Rae-Pretrain / trainer_state.json
Delta-Vector's picture
Upload folder using huggingface_hub
a2716df verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1376,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0007267441860465116,
"grad_norm": 9.074579620427265,
"learning_rate": 0.0,
"loss": 2.5726,
"step": 1
},
{
"epoch": 0.0014534883720930232,
"grad_norm": 9.325884168450177,
"learning_rate": 2.0000000000000002e-07,
"loss": 2.5915,
"step": 2
},
{
"epoch": 0.002180232558139535,
"grad_norm": 9.577883515575813,
"learning_rate": 4.0000000000000003e-07,
"loss": 2.5166,
"step": 3
},
{
"epoch": 0.0029069767441860465,
"grad_norm": 9.081960014952461,
"learning_rate": 6.000000000000001e-07,
"loss": 2.5275,
"step": 4
},
{
"epoch": 0.003633720930232558,
"grad_norm": 9.146732295895594,
"learning_rate": 8.000000000000001e-07,
"loss": 2.6068,
"step": 5
},
{
"epoch": 0.00436046511627907,
"grad_norm": 8.878352797679302,
"learning_rate": 1.0000000000000002e-06,
"loss": 2.5602,
"step": 6
},
{
"epoch": 0.005087209302325582,
"grad_norm": 8.7892387976441,
"learning_rate": 1.2000000000000002e-06,
"loss": 2.5768,
"step": 7
},
{
"epoch": 0.005813953488372093,
"grad_norm": 8.439946529584597,
"learning_rate": 1.4000000000000001e-06,
"loss": 2.571,
"step": 8
},
{
"epoch": 0.006540697674418605,
"grad_norm": 7.525421260655473,
"learning_rate": 1.6000000000000001e-06,
"loss": 2.5806,
"step": 9
},
{
"epoch": 0.007267441860465116,
"grad_norm": 7.420995687542313,
"learning_rate": 1.8000000000000001e-06,
"loss": 2.5632,
"step": 10
},
{
"epoch": 0.007994186046511628,
"grad_norm": 6.529233432613504,
"learning_rate": 2.0000000000000003e-06,
"loss": 2.5702,
"step": 11
},
{
"epoch": 0.00872093023255814,
"grad_norm": 5.076235841533463,
"learning_rate": 2.2e-06,
"loss": 2.4557,
"step": 12
},
{
"epoch": 0.00944767441860465,
"grad_norm": 4.530065999910271,
"learning_rate": 2.4000000000000003e-06,
"loss": 2.4625,
"step": 13
},
{
"epoch": 0.010174418604651164,
"grad_norm": 3.757971086858522,
"learning_rate": 2.6e-06,
"loss": 2.5126,
"step": 14
},
{
"epoch": 0.010901162790697675,
"grad_norm": 3.0710603482243517,
"learning_rate": 2.8000000000000003e-06,
"loss": 2.4419,
"step": 15
},
{
"epoch": 0.011627906976744186,
"grad_norm": 2.8230945104887333,
"learning_rate": 3e-06,
"loss": 2.437,
"step": 16
},
{
"epoch": 0.012354651162790697,
"grad_norm": 2.6279435114027168,
"learning_rate": 3.2000000000000003e-06,
"loss": 2.3592,
"step": 17
},
{
"epoch": 0.01308139534883721,
"grad_norm": 2.5161015684914774,
"learning_rate": 3.4000000000000005e-06,
"loss": 2.3957,
"step": 18
},
{
"epoch": 0.013808139534883721,
"grad_norm": 2.4278917076308475,
"learning_rate": 3.6000000000000003e-06,
"loss": 2.3352,
"step": 19
},
{
"epoch": 0.014534883720930232,
"grad_norm": 2.348998890488855,
"learning_rate": 3.8000000000000005e-06,
"loss": 2.4372,
"step": 20
},
{
"epoch": 0.015261627906976744,
"grad_norm": 2.4780736413377396,
"learning_rate": 4.000000000000001e-06,
"loss": 2.3838,
"step": 21
},
{
"epoch": 0.015988372093023256,
"grad_norm": 2.5057556166741604,
"learning_rate": 4.2000000000000004e-06,
"loss": 2.4095,
"step": 22
},
{
"epoch": 0.016715116279069766,
"grad_norm": 2.3191989107995648,
"learning_rate": 4.4e-06,
"loss": 2.3799,
"step": 23
},
{
"epoch": 0.01744186046511628,
"grad_norm": 2.187761956889251,
"learning_rate": 4.600000000000001e-06,
"loss": 2.4215,
"step": 24
},
{
"epoch": 0.018168604651162792,
"grad_norm": 1.670881599245353,
"learning_rate": 4.800000000000001e-06,
"loss": 2.395,
"step": 25
},
{
"epoch": 0.0188953488372093,
"grad_norm": 1.5365760213246764,
"learning_rate": 5e-06,
"loss": 2.4282,
"step": 26
},
{
"epoch": 0.019622093023255814,
"grad_norm": 1.657846538045198,
"learning_rate": 5.2e-06,
"loss": 2.4198,
"step": 27
},
{
"epoch": 0.020348837209302327,
"grad_norm": 1.5638927670066207,
"learning_rate": 5.400000000000001e-06,
"loss": 2.3976,
"step": 28
},
{
"epoch": 0.021075581395348836,
"grad_norm": 1.5407040415402946,
"learning_rate": 5.600000000000001e-06,
"loss": 2.4409,
"step": 29
},
{
"epoch": 0.02180232558139535,
"grad_norm": 1.4066346512940076,
"learning_rate": 5.8e-06,
"loss": 2.3676,
"step": 30
},
{
"epoch": 0.02252906976744186,
"grad_norm": 1.3673414382471556,
"learning_rate": 6e-06,
"loss": 2.3971,
"step": 31
},
{
"epoch": 0.023255813953488372,
"grad_norm": 1.495128974221532,
"learning_rate": 6.200000000000001e-06,
"loss": 2.4844,
"step": 32
},
{
"epoch": 0.023982558139534885,
"grad_norm": 2.2845356742596397,
"learning_rate": 6.4000000000000006e-06,
"loss": 2.3814,
"step": 33
},
{
"epoch": 0.024709302325581394,
"grad_norm": 1.443466209692566,
"learning_rate": 6.600000000000001e-06,
"loss": 2.4134,
"step": 34
},
{
"epoch": 0.025436046511627907,
"grad_norm": 1.7492603796870134,
"learning_rate": 6.800000000000001e-06,
"loss": 2.3634,
"step": 35
},
{
"epoch": 0.02616279069767442,
"grad_norm": 1.6767321941082,
"learning_rate": 7e-06,
"loss": 2.4547,
"step": 36
},
{
"epoch": 0.02688953488372093,
"grad_norm": 1.4015556883781917,
"learning_rate": 7.2000000000000005e-06,
"loss": 2.3875,
"step": 37
},
{
"epoch": 0.027616279069767442,
"grad_norm": 1.3729850201136446,
"learning_rate": 7.4e-06,
"loss": 2.3611,
"step": 38
},
{
"epoch": 0.028343023255813952,
"grad_norm": 1.4151893697508262,
"learning_rate": 7.600000000000001e-06,
"loss": 2.3659,
"step": 39
},
{
"epoch": 0.029069767441860465,
"grad_norm": 1.475208153189891,
"learning_rate": 7.800000000000002e-06,
"loss": 2.371,
"step": 40
},
{
"epoch": 0.029796511627906978,
"grad_norm": 1.5523241828124394,
"learning_rate": 8.000000000000001e-06,
"loss": 2.3868,
"step": 41
},
{
"epoch": 0.030523255813953487,
"grad_norm": 1.3127492176118258,
"learning_rate": 8.2e-06,
"loss": 2.4054,
"step": 42
},
{
"epoch": 0.03125,
"grad_norm": 1.2689101037621893,
"learning_rate": 8.400000000000001e-06,
"loss": 2.3864,
"step": 43
},
{
"epoch": 0.03197674418604651,
"grad_norm": 1.335611750054927,
"learning_rate": 8.6e-06,
"loss": 2.442,
"step": 44
},
{
"epoch": 0.032703488372093026,
"grad_norm": 1.2452087861474146,
"learning_rate": 8.8e-06,
"loss": 2.3697,
"step": 45
},
{
"epoch": 0.03343023255813953,
"grad_norm": 1.4439605178224628,
"learning_rate": 9e-06,
"loss": 2.3734,
"step": 46
},
{
"epoch": 0.034156976744186045,
"grad_norm": 1.2927603151316898,
"learning_rate": 9.200000000000002e-06,
"loss": 2.4011,
"step": 47
},
{
"epoch": 0.03488372093023256,
"grad_norm": 1.185649482589594,
"learning_rate": 9.4e-06,
"loss": 2.3013,
"step": 48
},
{
"epoch": 0.03561046511627907,
"grad_norm": 1.6520257001233214,
"learning_rate": 9.600000000000001e-06,
"loss": 2.3405,
"step": 49
},
{
"epoch": 0.036337209302325583,
"grad_norm": 1.415240980367823,
"learning_rate": 9.800000000000001e-06,
"loss": 2.3749,
"step": 50
},
{
"epoch": 0.037063953488372096,
"grad_norm": 1.1599368922180915,
"learning_rate": 1e-05,
"loss": 2.3551,
"step": 51
},
{
"epoch": 0.0377906976744186,
"grad_norm": 1.3457151082199426,
"learning_rate": 9.9999859669361e-06,
"loss": 2.3236,
"step": 52
},
{
"epoch": 0.038517441860465115,
"grad_norm": 1.3225280521166263,
"learning_rate": 9.999943867823174e-06,
"loss": 2.3406,
"step": 53
},
{
"epoch": 0.03924418604651163,
"grad_norm": 1.3072339789726655,
"learning_rate": 9.999873702897528e-06,
"loss": 2.3821,
"step": 54
},
{
"epoch": 0.03997093023255814,
"grad_norm": 1.4674401083737256,
"learning_rate": 9.999775472553019e-06,
"loss": 2.3389,
"step": 55
},
{
"epoch": 0.040697674418604654,
"grad_norm": 7.364273715054014,
"learning_rate": 9.999649177341036e-06,
"loss": 2.3676,
"step": 56
},
{
"epoch": 0.04142441860465116,
"grad_norm": 1.4164513349686614,
"learning_rate": 9.999494817970498e-06,
