    },
    {
      "epoch": 0.1279317697228145,
      "grad_norm": 2.5980541706085205,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.9328,
      "step": 10
    },
    {
      "epoch": 0.14072494669509594,
      "grad_norm": 4.253756999969482,
      "learning_rate": 4.583333333333333e-06,
      "loss": 0.9871,
      "step": 11
    },
    {
      "epoch": 0.1535181236673774,
      "grad_norm": 4.215671062469482,
      "learning_rate": 5e-06,
      "loss": 0.9252,
      "step": 12
    },
    {
      "epoch": 0.16631130063965885,
      "grad_norm": 4.390238285064697,
      "learning_rate": 5.416666666666667e-06,
      "loss": 0.9241,
      "step": 13
    },
    {
      "epoch": 0.1791044776119403,
      "grad_norm": 3.742227077484131,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.9443,
      "step": 14
    },
    {
      "epoch": 0.19189765458422176,
      "grad_norm": 3.979121446609497,
      "learning_rate": 6.25e-06,
      "loss": 0.946,
      "step": 15
    },
    {
      "epoch": 0.2046908315565032,
      "grad_norm": 2.8892107009887695,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.8909,
      "step": 16
    },
    {
      "epoch": 0.21748400852878466,
      "grad_norm": 2.102924346923828,
      "learning_rate": 7.083333333333335e-06,
      "loss": 0.8488,
      "step": 17
    },
    {
      "epoch": 0.2302771855010661,
      "grad_norm": 2.0470516681671143,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.9146,
      "step": 18
    },
    {
      "epoch": 0.24307036247334754,
      "grad_norm": 1.9300282001495361,
      "learning_rate": 7.916666666666667e-06,
      "loss": 0.8769,
      "step": 19
    },
    {
      "epoch": 0.255863539445629,
      "grad_norm": 1.6612228155136108,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.8631,
      "step": 20
    },
    {
      "epoch": 0.26865671641791045,
      "grad_norm": 1.4431664943695068,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.8269,
      "step": 21
    },
    {
      "epoch": 0.2814498933901919,
      "grad_norm": 1.2628264427185059,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.8353,
      "step": 22
    },
    {
      "epoch": 0.2942430703624733,
      "grad_norm": 1.1078248023986816,
      "learning_rate": 9.583333333333335e-06,
      "loss": 0.8076,
      "step": 23
    },
    {
      "epoch": 0.3070362473347548,
      "grad_norm": 1.111648440361023,
      "learning_rate": 1e-05,
      "loss": 0.9005,
      "step": 24
    },
    {
      "epoch": 0.31982942430703626,
      "grad_norm": 0.8681544065475464,
      "learning_rate": 9.999440509051367e-06,
      "loss": 0.8255,
      "step": 25
    },
    {
      "epoch": 0.3326226012793177,
      "grad_norm": 1.1207367181777954,
      "learning_rate": 9.997762161417517e-06,
      "loss": 0.8363,
      "step": 26
    },
    {
      "epoch": 0.34541577825159914,
      "grad_norm": 0.9773739576339722,
      "learning_rate": 9.994965332706574e-06,
      "loss": 0.7949,
      "step": 27
    },
    {
      "epoch": 0.3582089552238806,
      "grad_norm": 0.7835165858268738,
      "learning_rate": 9.991050648838676e-06,
      "loss": 0.7652,
      "step": 28
    },
    {
      "epoch": 0.37100213219616207,
      "grad_norm": 1.0131946802139282,
      "learning_rate": 9.986018985905901e-06,
      "loss": 0.8445,
      "step": 29
    },
    {
      "epoch": 0.3837953091684435,
      "grad_norm": 0.8833023905754089,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.7808,
      "step": 30
    },
    {
      "epoch": 0.39658848614072495,
      "grad_norm": 0.8186467289924622,
      "learning_rate": 9.972609476841368e-06,
      "loss": 0.8133,
      "step": 31
    },
    {
      "epoch": 0.4093816631130064,
      "grad_norm": 0.8074617385864258,
      "learning_rate": 9.964234631709188e-06,
      "loss": 0.7957,
      "step": 32
    },
    {
      "epoch": 0.42217484008528783,
      "grad_norm": 0.8177934288978577,
      "learning_rate": 9.954748808839675e-06,
      "loss": 0.8108,
      "step": 33
    },
    {
      "epoch": 0.4349680170575693,
      "grad_norm": 0.5792050957679749,
      "learning_rate": 9.944154131125643e-06,
      "loss": 0.7906,
      "step": 34
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 0.7301996946334839,
      "learning_rate": 9.932452969617607e-06,
      "loss": 0.8191,
      "step": 35
    },
    {
      "epoch": 0.4605543710021322,
      "grad_norm": 0.6449064612388611,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.7929,
      "step": 36
    },
    {
      "epoch": 0.47334754797441364,
      "grad_norm": 0.6409065127372742,
      "learning_rate": 9.905741916970863e-06,
      "loss": 0.853,
      "step": 37
    },
    {
      "epoch": 0.4861407249466951,
      "grad_norm": 0.5154856443405151,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.7733,