"loss": 2.3726,
"step": 57
},
{
"epoch": 0.04215116279069767,
"grad_norm": 1.216986905939288,
"learning_rate": 9.999312395307861e-06,
"loss": 2.3642,
"step": 58
},
{
"epoch": 0.042877906976744186,
"grad_norm": 1.0885159567206724,
"learning_rate": 9.999101910377107e-06,
"loss": 2.3726,
"step": 59
},
{
"epoch": 0.0436046511627907,
"grad_norm": 1.2741330951835304,
"learning_rate": 9.998863364359734e-06,
"loss": 2.3776,
"step": 60
},
{
"epoch": 0.04433139534883721,
"grad_norm": 1.1966086656900674,
"learning_rate": 9.998596758594752e-06,
"loss": 2.4317,
"step": 61
},
{
"epoch": 0.04505813953488372,
"grad_norm": 1.336165326700706,
"learning_rate": 9.998302094578685e-06,
"loss": 2.3995,
"step": 62
},
{
"epoch": 0.04578488372093023,
"grad_norm": 1.1879637364009348,
"learning_rate": 9.997979373965542e-06,
"loss": 2.3422,
"step": 63
},
{
"epoch": 0.046511627906976744,
"grad_norm": 1.1912341449378812,
"learning_rate": 9.99762859856683e-06,
"loss": 2.4091,
"step": 64
},
{
"epoch": 0.047238372093023256,
"grad_norm": 1.2930140030336172,
"learning_rate": 9.997249770351531e-06,
"loss": 2.3575,
"step": 65
},
{
"epoch": 0.04796511627906977,
"grad_norm": 1.194601702985452,
"learning_rate": 9.996842891446092e-06,
"loss": 2.3894,
"step": 66
},
{
"epoch": 0.04869186046511628,
"grad_norm": 1.2151890579244191,
"learning_rate": 9.996407964134416e-06,
"loss": 2.3649,
"step": 67
},
{
"epoch": 0.04941860465116279,
"grad_norm": 1.2265501897788895,
"learning_rate": 9.995944990857848e-06,
"loss": 2.3229,
"step": 68
},
{
"epoch": 0.0501453488372093,
"grad_norm": 1.330613412551332,
"learning_rate": 9.995453974215164e-06,
"loss": 2.3961,
"step": 69
},
{
"epoch": 0.050872093023255814,
"grad_norm": 1.278238260413937,
"learning_rate": 9.994934916962547e-06,
"loss": 2.4385,
"step": 70
},
{
"epoch": 0.05159883720930233,
"grad_norm": 1.1586366855530064,
"learning_rate": 9.994387822013586e-06,
"loss": 2.3732,
"step": 71
},
{
"epoch": 0.05232558139534884,
"grad_norm": 1.217932106034384,
"learning_rate": 9.993812692439247e-06,
"loss": 2.3922,
"step": 72
},
{
"epoch": 0.053052325581395346,
"grad_norm": 1.1455244475688116,
"learning_rate": 9.99320953146786e-06,
"loss": 2.3852,
"step": 73
},
{
"epoch": 0.05377906976744186,
"grad_norm": 1.1743872650502356,
"learning_rate": 9.992578342485107e-06,
"loss": 2.3637,
"step": 74
},
{
"epoch": 0.05450581395348837,
"grad_norm": 1.2681205469149612,
"learning_rate": 9.991919129033994e-06,
"loss": 2.4364,
"step": 75
},
{
"epoch": 0.055232558139534885,
"grad_norm": 1.2522437864600409,
"learning_rate": 9.99123189481483e-06,
"loss": 2.3807,
"step": 76
},
{
"epoch": 0.0559593023255814,
"grad_norm": 1.1756390549643723,
"learning_rate": 9.990516643685222e-06,
"loss": 2.323,
"step": 77
},
{
"epoch": 0.056686046511627904,
"grad_norm": 1.388760289434306,
"learning_rate": 9.98977337966003e-06,
"loss": 2.3559,
"step": 78
},
{
"epoch": 0.057412790697674417,
"grad_norm": 1.2560913594757854,
"learning_rate": 9.989002106911368e-06,
"loss": 2.3606,
"step": 79
},
{
"epoch": 0.05813953488372093,
"grad_norm": 1.354056708770636,
"learning_rate": 9.988202829768562e-06,
"loss": 2.3182,
"step": 80
},
{
"epoch": 0.05886627906976744,
"grad_norm": 1.5428329715782925,
"learning_rate": 9.987375552718133e-06,
"loss": 2.3682,
"step": 81
},
{
"epoch": 0.059593023255813955,
"grad_norm": 1.821989967160763,
"learning_rate": 9.986520280403775e-06,
"loss": 2.3528,
"step": 82
},
{
"epoch": 0.06031976744186047,
"grad_norm": 1.2376733004557532,
"learning_rate": 9.985637017626326e-06,
"loss": 2.3695,
"step": 83
},
{
"epoch": 0.061046511627906974,
"grad_norm": 1.173034231871182,
"learning_rate": 9.984725769343737e-06,
"loss": 2.279,
"step": 84
},
{
"epoch": 0.06177325581395349,
"grad_norm": 1.1918956966775534,
"learning_rate": 9.983786540671052e-06,
"loss": 2.3667,
"step": 85
},
{
"epoch": 0.0625,
"grad_norm": 1.2273572948900426,
"learning_rate": 9.982819336880369e-06,
"loss": 2.3373,
"step": 86
},
{
"epoch": 0.06322674418604651,
"grad_norm": 1.1231561709191837,
"learning_rate": 9.981824163400827e-06,
"loss": 2.3194,
"step": 87
},
{
"epoch": 0.06395348837209303,
"grad_norm": 1.202004965525141,
"learning_rate": 9.980801025818556e-06,
"loss": 2.3587,
"step": 88
},
{
"epoch": 0.06468023255813954,
"grad_norm": 1.2282314740315383,
"learning_rate": 9.979749929876658e-06,
"loss": 2.3797,
"step": 89
},
{
"epoch": 0.06540697674418605,
"grad_norm": 1.3144649205068073,
"learning_rate": 9.978670881475173e-06,
"loss": 2.4046,
"step": 90
},
{
"epoch": 0.06613372093023256,
"grad_norm": 1.3880094986440303,
"learning_rate": 9.977563886671043e-06,
"loss": 2.3486,
"step": 91
},
{
"epoch": 0.06686046511627906,
"grad_norm": 1.327629914175587,
"learning_rate": 9.976428951678077e-06,
"loss": 2.3443,
"step": 92
},
{
"epoch": 0.06758720930232558,
"grad_norm": 1.4219965859285533,
"learning_rate": 9.975266082866923e-06,
"loss": 2.4261,
"step": 93
},
{
"epoch": 0.06831395348837209,
"grad_norm": 1.2182939431389077,
"learning_rate": 9.974075286765027e-06,
"loss": 2.3691,
"step": 94
},
{
"epoch": 0.0690406976744186,
"grad_norm": 1.1061466662626884,
"learning_rate": 9.972856570056594e-06,
"loss": 2.3745,
"step": 95
},
{
"epoch": 0.06976744186046512,
"grad_norm": 1.1094985189262911,
"learning_rate": 9.971609939582556e-06,
"loss": 2.3647,
"step": 96
},
{
"epoch": 0.07049418604651163,
"grad_norm": 1.1428377874275164,
"learning_rate": 9.970335402340534e-06,
"loss": 2.326,
"step": 97
},
{
"epoch": 0.07122093023255814,
"grad_norm": 1.2916043783430748,
"learning_rate": 9.969032965484789e-06,
"loss": 2.4184,
"step": 98
},
{
"epoch": 0.07194767441860465,
"grad_norm": 1.1245599922593577,
"learning_rate": 9.967702636326195e-06,
"loss": 2.3263,
"step": 99
},
{
"epoch": 0.07267441860465117,
"grad_norm": 1.1704821318393857,
"learning_rate": 9.96634442233219e-06,
"loss": 2.3092,
"step": 100
},
{
"epoch": 0.07340116279069768,
"grad_norm": 1.4100861747980609,
"learning_rate": 9.964958331126735e-06,
"loss": 2.3648,
"step": 101
},
{
"epoch": 0.07412790697674419,
"grad_norm": 1.2225339934583712,
"learning_rate": 9.96354437049027e-06,
"loss": 2.3505,
"step": 102
},
{
"epoch": 0.07485465116279069,
"grad_norm": 1.1411365305964143,
"learning_rate": 9.96210254835968e-06,
"loss": 2.3503,
"step": 103
},
{
"epoch": 0.0755813953488372,
"grad_norm": 1.1283411831087642,
"learning_rate": 9.960632872828233e-06,
"loss": 2.3763,
"step": 104
},
{
"epoch": 0.07630813953488372,
"grad_norm": 1.4669332760539098,
"learning_rate": 9.959135352145552e-06,
"loss": 2.3131,
"step": 105
},
{
"epoch": 0.07703488372093023,
"grad_norm": 1.1790811219137445,
"learning_rate": 9.957609994717559e-06,
"loss": 2.3473,
"step": 106
},
{
"epoch": 0.07776162790697674,
"grad_norm": 1.1041919220536518,
"learning_rate": 9.956056809106426e-06,
"loss": 2.2556,
"step": 107
},
{
"epoch": 0.07848837209302326,
"grad_norm": 1.3219326629276105,
"learning_rate": 9.954475804030539e-06,
"loss": 2.3901,
"step": 108
},
{
"epoch": 0.07921511627906977,
"grad_norm": 1.1045288858846065,
"learning_rate": 9.952866988364431e-06,
"loss": 2.3432,
"step": 109
},
{
"epoch": 0.07994186046511628,
"grad_norm": 1.2740973453849955,
"learning_rate": 9.95123037113875e-06,
"loss": 2.4174,
"step": 110
},
{
"epoch": 0.0806686046511628,
"grad_norm": 1.192838557735101,
"learning_rate": 9.9495659615402e-06,
"loss": 2.378,
"step": 111
},
{
"epoch": 0.08139534883720931,
"grad_norm": 1.1316106615191093,
"learning_rate": 9.947873768911483e-06,
"loss": 2.3669,
"step": 112
},
{
"epoch": 0.08212209302325581,
"grad_norm": 1.1953371633708236,
"learning_rate": 9.946153802751257e-06,
"loss": 2.3339,
"step": 113
},
{
"epoch": 0.08284883720930232,
"grad_norm": 1.1589422598242156,
"learning_rate": 9.944406072714086e-06,
"loss": 2.3633,
"step": 114
},
{
"epoch": 0.08357558139534883,
"grad_norm": 1.1665993258833627,
"learning_rate": 9.942630588610368e-06,
"loss": 2.3718,
"step": 115
},
{
"epoch": 0.08430232558139535,
"grad_norm": 1.2252914448806391,
"learning_rate": 9.940827360406297e-06,
"loss": 2.3757,
"step": 116
},
{
"epoch": 0.08502906976744186,
"grad_norm": 1.99966100562339,
"learning_rate": 9.938996398223802e-06,
"loss": 2.3849,
"step": 117
},
{
"epoch": 0.08575581395348837,
"grad_norm": 1.2354896948177605,
"learning_rate": 9.937137712340483e-06,
"loss": 2.3991,
"step": 118
},
{
"epoch": 0.08648255813953488,
"grad_norm": 1.246354759029233,
"learning_rate": 9.935251313189564e-06,
"loss": 2.4023,
"step": 119
},
{
"epoch": 0.0872093023255814,
"grad_norm": 1.2116770658459601,
"learning_rate": 9.933337211359833e-06,
"loss": 2.3887,
"step": 120
},
{
"epoch": 0.08793604651162791,
"grad_norm": 1.2249695456790302,
"learning_rate": 9.931395417595568e-06,
"loss": 2.373,
"step": 121
},
{
"epoch": 0.08866279069767442,
"grad_norm": 1.0642820314257344,
"learning_rate": 9.929425942796502e-06,
"loss": 2.3351,
"step": 122
},
{
"epoch": 0.08938953488372094,
"grad_norm": 1.2743502150368227,
"learning_rate": 9.927428798017738e-06,
"loss": 2.3391,
"step": 123
},
{
"epoch": 0.09011627906976744,
"grad_norm": 1.3340646666078864,
"learning_rate": 9.925403994469702e-06,
"loss": 2.4218,
"step": 124
},
{
"epoch": 0.09084302325581395,
"grad_norm": 1.566724790718566,
"learning_rate": 9.92335154351807e-06,
"loss": 2.3399,
"step": 125
},
{
"epoch": 0.09156976744186046,
"grad_norm": 1.0741534822566858,
"learning_rate": 9.921271456683716e-06,
"loss": 2.4141,
"step": 126
},
{
"epoch": 0.09229651162790697,
"grad_norm": 1.1975939734272736,
"learning_rate": 9.919163745642633e-06,
"loss": 2.3331,
"step": 127
},
{
"epoch": 0.09302325581395349,
"grad_norm": 1.26408679935078,
"learning_rate": 9.91702842222588e-06,
"loss": 2.3798,
"step": 128
},
{
"epoch": 0.09375,
"grad_norm": 1.2052694508772792,
"learning_rate": 9.91486549841951e-06,
"loss": 2.3746,
"step": 129
},
{
"epoch": 0.09447674418604651,
"grad_norm": 1.1489590029014498,
"learning_rate": 9.912674986364502e-06,
"loss": 2.3986,
"step": 130
},
{
"epoch": 0.09520348837209303,
"grad_norm": 1.2440161112749069,
"learning_rate": 9.91045689835669e-06,
"loss": 2.3182,
"step": 131
},
{
"epoch": 0.09593023255813954,
"grad_norm": 1.2574929722039276,
"learning_rate": 9.908211246846708e-06,
"loss": 2.3595,
"step": 132
},
{
"epoch": 0.09665697674418605,
"grad_norm": 1.1422750235883177,
"learning_rate": 9.905938044439904e-06,
"loss": 2.3402,
"step": 133
},
{
"epoch": 0.09738372093023256,
"grad_norm": 1.1320332374781519,
"learning_rate": 9.903637303896272e-06,
"loss": 2.3584,
"step": 134
},
{
"epoch": 0.09811046511627906,
"grad_norm": 1.182873125632428,
"learning_rate": 9.901309038130392e-06,
"loss": 2.3266,
"step": 135
},
{
"epoch": 0.09883720930232558,
"grad_norm": 1.155380615425545,
"learning_rate": 9.89895326021134e-06,
"loss": 2.3379,
"step": 136
},
{
"epoch": 0.09956395348837209,
"grad_norm": 1.0407717760927417,
"learning_rate": 9.896569983362632e-06,
"loss": 2.3071,
"step": 137
},
{
"epoch": 0.1002906976744186,
"grad_norm": 1.1815278271760303,
"learning_rate": 9.894159220962138e-06,
"loss": 2.3706,
"step": 138
},
{
"epoch": 0.10101744186046512,
"grad_norm": 1.7394010037009506,
"learning_rate": 9.891720986542011e-06,
"loss": 2.2908,
"step": 139
},
{
"epoch": 0.10174418604651163,
"grad_norm": 1.247501930958456,
"learning_rate": 9.889255293788613e-06,
"loss": 2.2802,
"step": 140
},
{
"epoch": 0.10247093023255814,
"grad_norm": 1.376771705154438,
"learning_rate": 9.886762156542428e-06,
"loss": 2.3189,
"step": 141
},
{
"epoch": 0.10319767441860465,
"grad_norm": 1.0539145361934386,
"learning_rate": 9.884241588798004e-06,
"loss": 2.3588,
"step": 142
},
{
"epoch": 0.10392441860465117,
"grad_norm": 1.117473895172851,
"learning_rate": 9.881693604703853e-06,
"loss": 2.3647,
"step": 143
},
{
"epoch": 0.10465116279069768,
"grad_norm": 1.1230253756584534,
"learning_rate": 9.879118218562384e-06,
"loss": 2.3987,
"step": 144
},
{
"epoch": 0.10537790697674419,
"grad_norm": 1.0869199427758438,
"learning_rate": 9.876515444829822e-06,
"loss": 2.3608,
"step": 145
},
{
"epoch": 0.10610465116279069,
"grad_norm": 1.175972616599383,
"learning_rate": 9.873885298116123e-06,
"loss": 2.3339,
"step": 146
},
{
"epoch": 0.1068313953488372,
"grad_norm": 1.1287432192217346,
"learning_rate": 9.871227793184893e-06,
"loss": 2.3134,
"step": 147
},
{
"epoch": 0.10755813953488372,
"grad_norm": 1.455085174866513,
"learning_rate": 9.868542944953304e-06,
"loss": 2.3109,
"step": 148
},
{
"epoch": 0.10828488372093023,
"grad_norm": 1.175215614103579,
"learning_rate": 9.865830768492019e-06,
"loss": 2.3118,
"step": 149
},
{
"epoch": 0.10901162790697674,
"grad_norm": 1.1557846646815062,
"learning_rate": 9.863091279025095e-06,
"loss": 2.2978,
"step": 150
},
{
"epoch": 0.10973837209302326,
"grad_norm": 1.1458120293970389,
"learning_rate": 9.860324491929905e-06,
"loss": 2.3072,
"step": 151
},
{
"epoch": 0.11046511627906977,
"grad_norm": 1.3189900091924425,
"learning_rate": 9.857530422737045e-06,
"loss": 2.3174,
"step": 152
},
{
"epoch": 0.11119186046511628,
"grad_norm": 1.1143078767141845,
"learning_rate": 9.854709087130261e-06,
"loss": 2.3583,
"step": 153
},
{
"epoch": 0.1119186046511628,
"grad_norm": 1.1170390691615044,
"learning_rate": 9.851860500946342e-06,
"loss": 2.3226,
"step": 154
},
{
"epoch": 0.11264534883720931,
"grad_norm": 1.1117837847322765,
"learning_rate": 9.848984680175049e-06,
"loss": 2.3086,
"step": 155
},
{
"epoch": 0.11337209302325581,
"grad_norm": 1.5822415450769225,
"learning_rate": 9.846081640959008e-06,
"loss": 2.3525,
"step": 156
},
{
"epoch": 0.11409883720930232,
"grad_norm": 1.4781679250339181,
"learning_rate": 9.843151399593636e-06,
"loss": 2.3371,
"step": 157
},
{
"epoch": 0.11482558139534883,
"grad_norm": 1.1460798573233948,
"learning_rate": 9.840193972527037e-06,
"loss": 2.238,
"step": 158
},
{
"epoch": 0.11555232558139535,
"grad_norm": 1.2596341406617226,
"learning_rate": 9.837209376359918e-06,
"loss": 2.3571,
"step": 159
},
{
"epoch": 0.11627906976744186,
"grad_norm": 1.1138886353351996,
"learning_rate": 9.834197627845488e-06,
"loss": 2.3476,
"step": 160
},
{
"epoch": 0.11700581395348837,
"grad_norm": 1.141413584651315,
"learning_rate": 9.831158743889373e-06,
"loss": 2.3456,
"step": 161
},
{
"epoch": 0.11773255813953488,
"grad_norm": 1.1691072792164279,
"learning_rate": 9.828092741549513e-06,
"loss": 2.3417,
"step": 162
},
{
"epoch": 0.1184593023255814,
"grad_norm": 1.2184258738041083,
"learning_rate": 9.82499963803607e-06,
"loss": 2.267,
"step": 163
},
{
"epoch": 0.11918604651162791,
"grad_norm": 1.1576173045243061,
"learning_rate": 9.821879450711336e-06,
"loss": 2.3378,
"step": 164
},
{
"epoch": 0.11991279069767442,
"grad_norm": 1.0187703916093862,
"learning_rate": 9.81873219708962e-06,
"loss": 2.3258,
"step": 165
},
{
"epoch": 0.12063953488372094,
"grad_norm": 1.1463683129424287,
"learning_rate": 9.815557894837171e-06,
"loss": 2.3341,
"step": 166
},
{
"epoch": 0.12136627906976744,
"grad_norm": 1.1232938117798668,
"learning_rate": 9.81235656177206e-06,
"loss": 2.3387,
"step": 167
},
{
"epoch": 0.12209302325581395,
"grad_norm": 1.448813798580779,
"learning_rate": 9.809128215864096e-06,
"loss": 2.3542,
"step": 168
},
{
"epoch": 0.12281976744186046,
"grad_norm": 1.2584761113731575,
"learning_rate": 9.80587287523471e-06,
"loss": 2.328,
"step": 169
},
{
"epoch": 0.12354651162790697,