      "step": 38
    },
    {
      "epoch": 0.4989339019189765,
      "grad_norm": 0.6274257302284241,
      "learning_rate": 9.874639560909118e-06,
      "loss": 0.8247,
      "step": 39
    },
    {
      "epoch": 0.511727078891258,
      "grad_norm": 0.636122465133667,
      "learning_rate": 9.857450191464337e-06,
      "loss": 0.8102,
      "step": 40
    },
    {
      "epoch": 0.5245202558635395,
      "grad_norm": 0.5821788311004639,
      "learning_rate": 9.839173742253334e-06,
      "loss": 0.7774,
      "step": 41
    },
    {
      "epoch": 0.5373134328358209,
      "grad_norm": 0.5190237760543823,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.766,
      "step": 42
    },
    {
      "epoch": 0.5501066098081023,
      "grad_norm": 0.49539390206336975,
      "learning_rate": 9.799376207714446e-06,
      "loss": 0.7937,
      "step": 43
    },
    {
      "epoch": 0.5628997867803838,
      "grad_norm": 0.5138888359069824,
      "learning_rate": 9.777864028930705e-06,
      "loss": 0.7861,
      "step": 44
    },
    {
      "epoch": 0.5756929637526652,
      "grad_norm": 0.49708378314971924,
      "learning_rate": 9.755282581475769e-06,
      "loss": 0.7356,
      "step": 45
    },
    {
      "epoch": 0.5884861407249466,
      "grad_norm": 0.5589980483055115,
      "learning_rate": 9.731636918995821e-06,
      "loss": 0.7157,
      "step": 46
    },
    {
      "epoch": 0.6012793176972282,
      "grad_norm": 0.5140479207038879,
      "learning_rate": 9.706932333304518e-06,
      "loss": 0.7599,
      "step": 47
    },
    {
      "epoch": 0.6140724946695096,
      "grad_norm": 0.5195008516311646,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.7896,
      "step": 48
    },
    {
      "epoch": 0.6268656716417911,
      "grad_norm": 0.510276734828949,
      "learning_rate": 9.654368743221022e-06,
      "loss": 0.7775,
      "step": 49
    },
    {
      "epoch": 0.6396588486140725,
      "grad_norm": 0.6014343500137329,
      "learning_rate": 9.626521502369984e-06,
      "loss": 0.7615,
      "step": 50
    },
    {
      "epoch": 0.652452025586354,
      "grad_norm": 0.5976372361183167,
      "learning_rate": 9.597638862757255e-06,
      "loss": 0.8036,
      "step": 51
    },
    {
      "epoch": 0.6652452025586354,
      "grad_norm": 0.6337112188339233,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.8269,
      "step": 52
    },
    {
      "epoch": 0.6780383795309168,
      "grad_norm": 0.4970976710319519,
      "learning_rate": 9.536793472839325e-06,
      "loss": 0.7883,
      "step": 53
    },
    {
      "epoch": 0.6908315565031983,
      "grad_norm": 0.5633640289306641,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.7726,
      "step": 54
    },
    {
      "epoch": 0.7036247334754797,
      "grad_norm": 0.6632335186004639,
      "learning_rate": 9.471887038331686e-06,
      "loss": 0.7631,
      "step": 55
    },
    {
      "epoch": 0.7164179104477612,
      "grad_norm": 0.6567376852035522,
      "learning_rate": 9.437928945022772e-06,
      "loss": 0.7916,
      "step": 56
    },
    {
      "epoch": 0.7292110874200426,
      "grad_norm": 0.4805288016796112,
      "learning_rate": 9.40297765928369e-06,
      "loss": 0.7652,
      "step": 57
    },
    {
      "epoch": 0.7420042643923241,
      "grad_norm": 0.5745772123336792,
      "learning_rate": 9.36704100308565e-06,
      "loss": 0.7946,
      "step": 58
    },
    {
      "epoch": 0.7547974413646056,
      "grad_norm": 0.5194308161735535,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.7118,
      "step": 59
    },
    {
      "epoch": 0.767590618336887,
      "grad_norm": 0.4913547933101654,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.7412,
      "step": 60
    },
    {
      "epoch": 0.7803837953091685,
      "grad_norm": 0.5241764187812805,
      "learning_rate": 9.253400328436699e-06,
      "loss": 0.7634,
      "step": 61
    },
    {
      "epoch": 0.7931769722814499,
      "grad_norm": 0.5345350503921509,
      "learning_rate": 9.213604793270196e-06,
      "loss": 0.7353,
      "step": 62
    },
    {
      "epoch": 0.8059701492537313,
      "grad_norm": 0.5227858424186707,
      "learning_rate": 9.172866268606514e-06,
      "loss": 0.7697,
      "step": 63
    },
    {
      "epoch": 0.8187633262260128,
      "grad_norm": 0.4864361584186554,
      "learning_rate": 9.131193871579975e-06,
      "loss": 0.8112,
      "step": 64
    },
    {
      "epoch": 0.8315565031982942,
      "grad_norm": 0.5334176421165466,
      "learning_rate": 9.088596928322158e-06,
      "loss": 0.731,
      "step": 65
    },
    {
      "epoch": 0.8443496801705757,
      "grad_norm": 0.4496009945869446,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.7206,
      "step": 66
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.4755212068557739,