"grad_norm": 1.1315037986766692,
"learning_rate": 9.802590558156863e-06,
"loss": 2.3566,
"step": 170
},
{
"epoch": 0.12427325581395349,
"grad_norm": 1.124447003450953,
"learning_rate": 9.79928128305494e-06,
"loss": 2.3846,
"step": 171
},
{
"epoch": 0.125,
"grad_norm": 1.3268576225981397,
"learning_rate": 9.795945068504654e-06,
"loss": 2.3277,
"step": 172
},
{
"epoch": 0.1257267441860465,
"grad_norm": 1.104840140367407,
"learning_rate": 9.792581933232924e-06,
"loss": 2.3476,
"step": 173
},
{
"epoch": 0.12645348837209303,
"grad_norm": 1.2928119161216733,
"learning_rate": 9.789191896117786e-06,
"loss": 2.3244,
"step": 174
},
{
"epoch": 0.12718023255813954,
"grad_norm": 1.2396436814423948,
"learning_rate": 9.78577497618829e-06,
"loss": 2.353,
"step": 175
},
{
"epoch": 0.12790697674418605,
"grad_norm": 1.103208759457441,
"learning_rate": 9.782331192624372e-06,
"loss": 2.383,
"step": 176
},
{
"epoch": 0.12863372093023256,
"grad_norm": 2.0646273194959797,
"learning_rate": 9.778860564756769e-06,
"loss": 2.3285,
"step": 177
},
{
"epoch": 0.12936046511627908,
"grad_norm": 1.1841117516704331,
"learning_rate": 9.775363112066897e-06,
"loss": 2.3602,
"step": 178
},
{
"epoch": 0.1300872093023256,
"grad_norm": 1.465105518576992,
"learning_rate": 9.771838854186748e-06,
"loss": 2.3446,
"step": 179
},
{
"epoch": 0.1308139534883721,
"grad_norm": 1.2274769216909944,
"learning_rate": 9.768287810898773e-06,
"loss": 2.3527,
"step": 180
},
{
"epoch": 0.13154069767441862,
"grad_norm": 1.778850283425946,
"learning_rate": 9.764710002135784e-06,
"loss": 2.4089,
"step": 181
},
{
"epoch": 0.13226744186046513,
"grad_norm": 1.2955538543537204,
"learning_rate": 9.761105447980824e-06,
"loss": 2.3727,
"step": 182
},
{
"epoch": 0.13299418604651161,
"grad_norm": 1.8038989790897118,
"learning_rate": 9.757474168667072e-06,
"loss": 2.3779,
"step": 183
},
{
"epoch": 0.13372093023255813,
"grad_norm": 1.4246329229100207,
"learning_rate": 9.753816184577715e-06,
"loss": 2.3381,
"step": 184
},
{
"epoch": 0.13444767441860464,
"grad_norm": 1.5788899090032442,
"learning_rate": 9.750131516245844e-06,
"loss": 2.2993,
"step": 185
},
{
"epoch": 0.13517441860465115,
"grad_norm": 1.485106172355965,
"learning_rate": 9.746420184354334e-06,
"loss": 2.3213,
"step": 186
},
{
"epoch": 0.13590116279069767,
"grad_norm": 1.424583265290273,
"learning_rate": 9.742682209735727e-06,
"loss": 2.3239,
"step": 187
},
{
"epoch": 0.13662790697674418,
"grad_norm": 1.3327881052111739,
"learning_rate": 9.738917613372121e-06,
"loss": 2.3294,
"step": 188
},
{
"epoch": 0.1373546511627907,
"grad_norm": 1.28657064764393,
"learning_rate": 9.73512641639504e-06,
"loss": 2.3268,
"step": 189
},
{
"epoch": 0.1380813953488372,
"grad_norm": 1.325094868563775,
"learning_rate": 9.731308640085329e-06,
"loss": 2.3798,
"step": 190
},
{
"epoch": 0.13880813953488372,
"grad_norm": 1.417161190490927,
"learning_rate": 9.72746430587303e-06,
"loss": 2.3363,
"step": 191
},
{
"epoch": 0.13953488372093023,
"grad_norm": 1.44872426301407,
"learning_rate": 9.723593435337252e-06,
"loss": 2.3349,
"step": 192
},
{
"epoch": 0.14026162790697674,
"grad_norm": 1.5448029589257748,
"learning_rate": 9.719696050206072e-06,
"loss": 2.3632,
"step": 193
},
{
"epoch": 0.14098837209302326,
"grad_norm": 1.5859763359793375,
"learning_rate": 9.715772172356388e-06,
"loss": 2.2827,
"step": 194
},
{
"epoch": 0.14171511627906977,
"grad_norm": 1.184747752110052,
"learning_rate": 9.711821823813812e-06,
"loss": 2.3393,
"step": 195
},
{
"epoch": 0.14244186046511628,
"grad_norm": 1.4867726022411942,
"learning_rate": 9.70784502675254e-06,
"loss": 2.3275,
"step": 196
},
{
"epoch": 0.1431686046511628,
"grad_norm": 1.1538729596914152,
"learning_rate": 9.703841803495234e-06,
"loss": 2.3359,
"step": 197
},
{
"epoch": 0.1438953488372093,
"grad_norm": 1.6929885708609627,
"learning_rate": 9.699812176512887e-06,
"loss": 2.3807,
"step": 198
},
{
"epoch": 0.14462209302325582,
"grad_norm": 1.1679128487047015,
"learning_rate": 9.695756168424703e-06,
"loss": 2.2929,
"step": 199
},
{
"epoch": 0.14534883720930233,
"grad_norm": 3.228739152070278,
"learning_rate": 9.691673801997974e-06,
"loss": 2.3188,
"step": 200
},
{
"epoch": 0.14607558139534885,
"grad_norm": 2.0854780633686856,
"learning_rate": 9.68756510014794e-06,
"loss": 2.3525,
"step": 201
},
{
"epoch": 0.14680232558139536,
"grad_norm": 1.4863284919761375,
"learning_rate": 9.683430085937672e-06,
"loss": 2.3462,
"step": 202
},
{
"epoch": 0.14752906976744187,
"grad_norm": 1.7980518257689877,
"learning_rate": 9.67926878257794e-06,
"loss": 2.3268,
"step": 203
},
{
"epoch": 0.14825581395348839,
"grad_norm": 1.4394065482189453,
"learning_rate": 9.675081213427076e-06,
"loss": 2.3802,
"step": 204
},
{
"epoch": 0.14898255813953487,
"grad_norm": 2.281476790918211,
"learning_rate": 9.67086740199085e-06,
"loss": 2.2903,
"step": 205
},
{
"epoch": 0.14970930232558138,
"grad_norm": 2.1550730313406596,
"learning_rate": 9.666627371922335e-06,
"loss": 2.3204,
"step": 206
},
{
"epoch": 0.1504360465116279,
"grad_norm": 1.1250028145242355,
"learning_rate": 9.66236114702178e-06,
"loss": 2.3396,
"step": 207
},
{
"epoch": 0.1511627906976744,
"grad_norm": 1.547403727782741,
"learning_rate": 9.658068751236464e-06,
"loss": 2.3074,
"step": 208
},
{
"epoch": 0.15188953488372092,
"grad_norm": 1.274380714094946,
"learning_rate": 9.653750208660577e-06,
"loss": 2.3541,
"step": 209
},
{
"epoch": 0.15261627906976744,
"grad_norm": 2.134091901741687,
"learning_rate": 9.649405543535067e-06,
"loss": 2.3642,
"step": 210
},
{
"epoch": 0.15334302325581395,
"grad_norm": 1.8269953720467489,
"learning_rate": 9.645034780247521e-06,
"loss": 2.2898,
"step": 211
},
{
"epoch": 0.15406976744186046,
"grad_norm": 1.2440228463941307,
"learning_rate": 9.640637943332025e-06,
"loss": 2.3005,
"step": 212
},
{
"epoch": 0.15479651162790697,
"grad_norm": 1.6575430787365315,
"learning_rate": 9.636215057469009e-06,
"loss": 2.3413,
"step": 213
},
{
"epoch": 0.1555232558139535,
"grad_norm": 1.1243540368900853,
"learning_rate": 9.631766147485131e-06,
"loss": 2.3274,
"step": 214
},
{
"epoch": 0.15625,
"grad_norm": 1.5183470450257768,
"learning_rate": 9.627291238353127e-06,
"loss": 2.3051,
"step": 215
},
{
"epoch": 0.1569767441860465,
"grad_norm": 1.5119397631588836,
"learning_rate": 9.622790355191672e-06,
"loss": 2.3259,
"step": 216
},
{
"epoch": 0.15770348837209303,
"grad_norm": 1.4922176869104902,
"learning_rate": 9.618263523265238e-06,
"loss": 2.3639,
"step": 217
},
{
"epoch": 0.15843023255813954,
"grad_norm": 1.4730383961473745,
"learning_rate": 9.613710767983953e-06,
"loss": 2.3338,
"step": 218
},
{
"epoch": 0.15915697674418605,
"grad_norm": 1.2439501029680762,
"learning_rate": 9.609132114903458e-06,
"loss": 2.3107,
"step": 219
},
{
"epoch": 0.15988372093023256,
"grad_norm": 1.2143771364643452,
"learning_rate": 9.60452758972477e-06,
"loss": 2.3044,
"step": 220
},
{
"epoch": 0.16061046511627908,
"grad_norm": 1.2334867243523404,
"learning_rate": 9.599897218294122e-06,
"loss": 2.2814,
"step": 221
},
{
"epoch": 0.1613372093023256,
"grad_norm": 1.156597773993428,
"learning_rate": 9.595241026602836e-06,
"loss": 2.2877,
"step": 222
},
{
"epoch": 0.1620639534883721,
"grad_norm": 1.2761713517213147,
"learning_rate": 9.590559040787168e-06,
"loss": 2.328,
"step": 223
},
{
"epoch": 0.16279069767441862,
"grad_norm": 1.7761321590409704,
"learning_rate": 9.585851287128157e-06,
"loss": 2.3159,
"step": 224
},
{
"epoch": 0.16351744186046513,
"grad_norm": 1.163745695967623,
"learning_rate": 9.581117792051487e-06,
"loss": 2.3469,
"step": 225
},
{
"epoch": 0.16424418604651161,
"grad_norm": 1.133988790541161,
"learning_rate": 9.576358582127334e-06,
"loss": 2.3531,
"step": 226
},
{
"epoch": 0.16497093023255813,
"grad_norm": 1.1024750427768464,
"learning_rate": 9.57157368407022e-06,
"loss": 2.3779,
"step": 227
},
{
"epoch": 0.16569767441860464,
"grad_norm": 1.2517859254831207,
"learning_rate": 9.56676312473885e-06,
"loss": 2.2703,
"step": 228
},
{
"epoch": 0.16642441860465115,
"grad_norm": 1.073885261058524,
"learning_rate": 9.561926931135985e-06,
"loss": 2.3131,
"step": 229
},
{
"epoch": 0.16715116279069767,
"grad_norm": 1.1127469264086771,
"learning_rate": 9.557065130408267e-06,
"loss": 2.3835,
"step": 230
},
{
"epoch": 0.16787790697674418,
"grad_norm": 1.1959973006506952,
"learning_rate": 9.552177749846083e-06,
"loss": 2.3338,
"step": 231
},
{
"epoch": 0.1686046511627907,
"grad_norm": 1.1058947069586706,
"learning_rate": 9.5472648168834e-06,
"loss": 2.3255,
"step": 232
},
{
"epoch": 0.1693313953488372,
"grad_norm": 1.116754555488762,
"learning_rate": 9.542326359097619e-06,
"loss": 2.3494,
"step": 233
},
{
"epoch": 0.17005813953488372,
"grad_norm": 1.0876854570886407,
"learning_rate": 9.537362404209419e-06,
"loss": 2.3791,
"step": 234
},
{
"epoch": 0.17078488372093023,
"grad_norm": 1.2137372127091681,
"learning_rate": 9.532372980082598e-06,
"loss": 2.2718,
"step": 235
},
{
"epoch": 0.17151162790697674,
"grad_norm": 1.129260401791111,
"learning_rate": 9.527358114723917e-06,
"loss": 2.3639,
"step": 236
},
{
"epoch": 0.17223837209302326,
"grad_norm": 1.2340740847840952,
"learning_rate": 9.522317836282949e-06,
"loss": 2.3132,
"step": 237
},
{
"epoch": 0.17296511627906977,
"grad_norm": 1.1219295749886464,
"learning_rate": 9.517252173051912e-06,
"loss": 2.3437,
"step": 238
},
{
"epoch": 0.17369186046511628,
"grad_norm": 1.1822380144608287,
"learning_rate": 9.512161153465518e-06,
"loss": 2.3267,
"step": 239
},
{
"epoch": 0.1744186046511628,
"grad_norm": 1.1180365462119664,
"learning_rate": 9.507044806100806e-06,
"loss": 2.2838,
"step": 240
},
{
"epoch": 0.1751453488372093,
"grad_norm": 1.2103630698428618,
"learning_rate": 9.501903159676993e-06,
"loss": 2.3157,
"step": 241
},
{
"epoch": 0.17587209302325582,
"grad_norm": 1.6042659029711763,
"learning_rate": 9.496736243055293e-06,
"loss": 2.302,
"step": 242
},
{
"epoch": 0.17659883720930233,
"grad_norm": 1.142876951018072,
"learning_rate": 9.491544085238778e-06,
"loss": 2.3086,
"step": 243
},
{
"epoch": 0.17732558139534885,
"grad_norm": 1.2940579158130927,
"learning_rate": 9.486326715372201e-06,
"loss": 2.2899,
"step": 244
},
{
"epoch": 0.17805232558139536,
"grad_norm": 1.2892835086318652,
"learning_rate": 9.481084162741835e-06,
"loss": 2.3795,
"step": 245
},
{
"epoch": 0.17877906976744187,
"grad_norm": 1.0727384912534383,
"learning_rate": 9.475816456775313e-06,
"loss": 2.3633,
"step": 246
},
{
"epoch": 0.17950581395348839,
"grad_norm": 1.1365478569616185,
"learning_rate": 9.470523627041452e-06,
"loss": 2.3659,
"step": 247
},
{
"epoch": 0.18023255813953487,
"grad_norm": 1.1291795803936022,
"learning_rate": 9.465205703250105e-06,
"loss": 2.325,
"step": 248
},
{
"epoch": 0.18095930232558138,
"grad_norm": 1.2835767823611957,
"learning_rate": 9.459862715251973e-06,
"loss": 2.2822,
"step": 249
},
{
"epoch": 0.1816860465116279,
"grad_norm": 1.2029135874825383,
"learning_rate": 9.454494693038455e-06,
"loss": 2.3935,
"step": 250
},
{
"epoch": 0.1824127906976744,
"grad_norm": 1.134998330696039,
"learning_rate": 9.44910166674147e-06,
"loss": 2.3326,
"step": 251
},
{
"epoch": 0.18313953488372092,
"grad_norm": 1.2024738851493837,
"learning_rate": 9.44368366663329e-06,
"loss": 2.3289,
"step": 252
},
{
"epoch": 0.18386627906976744,
"grad_norm": 1.204843905841385,
"learning_rate": 9.438240723126376e-06,
"loss": 2.3165,
"step": 253
},
{
"epoch": 0.18459302325581395,
"grad_norm": 1.364897858166092,
"learning_rate": 9.43277286677319e-06,
"loss": 2.3432,
"step": 254
},
{
"epoch": 0.18531976744186046,
"grad_norm": 1.1892133781450915,
"learning_rate": 9.427280128266049e-06,
"loss": 2.329,
"step": 255
},
{
"epoch": 0.18604651162790697,
"grad_norm": 1.2959664242769635,
"learning_rate": 9.421762538436933e-06,
"loss": 2.3469,
"step": 256
},
{
"epoch": 0.1867732558139535,
"grad_norm": 1.4188951604829714,
"learning_rate": 9.416220128257317e-06,
"loss": 2.3602,
"step": 257
},
{
"epoch": 0.1875,
"grad_norm": 1.15071972840638,
"learning_rate": 9.410652928837998e-06,
"loss": 2.3189,
"step": 258
},
{
"epoch": 0.1882267441860465,
"grad_norm": 1.1635890361824703,
"learning_rate": 9.405060971428924e-06,
"loss": 2.3282,
"step": 259
},
{
"epoch": 0.18895348837209303,
"grad_norm": 1.173051490771907,
"learning_rate": 9.399444287419012e-06,
"loss": 2.3391,
"step": 260
},
{
"epoch": 0.18968023255813954,
"grad_norm": 1.4480462335454614,
"learning_rate": 9.393802908335978e-06,
"loss": 2.3273,
"step": 261
},
{
"epoch": 0.19040697674418605,
"grad_norm": 1.3645358660521583,
"learning_rate": 9.388136865846153e-06,
"loss": 2.3126,
"step": 262
},
{
"epoch": 0.19113372093023256,
"grad_norm": 1.3259099589635823,
"learning_rate": 9.382446191754313e-06,
"loss": 2.3002,
"step": 263
},
{
"epoch": 0.19186046511627908,
"grad_norm": 1.2243958605804686,
"learning_rate": 9.376730918003495e-06,
"loss": 2.2832,
"step": 264
},
{
"epoch": 0.1925872093023256,
"grad_norm": 1.3154278911242208,
"learning_rate": 9.370991076674821e-06,
"loss": 2.3551,
"step": 265
},
{
"epoch": 0.1933139534883721,
"grad_norm": 1.1106953939990531,
"learning_rate": 9.36522669998731e-06,
"loss": 2.2829,
"step": 266
},
{
"epoch": 0.19404069767441862,
"grad_norm": 1.1959726898905372,
"learning_rate": 9.359437820297716e-06,
"loss": 2.3021,
"step": 267
},
{
"epoch": 0.19476744186046513,
"grad_norm": 1.227387261796965,
"learning_rate": 9.353624470100321e-06,
"loss": 2.3884,
"step": 268
},
{
"epoch": 0.19549418604651161,
"grad_norm": 1.4504794371079932,
"learning_rate": 9.347786682026774e-06,
"loss": 2.342,
"step": 269
},
{
"epoch": 0.19622093023255813,
"grad_norm": 1.1512546090637321,
"learning_rate": 9.341924488845892e-06,
"loss": 2.3033,
"step": 270
},
{
"epoch": 0.19694767441860464,
"grad_norm": 1.4217379552780167,
"learning_rate": 9.336037923463494e-06,
"loss": 2.3293,
"step": 271
},
{
"epoch": 0.19767441860465115,
"grad_norm": 1.3571520758260824,
"learning_rate": 9.330127018922195e-06,
"loss": 2.2994,
"step": 272
},
{
"epoch": 0.19840116279069767,
"grad_norm": 1.4540212433725141,
"learning_rate": 9.324191808401235e-06,
"loss": 2.2782,
"step": 273
},
{
"epoch": 0.19912790697674418,
"grad_norm": 1.4077888491936041,
"learning_rate": 9.31823232521629e-06,
"loss": 2.4051,
"step": 274
},
{
"epoch": 0.1998546511627907,
"grad_norm": 1.2165690102619144,
"learning_rate": 9.312248602819284e-06,
"loss": 2.3455,
"step": 275
},
{
"epoch": 0.2005813953488372,
"grad_norm": 1.2756272884495146,
"learning_rate": 9.306240674798203e-06,
"loss": 2.3318,
"step": 276
},
{
"epoch": 0.20130813953488372,
"grad_norm": 1.56568088675301,
"learning_rate": 9.300208574876897e-06,
"loss": 2.3476,
"step": 277
},
{
"epoch": 0.20203488372093023,
"grad_norm": 1.2769102850208098,
"learning_rate": 9.294152336914907e-06,
"loss": 2.3364,
"step": 278
},
{
"epoch": 0.20276162790697674,
"grad_norm": 1.1982423239697224,
"learning_rate": 9.288071994907262e-06,
"loss": 2.3306,
"step": 279
},
{
"epoch": 0.20348837209302326,
"grad_norm": 1.1679661222509305,
"learning_rate": 9.281967582984292e-06,
"loss": 2.3281,
"step": 280
},
{
"epoch": 0.20421511627906977,
"grad_norm": 1.4369131979668328,
"learning_rate": 9.275839135411439e-06,
"loss": 2.362,
"step": 281
},
{
"epoch": 0.20494186046511628,
"grad_norm": 1.1740960575913133,
"learning_rate": 9.269686686589063e-06,