      "learning_rate": 9.000667740056033e-06,
      "loss": 0.7843,
      "step": 67
    },
    {
      "epoch": 0.8699360341151386,
      "grad_norm": 0.45297691226005554,
      "learning_rate": 8.955355173281709e-06,
      "loss": 0.76,
      "step": 68
    },
    {
      "epoch": 0.8827292110874201,
      "grad_norm": 0.4970425069332123,
      "learning_rate": 8.90915741234015e-06,
      "loss": 0.7674,
      "step": 69
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 0.5083153247833252,
      "learning_rate": 8.862084796122998e-06,
      "loss": 0.7656,
      "step": 70
    },
    {
      "epoch": 0.908315565031983,
      "grad_norm": 0.45638445019721985,
      "learning_rate": 8.814147859311333e-06,
      "loss": 0.7303,
      "step": 71
    },
    {
      "epoch": 0.9211087420042644,
      "grad_norm": 0.4436575472354889,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.7295,
      "step": 72
    },
    {
      "epoch": 0.9339019189765458,
      "grad_norm": 0.5214686393737793,
      "learning_rate": 8.715724127386971e-06,
      "loss": 0.7075,
      "step": 73
    },
    {
      "epoch": 0.9466950959488273,
      "grad_norm": 0.48304834961891174,
      "learning_rate": 8.665259359149132e-06,
      "loss": 0.7279,
      "step": 74
    },
    {
      "epoch": 0.9594882729211087,
      "grad_norm": 0.5089297294616699,
      "learning_rate": 8.613974319136959e-06,
      "loss": 0.7424,
      "step": 75
    },
    {
      "epoch": 0.9722814498933902,
      "grad_norm": 0.49402785301208496,
      "learning_rate": 8.561880484756726e-06,
      "loss": 0.7957,
      "step": 76
    },
    {
      "epoch": 0.9850746268656716,
      "grad_norm": 0.5247993469238281,
      "learning_rate": 8.508989514419959e-06,
      "loss": 0.7539,
      "step": 77
    },
    {
      "epoch": 0.997867803837953,
      "grad_norm": 0.4710896611213684,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.7057,
      "step": 78
    },
    {
      "epoch": 1.0106609808102345,
      "grad_norm": 1.7053124904632568,
      "learning_rate": 8.400863688854598e-06,
      "loss": 1.2963,
      "step": 79
    },
    {
      "epoch": 1.023454157782516,
      "grad_norm": 0.5480591654777527,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.7255,
      "step": 80
    },
    {
      "epoch": 1.0362473347547974,
      "grad_norm": 0.5014137029647827,
      "learning_rate": 8.289693629698564e-06,
      "loss": 0.7831,
      "step": 81
    },
    {
      "epoch": 1.049040511727079,
      "grad_norm": 0.4693015515804291,
      "learning_rate": 8.232998006078998e-06,
      "loss": 0.7265,
      "step": 82
    },
    {
      "epoch": 1.0618336886993602,
      "grad_norm": 0.4948160648345947,
      "learning_rate": 8.175578849210894e-06,
      "loss": 0.7189,
      "step": 83
    },
    {
      "epoch": 1.0746268656716418,
      "grad_norm": 0.6815115213394165,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.7964,
      "step": 84
    },
    {
      "epoch": 1.0874200426439233,
      "grad_norm": 0.44852644205093384,
      "learning_rate": 8.058621495575032e-06,
      "loss": 0.6198,
      "step": 85
    },
    {
      "epoch": 1.1002132196162047,
      "grad_norm": 0.535753071308136,
      "learning_rate": 7.99910947343957e-06,
      "loss": 0.6642,
      "step": 86
    },
    {
      "epoch": 1.1130063965884862,
      "grad_norm": 0.46651574969291687,
      "learning_rate": 7.938926261462366e-06,
      "loss": 0.731,
      "step": 87
    },
    {
      "epoch": 1.1257995735607675,
      "grad_norm": 0.5469245910644531,
      "learning_rate": 7.87808532842837e-06,
      "loss": 0.794,
      "step": 88
    },
    {
      "epoch": 1.138592750533049,
      "grad_norm": 0.5276331901550293,
      "learning_rate": 7.81660029031811e-06,
      "loss": 0.6953,
      "step": 89
    },
    {
      "epoch": 1.1513859275053304,
      "grad_norm": 0.42721840739250183,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.7541,
      "step": 90
    },
    {
      "epoch": 1.164179104477612,
      "grad_norm": 0.4775303602218628,
      "learning_rate": 7.691753080453413e-06,
      "loss": 0.704,
      "step": 91
    },
    {
      "epoch": 1.1769722814498933,
      "grad_norm": 0.5940173268318176,
      "learning_rate": 7.628418849052523e-06,
      "loss": 0.778,
      "step": 92
    },
    {
      "epoch": 1.1897654584221748,
      "grad_norm": 0.4526679217815399,
      "learning_rate": 7.564496387029532e-06,
      "loss": 0.7215,
      "step": 93
    },
    {
      "epoch": 1.2025586353944564,
      "grad_norm": 0.4627590775489807,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.7299,
      "step": 94
    },
    {
      "epoch": 1.2153518123667377,
      "grad_norm": 0.4942200183868408,
      "learning_rate": 7.434944122021837e-06,
      "loss": 0.7224,
      "step": 95
    },
    {
      "epoch": 1.2281449893390193,
      "grad_norm": 0.428196519613266,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.7617,
      "step": 96
    },
    {
      "epoch": 1.2409381663113006,
      "grad_norm": 0.4644979238510132,
      "learning_rate": 7.303212252253163e-06,
      "loss": 0.7441,
      "step": 97
    },
    {
      "epoch": 1.2537313432835822,
      "grad_norm": 0.5221542119979858,
      "learning_rate": 7.236565741578163e-06,
      "loss": 0.7311,
      "step": 98
    },
    {
      "epoch": 1.2665245202558635,
      "grad_norm": 0.43763238191604614,
      "learning_rate": 7.169418695587791e-06,
      "loss": 0.7698,
      "step": 99
    },
    {
      "epoch": 1.279317697228145,
      "grad_norm": 0.4271838366985321,
      "learning_rate": 7.101786141547829e-06,
      "loss": 0.6731,
      "step": 100
    },
    {
      "epoch": 1.2921108742004264,
      "grad_norm": 0.4979677200317383,
      "learning_rate": 7.033683215379002e-06,
      "loss": 0.7077,
      "step": 101
    },
    {
      "epoch": 1.304904051172708,
      "grad_norm": 0.4421441853046417,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.7422,
      "step": 102
    },
    {
      "epoch": 1.3176972281449895,
      "grad_norm": 0.44597184658050537,
      "learning_rate": 6.896127313264643e-06,
      "loss": 0.7129,
      "step": 103
    },
    {
      "epoch": 1.3304904051172708,
      "grad_norm": 0.48497170209884644,
      "learning_rate": 6.8267051218319766e-06,
      "loss": 0.7248,
      "step": 104
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 0.5230187177658081,
      "learning_rate": 6.7568741204067145e-06,
      "loss": 0.6923,
      "step": 105
    },
    {
      "epoch": 1.3560767590618337,
      "grad_norm": 0.47476544976234436,
      "learning_rate": 6.686649936914151e-06,
      "loss": 0.732,
      "step": 106
    },
    {
      "epoch": 1.3688699360341152,
      "grad_norm": 0.40445131063461304,
      "learning_rate": 6.616048287272301e-06,
      "loss": 0.7672,
      "step": 107
    },
    {
      "epoch": 1.3816631130063965,
      "grad_norm": 0.4726698100566864,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.7898,
      "step": 108
    },
    {
      "epoch": 1.394456289978678,
      "grad_norm": 0.4922250211238861,
      "learning_rate": 6.473775872054522e-06,
      "loss": 0.7054,
      "step": 109
    },
    {
      "epoch": 1.4072494669509594,
      "grad_norm": 0.4099796712398529,
      "learning_rate": 6.402136946530014e-06,
      "loss": 0.6099,
      "step": 110
    },
    {
      "epoch": 1.420042643923241,
      "grad_norm": 0.5015040040016174,
      "learning_rate": 6.330184227833376e-06,
      "loss": 0.7295,
      "step": 111
    },
    {
      "epoch": 1.4328358208955223,
      "grad_norm": 0.43680623173713684,
      "learning_rate": 6.257933818722544e-06,
      "loss": 0.7267,
      "step": 112
    },
    {
      "epoch": 1.4456289978678039,
      "grad_norm": 0.4174647927284241,
      "learning_rate": 6.185401888577488e-06,
      "loss": 0.6789,
      "step": 113
    },
    {
      "epoch": 1.4584221748400852,
      "grad_norm": 0.45874616503715515,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.8178,
      "step": 114
    },
    {
      "epoch": 1.4712153518123667,
      "grad_norm": 0.39787718653678894,
      "learning_rate": 6.039558454088796e-06,
      "loss": 0.6645,
      "step": 115
    },
    {
      "epoch": 1.4840085287846483,
      "grad_norm": 0.6014237403869629,
      "learning_rate": 5.9662795889777666e-06,
      "loss": 0.7598,
      "step": 116
    },
    {
      "epoch": 1.4968017057569296,
      "grad_norm": 0.4093945026397705,
      "learning_rate": 5.892784473993184e-06,
      "loss": 0.6873,
      "step": 117
    },
    {
      "epoch": 1.509594882729211,
      "grad_norm": 0.48048579692840576,
      "learning_rate": 5.819089557075689e-06,
      "loss": 0.7689,
      "step": 118
    },
    {
      "epoch": 1.5223880597014925,
      "grad_norm": 0.45258212089538574,
      "learning_rate": 5.745211330880872e-06,
      "loss": 0.6531,
      "step": 119
    },
    {
      "epoch": 1.535181236673774,
      "grad_norm": 0.4489460289478302,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.7181,
      "step": 120
    },
    {
      "epoch": 1.5479744136460556,
      "grad_norm": 0.39668354392051697,
      "learning_rate": 5.596971122701221e-06,
      "loss": 0.6797,
      "step": 121
    },
    {
      "epoch": 1.560767590618337,
      "grad_norm": 0.3912845849990845,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.736,
      "step": 122
    },
    {
      "epoch": 1.5735607675906182,
      "grad_norm": 0.45017972588539124,
      "learning_rate": 5.448196544517168e-06,
      "loss": 0.7289,
      "step": 123
    },
    {
      "epoch": 1.5863539445628998,
      "grad_norm": 0.4025014042854309,