"loss": 2.3215,
"step": 282
},
{
"epoch": 0.2056686046511628,
"grad_norm": 1.2289991464520638,
"learning_rate": 9.263510271052243e-06,
"loss": 2.3116,
"step": 283
},
{
"epoch": 0.2063953488372093,
"grad_norm": 1.1411155539417916,
"learning_rate": 9.257309923470596e-06,
"loss": 2.3149,
"step": 284
},
{
"epoch": 0.20712209302325582,
"grad_norm": 1.3599826720526693,
"learning_rate": 9.251085678648072e-06,
"loss": 2.329,
"step": 285
},
{
"epoch": 0.20784883720930233,
"grad_norm": 1.1828225896311113,
"learning_rate": 9.244837571522758e-06,
"loss": 2.3205,
"step": 286
},
{
"epoch": 0.20857558139534885,
"grad_norm": 1.3890914247072965,
"learning_rate": 9.238565637166692e-06,
"loss": 2.3309,
"step": 287
},
{
"epoch": 0.20930232558139536,
"grad_norm": 1.123989679885759,
"learning_rate": 9.232269910785651e-06,
"loss": 2.3634,
"step": 288
},
{
"epoch": 0.21002906976744187,
"grad_norm": 1.3138347474424352,
"learning_rate": 9.225950427718974e-06,
"loss": 2.3247,
"step": 289
},
{
"epoch": 0.21075581395348839,
"grad_norm": 1.2441549630025073,
"learning_rate": 9.219607223439343e-06,
"loss": 2.2984,
"step": 290
},
{
"epoch": 0.21148255813953487,
"grad_norm": 1.3393817124230734,
"learning_rate": 9.213240333552589e-06,
"loss": 2.3352,
"step": 291
},
{
"epoch": 0.21220930232558138,
"grad_norm": 1.1731832742101422,
"learning_rate": 9.206849793797508e-06,
"loss": 2.3193,
"step": 292
},
{
"epoch": 0.2129360465116279,
"grad_norm": 1.250178641486035,
"learning_rate": 9.200435640045637e-06,
"loss": 2.3277,
"step": 293
},
{
"epoch": 0.2136627906976744,
"grad_norm": 2.0249942406767008,
"learning_rate": 9.193997908301069e-06,
"loss": 2.3586,
"step": 294
},
{
"epoch": 0.21438953488372092,
"grad_norm": 1.1502433960463287,
"learning_rate": 9.187536634700244e-06,
"loss": 2.3408,
"step": 295
},
{
"epoch": 0.21511627906976744,
"grad_norm": 1.5875909129160295,
"learning_rate": 9.181051855511749e-06,
"loss": 2.3472,
"step": 296
},
{
"epoch": 0.21584302325581395,
"grad_norm": 1.134211330712058,
"learning_rate": 9.174543607136111e-06,
"loss": 2.3133,
"step": 297
},
{
"epoch": 0.21656976744186046,
"grad_norm": 1.103871415303576,
"learning_rate": 9.168011926105598e-06,
"loss": 2.32,
"step": 298
},
{
"epoch": 0.21729651162790697,
"grad_norm": 1.143028922746832,
"learning_rate": 9.161456849084007e-06,
"loss": 2.3567,
"step": 299
},
{
"epoch": 0.2180232558139535,
"grad_norm": 1.1202287953390837,
"learning_rate": 9.154878412866465e-06,
"loss": 2.319,
"step": 300
},
{
"epoch": 0.21875,
"grad_norm": 1.202255264308543,
"learning_rate": 9.14827665437922e-06,
"loss": 2.3689,
"step": 301
},
{
"epoch": 0.2194767441860465,
"grad_norm": 1.1530745059451764,
"learning_rate": 9.141651610679427e-06,
"loss": 2.3254,
"step": 302
},
{
"epoch": 0.22020348837209303,
"grad_norm": 1.0561589105093774,
"learning_rate": 9.135003318954954e-06,
"loss": 2.3005,
"step": 303
},
{
"epoch": 0.22093023255813954,
"grad_norm": 1.3437348779767586,
"learning_rate": 9.12833181652416e-06,
"loss": 2.2729,
"step": 304
},
{
"epoch": 0.22165697674418605,
"grad_norm": 1.2231858657294308,
"learning_rate": 9.121637140835696e-06,
"loss": 2.3805,
"step": 305
},
{
"epoch": 0.22238372093023256,
"grad_norm": 1.1186730722598155,
"learning_rate": 9.114919329468283e-06,
"loss": 2.3557,
"step": 306
},
{
"epoch": 0.22311046511627908,
"grad_norm": 1.2141698238590986,
"learning_rate": 9.108178420130514e-06,
"loss": 2.3878,
"step": 307
},
{
"epoch": 0.2238372093023256,
"grad_norm": 1.265580530868728,
"learning_rate": 9.101414450660633e-06,
"loss": 2.363,
"step": 308
},
{
"epoch": 0.2245639534883721,
"grad_norm": 1.1350587482142922,
"learning_rate": 9.094627459026326e-06,
"loss": 2.3463,
"step": 309
},
{
"epoch": 0.22529069767441862,
"grad_norm": 1.2899696901740874,
"learning_rate": 9.087817483324507e-06,
"loss": 2.3027,
"step": 310
},
{
"epoch": 0.22601744186046513,
"grad_norm": 1.2739682151115492,
"learning_rate": 9.08098456178111e-06,
"loss": 2.2623,
"step": 311
},
{
"epoch": 0.22674418604651161,
"grad_norm": 1.1956689065395163,
"learning_rate": 9.074128732750859e-06,
"loss": 2.3667,
"step": 312
},
{
"epoch": 0.22747093023255813,
"grad_norm": 1.3494480852964184,
"learning_rate": 9.067250034717072e-06,
"loss": 2.3396,
"step": 313
},
{
"epoch": 0.22819767441860464,
"grad_norm": 1.2344725135239982,
"learning_rate": 9.060348506291432e-06,
"loss": 2.3715,
"step": 314
},
{
"epoch": 0.22892441860465115,
"grad_norm": 1.3058573727875196,
"learning_rate": 9.053424186213776e-06,
"loss": 2.3339,
"step": 315
},
{
"epoch": 0.22965116279069767,
"grad_norm": 1.2079684988702277,
"learning_rate": 9.046477113351871e-06,
"loss": 2.3097,
"step": 316
},
{
"epoch": 0.23037790697674418,
"grad_norm": 1.5686728723740095,
"learning_rate": 9.039507326701207e-06,
"loss": 2.3182,
"step": 317
},
{
"epoch": 0.2311046511627907,
"grad_norm": 1.132457656225435,
"learning_rate": 9.032514865384767e-06,
"loss": 2.3174,
"step": 318
},
{
"epoch": 0.2318313953488372,
"grad_norm": 1.1861921100803974,
"learning_rate": 9.025499768652817e-06,
"loss": 2.2587,
"step": 319
},
{
"epoch": 0.23255813953488372,
"grad_norm": 1.1734018256359529,
"learning_rate": 9.018462075882673e-06,
"loss": 2.3287,
"step": 320
},
{
"epoch": 0.23328488372093023,
"grad_norm": 1.1680024156022957,
"learning_rate": 9.011401826578492e-06,
"loss": 2.3259,
"step": 321
},
{
"epoch": 0.23401162790697674,
"grad_norm": 1.1241565327725624,
"learning_rate": 9.00431906037105e-06,
"loss": 2.3353,
"step": 322
},
{
"epoch": 0.23473837209302326,
"grad_norm": 1.1633195222967603,
"learning_rate": 8.997213817017508e-06,
"loss": 2.3811,
"step": 323
},
{
"epoch": 0.23546511627906977,
"grad_norm": 1.32393355925031,
"learning_rate": 8.990086136401199e-06,
"loss": 2.2933,
"step": 324
},
{
"epoch": 0.23619186046511628,
"grad_norm": 1.0782668479787727,
"learning_rate": 8.982936058531403e-06,
"loss": 2.2711,
"step": 325
},
{
"epoch": 0.2369186046511628,
"grad_norm": 1.1680539247789243,
"learning_rate": 8.975763623543121e-06,
"loss": 2.2925,
"step": 326
},
{
"epoch": 0.2376453488372093,
"grad_norm": 1.2209650870203186,
"learning_rate": 8.968568871696847e-06,
"loss": 2.3007,
"step": 327
},
{
"epoch": 0.23837209302325582,
"grad_norm": 1.193452920111548,
"learning_rate": 8.961351843378349e-06,
"loss": 2.2965,
"step": 328
},
{
"epoch": 0.23909883720930233,
"grad_norm": 1.2288630443793624,
"learning_rate": 8.95411257909843e-06,
"loss": 2.3947,
"step": 329
},
{
"epoch": 0.23982558139534885,
"grad_norm": 1.1891549047471393,
"learning_rate": 8.946851119492717e-06,
"loss": 2.3313,
"step": 330
},
{
"epoch": 0.24055232558139536,
"grad_norm": 1.3399860656299631,
"learning_rate": 8.939567505321418e-06,
"loss": 2.3004,
"step": 331
},
{
"epoch": 0.24127906976744187,
"grad_norm": 1.1419693810218348,
"learning_rate": 8.932261777469105e-06,
"loss": 2.3239,
"step": 332
},
{
"epoch": 0.24200581395348839,
"grad_norm": 1.476156520390232,
"learning_rate": 8.924933976944474e-06,
"loss": 2.2866,
"step": 333
},
{
"epoch": 0.24273255813953487,
"grad_norm": 1.1530937805371297,
"learning_rate": 8.917584144880124e-06,
"loss": 2.3271,
"step": 334
},
{
"epoch": 0.24345930232558138,
"grad_norm": 1.1115334863414545,
"learning_rate": 8.910212322532317e-06,
"loss": 2.3109,
"step": 335
},
{
"epoch": 0.2441860465116279,
"grad_norm": 1.2194859856345706,
"learning_rate": 8.902818551280758e-06,
"loss": 2.3286,
"step": 336
},
{
"epoch": 0.2449127906976744,
"grad_norm": 1.1114840199496325,
"learning_rate": 8.895402872628352e-06,
"loss": 2.3368,
"step": 337
},
{
"epoch": 0.24563953488372092,
"grad_norm": 1.2770635204204543,
"learning_rate": 8.887965328200975e-06,
"loss": 2.3251,
"step": 338