      "learning_rate": 5.373650467932122e-06,
      "loss": 0.7216,
      "step": 124
    },
    {
      "epoch": 1.5991471215351813,
      "grad_norm": 0.4008171558380127,
      "learning_rate": 5.299020769725172e-06,
      "loss": 0.6882,
      "step": 125
    },
    {
      "epoch": 1.6119402985074627,
      "grad_norm": 0.3734460175037384,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.7057,
      "step": 126
    },
    {
      "epoch": 1.624733475479744,
      "grad_norm": 0.41330069303512573,
      "learning_rate": 5.1495773308469935e-06,
      "loss": 0.6429,
      "step": 127
    },
    {
      "epoch": 1.6375266524520256,
      "grad_norm": 0.409628689289093,
      "learning_rate": 5.074797035076319e-06,
      "loss": 0.7052,
      "step": 128
    },
    {
      "epoch": 1.650319829424307,
      "grad_norm": 0.3930547535419464,
      "learning_rate": 5e-06,
      "loss": 0.773,
      "step": 129
    },
    {
      "epoch": 1.6631130063965884,
      "grad_norm": 0.47582224011421204,
      "learning_rate": 4.9252029649236835e-06,
      "loss": 0.7179,
      "step": 130
    },
    {
      "epoch": 1.67590618336887,
      "grad_norm": 0.4209410548210144,
      "learning_rate": 4.850422669153009e-06,
      "loss": 0.6379,
      "step": 131
    },
    {
      "epoch": 1.6886993603411513,
      "grad_norm": 0.4784112870693207,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.7232,
      "step": 132
    },
    {
      "epoch": 1.7014925373134329,
      "grad_norm": 0.44215840101242065,
      "learning_rate": 4.700979230274829e-06,
      "loss": 0.663,
      "step": 133
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.5071489810943604,
      "learning_rate": 4.626349532067879e-06,
      "loss": 0.7215,
      "step": 134
    },
    {
      "epoch": 1.7270788912579957,
      "grad_norm": 0.47784915566444397,
      "learning_rate": 4.551803455482833e-06,
      "loss": 0.788,
      "step": 135
    },
    {
      "epoch": 1.739872068230277,
      "grad_norm": 0.5307613611221313,
      "learning_rate": 4.477357683661734e-06,
      "loss": 0.745,
      "step": 136
    },
    {
      "epoch": 1.7526652452025586,
      "grad_norm": 0.455413281917572,
      "learning_rate": 4.4030288772987795e-06,
      "loss": 0.7129,
      "step": 137
    },
    {
      "epoch": 1.7654584221748402,
      "grad_norm": 0.4110308289527893,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.7328,
      "step": 138
    },
    {
      "epoch": 1.7782515991471215,
      "grad_norm": 0.5132991075515747,
      "learning_rate": 4.254788669119127e-06,
      "loss": 0.7738,
      "step": 139
    },
    {
      "epoch": 1.7910447761194028,
      "grad_norm": 0.4838980436325073,
      "learning_rate": 4.180910442924312e-06,
      "loss": 0.6317,
      "step": 140
    },
    {
      "epoch": 1.8038379530916844,
      "grad_norm": 0.4499025046825409,
      "learning_rate": 4.107215526006818e-06,
      "loss": 0.7288,
      "step": 141
    },
    {
      "epoch": 1.816631130063966,
      "grad_norm": 0.47396278381347656,
      "learning_rate": 4.033720411022235e-06,
      "loss": 0.776,
      "step": 142
    },
    {
      "epoch": 1.8294243070362475,
      "grad_norm": 0.3979196548461914,
      "learning_rate": 3.960441545911205e-06,
      "loss": 0.683,
      "step": 143
    },
    {
      "epoch": 1.8422174840085288,
      "grad_norm": 0.43219274282455444,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.7194,
      "step": 144
    },
    {
      "epoch": 1.8550106609808101,
      "grad_norm": 0.4215470254421234,
      "learning_rate": 3.8145981114225135e-06,
      "loss": 0.7744,
      "step": 145
    },
    {
      "epoch": 1.8678038379530917,
      "grad_norm": 0.4141583740711212,
      "learning_rate": 3.7420661812774577e-06,
      "loss": 0.7245,
      "step": 146
    },
    {
      "epoch": 1.8805970149253732,
      "grad_norm": 0.3775157332420349,
      "learning_rate": 3.669815772166625e-06,
      "loss": 0.6878,
      "step": 147
    },
    {
      "epoch": 1.8933901918976546,
      "grad_norm": 0.4012461304664612,
      "learning_rate": 3.5978630534699873e-06,
      "loss": 0.7315,
      "step": 148
    },
    {
      "epoch": 1.906183368869936,
      "grad_norm": 0.3893410861492157,
      "learning_rate": 3.526224127945479e-06,
      "loss": 0.6943,
      "step": 149
    },
    {
      "epoch": 1.9189765458422174,
      "grad_norm": 0.4317570626735687,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.6904,
      "step": 150
    },
    {
      "epoch": 1.931769722814499,
      "grad_norm": 0.4641903340816498,
      "learning_rate": 3.383951712727701e-06,
      "loss": 0.7495,
      "step": 151
    },
    {
      "epoch": 1.9445628997867805,
      "grad_norm": 0.4203890264034271,
      "learning_rate": 3.3133500630858507e-06,
      "loss": 0.7436,
      "step": 152
    },
    {
      "epoch": 1.9573560767590619,
      "grad_norm": 0.39781543612480164,
      "learning_rate": 3.2431258795932863e-06,
      "loss": 0.7042,
      "step": 153
    },
    {
      "epoch": 1.9701492537313432,
      "grad_norm": 0.4704498052597046,
      "learning_rate": 3.173294878168025e-06,
      "loss": 0.7379,
      "step": 154
    },
    {
      "epoch": 1.9829424307036247,
      "grad_norm": 0.39690351486206055,
      "learning_rate": 3.1038726867353587e-06,
      "loss": 0.6986,
      "step": 155
    },
    {
      "epoch": 1.9957356076759063,
      "grad_norm": 0.46861645579338074,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.7317,
      "step": 156
    },
    {
      "epoch": 2.008528784648188,
      "grad_norm": 1.1607420444488525,
      "learning_rate": 2.966316784621e-06,
      "loss": 1.2703,
      "step": 157
    },
    {
      "epoch": 2.021321961620469,
      "grad_norm": 0.38464683294296265,
      "learning_rate": 2.8982138584521734e-06,
      "loss": 0.7064,
      "step": 158
    },
    {
      "epoch": 2.0341151385927505,
      "grad_norm": 0.5022476315498352,
      "learning_rate": 2.83058130441221e-06,
      "loss": 0.7061,
      "step": 159
    },
    {
      "epoch": 2.046908315565032,
      "grad_norm": 0.5013148188591003,
      "learning_rate": 2.7634342584218364e-06,
      "loss": 0.7014,
      "step": 160
    },
    {
      "epoch": 2.0597014925373136,
      "grad_norm": 0.47904056310653687,
      "learning_rate": 2.6967877477468394e-06,
      "loss": 0.6972,
      "step": 161
    },
    {
      "epoch": 2.0724946695095947,
      "grad_norm": 0.41209492087364197,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.6418,
      "step": 162
    },
    {
      "epoch": 2.0852878464818763,
      "grad_norm": 0.46157318353652954,
      "learning_rate": 2.5650558779781635e-06,
      "loss": 0.7516,
      "step": 163
    },
    {
      "epoch": 2.098081023454158,
      "grad_norm": 0.42628005146980286,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.7129,
      "step": 164
    },
    {
      "epoch": 2.1108742004264394,
      "grad_norm": 0.45490726828575134,
      "learning_rate": 2.43550361297047e-06,
      "loss": 0.7031,
      "step": 165
    },
    {
      "epoch": 2.1236673773987205,
      "grad_norm": 0.4108670651912689,
      "learning_rate": 2.371581150947476e-06,
      "loss": 0.7059,
      "step": 166
    },
    {
      "epoch": 2.136460554371002,
      "grad_norm": 0.4316941797733307,
      "learning_rate": 2.3082469195465893e-06,
      "loss": 0.6921,
      "step": 167
    },
    {
      "epoch": 2.1492537313432836,
      "grad_norm": 0.4376332759857178,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.6459,
      "step": 168
    },
    {
      "epoch": 2.162046908315565,
      "grad_norm": 0.4335668981075287,
      "learning_rate": 2.1833997096818897e-06,
      "loss": 0.6815,
      "step": 169
    },
    {
      "epoch": 2.1748400852878467,
      "grad_norm": 0.37335702776908875,
      "learning_rate": 2.1219146715716332e-06,
      "loss": 0.6917,
      "step": 170
    },
    {
      "epoch": 2.1876332622601278,
      "grad_norm": 0.4028381109237671,
      "learning_rate": 2.061073738537635e-06,
      "loss": 0.6415,
      "step": 171
    },
    {
      "epoch": 2.2004264392324093,
      "grad_norm": 0.4265342056751251,
      "learning_rate": 2.0008905265604316e-06,
      "loss": 0.6476,
      "step": 172
    },
    {
      "epoch": 2.213219616204691,
      "grad_norm": 0.4345915615558624,
      "learning_rate": 1.941378504424968e-06,
      "loss": 0.8098,
      "step": 173
    },
    {
      "epoch": 2.2260127931769724,
      "grad_norm": 0.35505157709121704,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.6996,
      "step": 174
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 0.3520065248012543,
      "learning_rate": 1.8244211507891064e-06,
      "loss": 0.6946,
      "step": 175
    },
    {
      "epoch": 2.251599147121535,
      "grad_norm": 0.3407677114009857,
      "learning_rate": 1.7670019939210025e-06,
      "loss": 0.7051,
      "step": 176
    },
    {
      "epoch": 2.2643923240938166,
      "grad_norm": 0.39579153060913086,
      "learning_rate": 1.7103063703014372e-06,
      "loss": 0.6582,
      "step": 177
    },
    {
      "epoch": 2.277185501066098,
      "grad_norm": 0.3585205078125,
      "learning_rate": 1.6543469682057105e-06,
      "loss": 0.6591,
      "step": 178
    },
    {
      "epoch": 2.2899786780383797,
      "grad_norm": 0.3757244646549225,
      "learning_rate": 1.5991363111454023e-06,
      "loss": 0.6544,
      "step": 179
    },
    {
      "epoch": 2.302771855010661,
      "grad_norm": 0.34080037474632263,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.6695,
      "step": 180
    },