},
{
"epoch": 0.24636627906976744,
"grad_norm": 3.188412638877289,
"learning_rate": 8.880505959747245e-06,
"loss": 2.3463,
"step": 339
},
{
"epoch": 0.24709302325581395,
"grad_norm": 5.785905050744672,
"learning_rate": 8.873024809138272e-06,
"loss": 2.3595,
"step": 340
},
{
"epoch": 0.24781976744186046,
"grad_norm": 1.5583689080056162,
"learning_rate": 8.86552191836745e-06,
"loss": 2.3503,
"step": 341
},
{
"epoch": 0.24854651162790697,
"grad_norm": 1.9351660592413844,
"learning_rate": 8.857997329550195e-06,
"loss": 2.3714,
"step": 342
},
{
"epoch": 0.2492732558139535,
"grad_norm": 1.3609293978964716,
"learning_rate": 8.850451084923717e-06,
"loss": 2.2474,
"step": 343
},
{
"epoch": 0.25,
"grad_norm": 1.203918915159538,
"learning_rate": 8.842883226846792e-06,
"loss": 2.3455,
"step": 344
},
{
"epoch": 0.2507267441860465,
"grad_norm": 1.230555039754799,
"learning_rate": 8.835293797799517e-06,
"loss": 2.2984,
"step": 345
},
{
"epoch": 0.251453488372093,
"grad_norm": 1.2486863122908132,
"learning_rate": 8.827682840383065e-06,
"loss": 2.271,
"step": 346
},
{
"epoch": 0.25218023255813954,
"grad_norm": 1.5306570439449183,
"learning_rate": 8.82005039731946e-06,
"loss": 2.3261,
"step": 347
},
{
"epoch": 0.25290697674418605,
"grad_norm": 1.245699604524567,
"learning_rate": 8.812396511451324e-06,
"loss": 2.3407,
"step": 348
},
{
"epoch": 0.25363372093023256,
"grad_norm": 1.3300197575111206,
"learning_rate": 8.804721225741646e-06,
"loss": 2.3137,
"step": 349
},
{
"epoch": 0.2543604651162791,
"grad_norm": 1.3585363752309658,
"learning_rate": 8.797024583273536e-06,
"loss": 2.3609,
"step": 350
},
{
"epoch": 0.2550872093023256,
"grad_norm": 1.33120843382374,
"learning_rate": 8.789306627249985e-06,
"loss": 2.3286,
"step": 351
},
{
"epoch": 0.2558139534883721,
"grad_norm": 1.2345078288815257,
"learning_rate": 8.781567400993617e-06,
"loss": 2.2758,
"step": 352
},
{
"epoch": 0.2565406976744186,
"grad_norm": 1.175709121820161,
"learning_rate": 8.77380694794646e-06,
"loss": 2.3674,
"step": 353
},
{
"epoch": 0.25726744186046513,
"grad_norm": 1.1079781294472077,
"learning_rate": 8.766025311669685e-06,
"loss": 2.3371,
"step": 354
},
{
"epoch": 0.25799418604651164,
"grad_norm": 1.17685761578481,
"learning_rate": 8.75822253584337e-06,
"loss": 2.3144,
"step": 355
},
{
"epoch": 0.25872093023255816,
"grad_norm": 1.1866727563492487,
"learning_rate": 8.75039866426626e-06,
"loss": 2.2811,
"step": 356
},
{
"epoch": 0.25944767441860467,
"grad_norm": 1.0868526193912587,
"learning_rate": 8.742553740855507e-06,
"loss": 2.309,
"step": 357
},
{
"epoch": 0.2601744186046512,
"grad_norm": 1.2915335811110245,
"learning_rate": 8.734687809646437e-06,
"loss": 2.3078,
"step": 358
},
{
"epoch": 0.2609011627906977,
"grad_norm": 1.1957644387801363,
"learning_rate": 8.726800914792296e-06,
"loss": 2.3283,
"step": 359
},
{
"epoch": 0.2616279069767442,
"grad_norm": 1.1600839842194066,
"learning_rate": 8.718893100564002e-06,
"loss": 2.3648,
"step": 360
},
{
"epoch": 0.2623546511627907,
"grad_norm": 1.1209572775215828,
"learning_rate": 8.710964411349902e-06,
"loss": 2.3809,
"step": 361
},
{
"epoch": 0.26308139534883723,
"grad_norm": 1.1794327292754048,
"learning_rate": 8.703014891655518e-06,
"loss": 2.3331,
"step": 362
},
{
"epoch": 0.26380813953488375,
"grad_norm": 1.0931283110899968,
"learning_rate": 8.695044586103297e-06,
"loss": 2.3286,
"step": 363
},
{
"epoch": 0.26453488372093026,
"grad_norm": 1.1015327533697825,
"learning_rate": 8.687053539432358e-06,
"loss": 2.3201,
"step": 364
},
{
"epoch": 0.26526162790697677,
"grad_norm": 1.1199381476087624,
"learning_rate": 8.679041796498253e-06,
"loss": 2.3103,
"step": 365
},
{
"epoch": 0.26598837209302323,
"grad_norm": 1.271016400767226,
"learning_rate": 8.6710094022727e-06,
"loss": 2.3066,
"step": 366
},
{
"epoch": 0.26671511627906974,
"grad_norm": 1.2726894817812915,
"learning_rate": 8.66295640184334e-06,
"loss": 2.3731,
"step": 367
},
{
"epoch": 0.26744186046511625,
"grad_norm": 1.2171370056189053,
"learning_rate": 8.65488284041348e-06,
"loss": 2.3446,
"step": 368
},
{
"epoch": 0.26816860465116277,
"grad_norm": 1.1315553912514624,
"learning_rate": 8.646788763301842e-06,
"loss": 2.3218,
"step": 369
},
{
"epoch": 0.2688953488372093,
"grad_norm": 1.0962417262456845,
"learning_rate": 8.638674215942307e-06,
"loss": 2.2953,
"step": 370
},
{
"epoch": 0.2696220930232558,
"grad_norm": 1.2587797644690477,
"learning_rate": 8.630539243883659e-06,
"loss": 2.3589,
"step": 371
},
{
"epoch": 0.2703488372093023,
"grad_norm": 1.1337580563036955,
"learning_rate": 8.62238389278933e-06,
"loss": 2.3077,
"step": 372
},
{
"epoch": 0.2710755813953488,
"grad_norm": 1.4161156234882568,
"learning_rate": 8.61420820843715e-06,
"loss": 2.2961,
"step": 373
},
{
"epoch": 0.27180232558139533,
"grad_norm": 1.0912056466351248,
"learning_rate": 8.606012236719073e-06,
"loss": 2.309,
"step": 374
},
{
"epoch": 0.27252906976744184,
"grad_norm": 1.2422820682329003,
"learning_rate": 8.59779602364094e-06,
"loss": 2.3833,
"step": 375
},
{
"epoch": 0.27325581395348836,
"grad_norm": 1.1217055100092126,
"learning_rate": 8.58955961532221e-06,
"loss": 2.3043,
"step": 376
},
{
"epoch": 0.27398255813953487,
"grad_norm": 1.2187038590832011,
"learning_rate": 8.581303057995697e-06,
"loss": 2.2918,
"step": 377
},
{
"epoch": 0.2747093023255814,
"grad_norm": 1.3062238711621994,
"learning_rate": 8.573026398007323e-06,
"loss": 2.3038,
"step": 378
},
{
"epoch": 0.2754360465116279,
"grad_norm": 1.2074262903566075,
"learning_rate": 8.564729681815846e-06,
"loss": 2.2956,
"step": 379
},
{
"epoch": 0.2761627906976744,
"grad_norm": 1.225859930553607,
"learning_rate": 8.556412955992604e-06,
"loss": 2.3516,
"step": 380
},
{
"epoch": 0.2768895348837209,
"grad_norm": 1.0827364010463423,
"learning_rate": 8.548076267221258e-06,
"loss": 2.2869,
"step": 381
},
{
"epoch": 0.27761627906976744,
"grad_norm": 1.0860309795926768,
"learning_rate": 8.539719662297519e-06,
"loss": 2.3179,
"step": 382
},
{
"epoch": 0.27834302325581395,
"grad_norm": 1.2009020460108055,
"learning_rate": 8.531343188128896e-06,
"loss": 2.3978,
"step": 383
},
{
"epoch": 0.27906976744186046,
"grad_norm": 1.0723553238260994,
"learning_rate": 8.52294689173443e-06,
"loss": 2.3311,
"step": 384
},
{
"epoch": 0.279796511627907,
"grad_norm": 1.2308339588445456,
"learning_rate": 8.514530820244427e-06,
"loss": 2.3072,
"step": 385
},
{
"epoch": 0.2805232558139535,
"grad_norm": 1.443673625883687,
"learning_rate": 8.506095020900192e-06,
"loss": 2.3471,
"step": 386
},
{
"epoch": 0.28125,
"grad_norm": 1.2323755395885254,
"learning_rate": 8.497639541053769e-06,
"loss": 2.3082,
"step": 387
},
{
"epoch": 0.2819767441860465,
"grad_norm": 1.0985843059579208,
"learning_rate": 8.489164428167677e-06,
"loss": 2.3291,
"step": 388
},
{
"epoch": 0.282703488372093,
"grad_norm": 1.2185119408676808,
"learning_rate": 8.480669729814635e-06,
"loss": 2.2915,
"step": 389
},
{
"epoch": 0.28343023255813954,
"grad_norm": 1.2547102455831876,
"learning_rate": 8.472155493677299e-06,
"loss": 2.3284,
"step": 390
},
{
"epoch": 0.28415697674418605,
"grad_norm": 1.241863166940908,
"learning_rate": 8.463621767547998e-06,
"loss": 2.2481,
"step": 391
},
{
"epoch": 0.28488372093023256,
"grad_norm": 1.0588850602261963,
"learning_rate": 8.455068599328462e-06,
"loss": 2.3331,
"step": 392
},
{
"epoch": 0.2856104651162791,
"grad_norm": 1.1212857689423208,
"learning_rate": 8.446496037029555e-06,
"loss": 2.2994,
"step": 393
},
{
"epoch": 0.2863372093023256,
"grad_norm": 1.126135345225652,
"learning_rate": 8.437904128770999e-06,
"loss": 2.3206,
"step": 394
},
{
"epoch": 0.2870639534883721,
"grad_norm": 1.298548137211412,