    {
      "epoch": 2.3155650319829424,
      "grad_norm": 0.36547356843948364,
      "learning_rate": 1.4910104855800429e-06,
      "loss": 0.6983,
      "step": 181
    },
    {
      "epoch": 2.328358208955224,
      "grad_norm": 0.3294934630393982,
      "learning_rate": 1.438119515243277e-06,
      "loss": 0.6253,
      "step": 182
    },
    {
      "epoch": 2.3411513859275055,
      "grad_norm": 0.3605087399482727,
      "learning_rate": 1.3860256808630429e-06,
      "loss": 0.6788,
      "step": 183
    },
    {
      "epoch": 2.3539445628997866,
      "grad_norm": 0.39677152037620544,
      "learning_rate": 1.3347406408508695e-06,
      "loss": 0.7431,
      "step": 184
    },
    {
      "epoch": 2.366737739872068,
      "grad_norm": 0.3269940912723541,
      "learning_rate": 1.2842758726130283e-06,
      "loss": 0.6587,
      "step": 185
    },
    {
      "epoch": 2.3795309168443497,
      "grad_norm": 0.3385223150253296,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.7,
      "step": 186
    },
    {
      "epoch": 2.3923240938166312,
      "grad_norm": 0.35292384028434753,
      "learning_rate": 1.1858521406886674e-06,
      "loss": 0.7007,
      "step": 187
    },
    {
      "epoch": 2.405117270788913,
      "grad_norm": 0.34651538729667664,
      "learning_rate": 1.137915203877003e-06,
      "loss": 0.7469,
      "step": 188
    },
    {
      "epoch": 2.417910447761194,
      "grad_norm": 0.32731908559799194,
      "learning_rate": 1.0908425876598512e-06,
      "loss": 0.6608,
      "step": 189
    },
    {
      "epoch": 2.4307036247334755,
      "grad_norm": 0.3515162765979767,
      "learning_rate": 1.044644826718295e-06,
      "loss": 0.7566,
      "step": 190
    },
    {
      "epoch": 2.443496801705757,
      "grad_norm": 0.31811511516571045,
      "learning_rate": 9.993322599439692e-07,
      "loss": 0.6596,
      "step": 191
    },
    {
      "epoch": 2.4562899786780386,
      "grad_norm": 0.3392126262187958,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.7325,
      "step": 192
    },
    {
      "epoch": 2.4690831556503197,
      "grad_norm": 0.33756223320961,
      "learning_rate": 9.114030716778433e-07,
      "loss": 0.6592,
      "step": 193
    },
    {
      "epoch": 2.481876332622601,
      "grad_norm": 0.34133580327033997,
      "learning_rate": 8.688061284200266e-07,
      "loss": 0.7221,
      "step": 194
    },
    {
      "epoch": 2.4946695095948828,
      "grad_norm": 0.3212495446205139,
      "learning_rate": 8.271337313934869e-07,
      "loss": 0.6698,
      "step": 195
    },
    {
      "epoch": 2.5074626865671643,
      "grad_norm": 0.3041667640209198,
      "learning_rate": 7.863952067298042e-07,
      "loss": 0.691,
      "step": 196
    },
    {
      "epoch": 2.520255863539446,
      "grad_norm": 0.35703015327453613,
      "learning_rate": 7.465996715633028e-07,
      "loss": 0.7766,
      "step": 197
    },
    {
      "epoch": 2.533049040511727,
      "grad_norm": 0.3571200966835022,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.7563,
      "step": 198
    },
    {
      "epoch": 2.5458422174840085,
      "grad_norm": 0.33537036180496216,
      "learning_rate": 6.698729810778065e-07,
      "loss": 0.7013,
      "step": 199
    },
    {
      "epoch": 2.55863539445629,
      "grad_norm": 0.3055112063884735,
      "learning_rate": 6.329589969143518e-07,
      "loss": 0.6536,
      "step": 200
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.3426555097103119,
      "learning_rate": 5.9702234071631e-07,
      "loss": 0.7274,
      "step": 201
    },
    {
      "epoch": 2.5842217484008527,
      "grad_norm": 0.3432694673538208,
      "learning_rate": 5.620710549772295e-07,
      "loss": 0.6523,
      "step": 202
    },
    {
      "epoch": 2.5970149253731343,
      "grad_norm": 0.36352601647377014,
      "learning_rate": 5.281129616683167e-07,
      "loss": 0.7254,
      "step": 203
    },
    {
      "epoch": 2.609808102345416,
      "grad_norm": 0.3131335973739624,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.6832,
      "step": 204
    },
    {
      "epoch": 2.6226012793176974,
      "grad_norm": 0.3547823429107666,
      "learning_rate": 4.632065271606756e-07,
      "loss": 0.7226,
      "step": 205
    },
    {
      "epoch": 2.635394456289979,
      "grad_norm": 0.2958380877971649,
      "learning_rate": 4.322727117869951e-07,
      "loss": 0.6507,
      "step": 206
    },
    {
      "epoch": 2.64818763326226,
      "grad_norm": 0.2793337404727936,
      "learning_rate": 4.0236113724274716e-07,
      "loss": 0.6371,
      "step": 207
    },
    {
      "epoch": 2.6609808102345416,
      "grad_norm": 0.33492887020111084,
      "learning_rate": 3.734784976300165e-07,
      "loss": 0.7792,
      "step": 208
    },
    {
      "epoch": 2.673773987206823,
      "grad_norm": 0.3088694214820862,
      "learning_rate": 3.4563125677897936e-07,
      "loss": 0.5887,
      "step": 209
    },
    {
      "epoch": 2.6865671641791042,
      "grad_norm": 0.3331063389778137,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.6793,
      "step": 210
    },
    {
      "epoch": 2.699360341151386,
      "grad_norm": 0.3037247359752655,
      "learning_rate": 2.930676666954846e-07,
      "loss": 0.6553,
      "step": 211
    },
    {
      "epoch": 2.7121535181236673,
      "grad_norm": 0.3542158305644989,
      "learning_rate": 2.6836308100417874e-07,
      "loss": 0.7101,
      "step": 212
    },
    {
      "epoch": 2.724946695095949,
      "grad_norm": 0.3504466116428375,
      "learning_rate": 2.447174185242324e-07,
      "loss": 0.7189,
      "step": 213
    },
    {
      "epoch": 2.7377398720682304,
      "grad_norm": 0.30136266350746155,
      "learning_rate": 2.2213597106929608e-07,
      "loss": 0.6076,
      "step": 214
    },
    {
      "epoch": 2.750533049040512,
      "grad_norm": 0.3250851631164551,
      "learning_rate": 2.006237922855553e-07,
      "loss": 0.7334,
      "step": 215
    },
    {
      "epoch": 2.763326226012793,
      "grad_norm": 0.3266373574733734,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.7144,
      "step": 216
    },
    {
      "epoch": 2.7761194029850746,
      "grad_norm": 0.2980578541755676,
      "learning_rate": 1.6082625774666793e-07,
      "loss": 0.6139,
      "step": 217
    },
    {
      "epoch": 2.788912579957356,
      "grad_norm": 0.31971806287765503,
      "learning_rate": 1.4254980853566248e-07,
      "loss": 0.7111,
      "step": 218
    },
    {
      "epoch": 2.8017057569296373,
      "grad_norm": 0.30652180314064026,
      "learning_rate": 1.253604390908819e-07,
      "loss": 0.6719,
      "step": 219
    },
    {
      "epoch": 2.814498933901919,
      "grad_norm": 0.32493188977241516,
      "learning_rate": 1.0926199633097156e-07,
      "loss": 0.7834,
      "step": 220
    },
    {
      "epoch": 2.8272921108742004,
      "grad_norm": 0.30289193987846375,
      "learning_rate": 9.42580830291373e-08,
      "loss": 0.6486,
      "step": 221
    },
    {
      "epoch": 2.840085287846482,
      "grad_norm": 0.39349403977394104,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.6523,
      "step": 222
    },
    {
      "epoch": 2.8528784648187635,
      "grad_norm": 0.32352349162101746,
      "learning_rate": 6.75470303823933e-08,
      "loss": 0.708,
      "step": 223
    },
    {
      "epoch": 2.8656716417910446,
      "grad_norm": 0.31589657068252563,
      "learning_rate": 5.584586887435739e-08,
      "loss": 0.7025,
      "step": 224
    },
    {
      "epoch": 2.878464818763326,
      "grad_norm": 0.3233773708343506,
      "learning_rate": 4.52511911603265e-08,
      "loss": 0.7003,
      "step": 225
    },
    {
      "epoch": 2.8912579957356077,
      "grad_norm": 0.3037286400794983,
      "learning_rate": 3.576536829081323e-08,
      "loss": 0.7252,
      "step": 226
    },
    {
      "epoch": 2.9040511727078893,
      "grad_norm": 0.3222990334033966,
      "learning_rate": 2.7390523158633552e-08,
      "loss": 0.7079,
      "step": 227
    },
    {
      "epoch": 2.9168443496801704,
      "grad_norm": 0.30910524725914,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.6701,
      "step": 228
    },
    {
      "epoch": 2.929637526652452,
      "grad_norm": 0.29240769147872925,
      "learning_rate": 1.3981014094099354e-08,
      "loss": 0.6388,
      "step": 229
    },
    {
      "epoch": 2.9424307036247335,
      "grad_norm": 0.34161651134490967,
      "learning_rate": 8.949351161324227e-09,
      "loss": 0.7613,
      "step": 230
    },
    {
      "epoch": 2.955223880597015,
      "grad_norm": 0.299943745136261,
      "learning_rate": 5.034667293427053e-09,
      "loss": 0.6471,
      "step": 231
    },
    {
      "epoch": 2.9680170575692966,
      "grad_norm": 0.3515676259994507,
      "learning_rate": 2.237838582483387e-09,
      "loss": 0.7412,
      "step": 232
    },
    {
      "epoch": 2.9808102345415777,
      "grad_norm": 0.3143077790737152,
      "learning_rate": 5.594909486328348e-10,
      "loss": 0.6572,
      "step": 233
    },
    {
      "epoch": 2.9936034115138592,
      "grad_norm": 0.3188575506210327,
      "learning_rate": 0.0,
      "loss": 0.7006,
      "step": 234
    },
    {
      "epoch": 2.9936034115138592,
      "step": 234,
      "total_flos": 327743678562304.0,
      "train_loss": 0.7533083028263516,
      "train_runtime": 9147.0121,
      "train_samples_per_second": 2.46,
      "train_steps_per_second": 0.026
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 234,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 327743678562304.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}