"learning_rate": 8.429292922781115e-06,
"loss": 2.2919,
"step": 395
},
{
"epoch": 0.2877906976744186,
"grad_norm": 1.1255687465878497,
"learning_rate": 8.420662467396548e-06,
"loss": 2.3118,
"step": 396
},
{
"epoch": 0.28851744186046513,
"grad_norm": 1.1625027040610958,
"learning_rate": 8.412012811061985e-06,
"loss": 2.3196,
"step": 397
},
{
"epoch": 0.28924418604651164,
"grad_norm": 1.178266615842495,
"learning_rate": 8.403344002329901e-06,
"loss": 2.2574,
"step": 398
},
{
"epoch": 0.28997093023255816,
"grad_norm": 1.6209181315673342,
"learning_rate": 8.394656089860274e-06,
"loss": 2.3217,
"step": 399
},
{
"epoch": 0.29069767441860467,
"grad_norm": 1.4055391277869425,
"learning_rate": 8.385949122420318e-06,
"loss": 2.3053,
"step": 400
},
{
"epoch": 0.2914244186046512,
"grad_norm": 1.0837894019376972,
"learning_rate": 8.377223148884202e-06,
"loss": 2.3235,
"step": 401
},
{
"epoch": 0.2921511627906977,
"grad_norm": 1.1567143322087055,
"learning_rate": 8.368478218232787e-06,
"loss": 2.2682,
"step": 402
},
{
"epoch": 0.2928779069767442,
"grad_norm": 1.2833964025525515,
"learning_rate": 8.359714379553338e-06,
"loss": 2.3368,
"step": 403
},
{
"epoch": 0.2936046511627907,
"grad_norm": 1.1703928335892075,
"learning_rate": 8.350931682039262e-06,
"loss": 2.3692,
"step": 404
},
{
"epoch": 0.29433139534883723,
"grad_norm": 1.2278640949453006,
"learning_rate": 8.342130174989819e-06,
"loss": 2.3298,
"step": 405
},
{
"epoch": 0.29505813953488375,
"grad_norm": 1.2391820171446353,
"learning_rate": 8.333309907809852e-06,
"loss": 2.3221,
"step": 406
},
{
"epoch": 0.29578488372093026,
"grad_norm": 1.1591560003156445,
"learning_rate": 8.324470930009514e-06,
"loss": 2.2989,
"step": 407
},
{
"epoch": 0.29651162790697677,
"grad_norm": 1.069079000918171,
"learning_rate": 8.315613291203977e-06,
"loss": 2.3257,
"step": 408
},
{
"epoch": 0.29723837209302323,
"grad_norm": 1.0658003352432701,
"learning_rate": 8.306737041113169e-06,
"loss": 2.2969,
"step": 409
},
{
"epoch": 0.29796511627906974,
"grad_norm": 1.2018393379444972,
"learning_rate": 8.29784222956148e-06,
"loss": 2.2994,
"step": 410
},
{
"epoch": 0.29869186046511625,
"grad_norm": 1.0743593652723726,
"learning_rate": 8.288928906477497e-06,
"loss": 2.3918,
"step": 411
},
{
"epoch": 0.29941860465116277,
"grad_norm": 1.1791792326436397,
"learning_rate": 8.279997121893713e-06,
"loss": 2.2971,
"step": 412
},
{
"epoch": 0.3001453488372093,
"grad_norm": 1.068770185795778,
"learning_rate": 8.271046925946247e-06,
"loss": 2.3359,
"step": 413
},
{
"epoch": 0.3008720930232558,
"grad_norm": 1.3026079186055184,
"learning_rate": 8.262078368874566e-06,
"loss": 2.3961,
"step": 414
},
{
"epoch": 0.3015988372093023,
"grad_norm": 1.0548359339086584,
"learning_rate": 8.25309150102121e-06,
"loss": 2.2817,
"step": 415
},
{
"epoch": 0.3023255813953488,
"grad_norm": 1.2757882978370192,
"learning_rate": 8.244086372831492e-06,
"loss": 2.2907,
"step": 416
},
{
"epoch": 0.30305232558139533,
"grad_norm": 1.1608993852105884,
"learning_rate": 8.235063034853228e-06,
"loss": 2.3469,
"step": 417
},
{
"epoch": 0.30377906976744184,
"grad_norm": 1.6704712398247437,
"learning_rate": 8.226021537736449e-06,
"loss": 2.2899,
"step": 418
},
{
"epoch": 0.30450581395348836,
"grad_norm": 1.1076846726964071,
"learning_rate": 8.216961932233118e-06,
"loss": 2.2449,
"step": 419
},
{
"epoch": 0.30523255813953487,
"grad_norm": 1.2655876969977151,
"learning_rate": 8.207884269196845e-06,
"loss": 2.2987,
"step": 420
},
{
"epoch": 0.3059593023255814,
"grad_norm": 1.1019813151108213,
"learning_rate": 8.198788599582596e-06,
"loss": 2.3452,
"step": 421
},
{
"epoch": 0.3066860465116279,
"grad_norm": 1.781405867357273,
"learning_rate": 8.189674974446423e-06,
"loss": 2.308,
"step": 422
},
{
"epoch": 0.3074127906976744,
"grad_norm": 1.2350247568154873,
"learning_rate": 8.180543444945154e-06,
"loss": 2.3308,
"step": 423
},
{
"epoch": 0.3081395348837209,
"grad_norm": 1.9649756348954517,
"learning_rate": 8.171394062336127e-06,
"loss": 2.3159,
"step": 424
},
{
"epoch": 0.30886627906976744,
"grad_norm": 1.2739192151431729,
"learning_rate": 8.162226877976886e-06,
"loss": 2.2989,
"step": 425
},
{
"epoch": 0.30959302325581395,
"grad_norm": 1.129631923987475,
"learning_rate": 8.153041943324912e-06,
"loss": 2.3198,
"step": 426
},
{
"epoch": 0.31031976744186046,
"grad_norm": 1.3255063702568364,
"learning_rate": 8.143839309937307e-06,
"loss": 2.3381,
"step": 427
},
{
"epoch": 0.311046511627907,
"grad_norm": 1.1628422209438,
"learning_rate": 8.134619029470535e-06,
"loss": 2.3287,
"step": 428
},
{
"epoch": 0.3117732558139535,
"grad_norm": 1.1887971582610084,
"learning_rate": 8.125381153680103e-06,
"loss": 2.3326,
"step": 429
},
{
"epoch": 0.3125,
"grad_norm": 1.1714666094613004,
"learning_rate": 8.116125734420297e-06,
"loss": 2.3104,
"step": 430
},
{
"epoch": 0.3132267441860465,
"grad_norm": 1.1365583272218085,
"learning_rate": 8.10685282364387e-06,
"loss": 2.284,
"step": 431
},
{
"epoch": 0.313953488372093,
"grad_norm": 1.3559133401669676,
"learning_rate": 8.097562473401764e-06,
"loss": 2.3029,
"step": 432
},
{
"epoch": 0.31468023255813954,
"grad_norm": 1.4060183717002648,
"learning_rate": 8.088254735842808e-06,
"loss": 2.2876,
"step": 433
},
{
"epoch": 0.31540697674418605,
"grad_norm": 1.285917870149192,
"learning_rate": 8.078929663213432e-06,
"loss": 2.3057,
"step": 434
},
{
"epoch": 0.31613372093023256,
"grad_norm": 1.1249535937437272,
"learning_rate": 8.069587307857377e-06,
"loss": 2.3215,
"step": 435
},
{
"epoch": 0.3168604651162791,
"grad_norm": 1.0900905006814698,
"learning_rate": 8.060227722215385e-06,
"loss": 2.2907,
"step": 436
},
{
"epoch": 0.3175872093023256,
"grad_norm": 1.1568687109085039,
"learning_rate": 8.050850958824926e-06,
"loss": 2.3176,
"step": 437
},
{
"epoch": 0.3183139534883721,
"grad_norm": 1.2382798087134244,
"learning_rate": 8.041457070319884e-06,
"loss": 2.3037,
"step": 438
},
{
"epoch": 0.3190406976744186,
"grad_norm": 1.1937622328697128,
"learning_rate": 8.032046109430276e-06,
"loss": 2.3076,
"step": 439
},
{
"epoch": 0.31976744186046513,
"grad_norm": 1.1177244010663805,
"learning_rate": 8.02261812898195e-06,
"loss": 2.3324,
"step": 440
},
{
"epoch": 0.32049418604651164,
"grad_norm": 1.127296551947173,
"learning_rate": 8.013173181896283e-06,
"loss": 2.3078,
"step": 441
},
{
"epoch": 0.32122093023255816,
"grad_norm": 1.1455250336877478,
"learning_rate": 8.003711321189895e-06,
"loss": 2.3523,
"step": 442
},
{
"epoch": 0.32194767441860467,
"grad_norm": 1.1011738742836867,
"learning_rate": 7.994232599974346e-06,
"loss": 2.3044,
"step": 443
},
{
"epoch": 0.3226744186046512,
"grad_norm": 1.1216213823441015,
"learning_rate": 7.984737071455834e-06,
"loss": 2.2963,
"step": 444
},
{
"epoch": 0.3234011627906977,
"grad_norm": 1.2200505156378094,
"learning_rate": 7.975224788934903e-06,
"loss": 2.3331,
"step": 445
},
{
"epoch": 0.3241279069767442,
"grad_norm": 1.1574339311156163,
"learning_rate": 7.965695805806141e-06,
"loss": 2.2945,
"step": 446
},
{
"epoch": 0.3248546511627907,
"grad_norm": 1.1286977461580614,
"learning_rate": 7.95615017555788e-06,
"loss": 2.308,
"step": 447
},
{
"epoch": 0.32558139534883723,
"grad_norm": 1.0892889308004792,
"learning_rate": 7.946587951771894e-06,
"loss": 2.3267,
"step": 448
},
{
"epoch": 0.32630813953488375,
"grad_norm": 1.1576745403605158,
"learning_rate": 7.937009188123102e-06,
"loss": 2.3197,
"step": 449
},
{
"epoch": 0.32703488372093026,
"grad_norm": 1.4094383886534938,
"learning_rate": 7.927413938379268e-06,
"loss": 2.3073,
"step": 450
},
{
"epoch": 0.32776162790697677,
"grad_norm": 1.1812098015475305,
"learning_rate": 7.917802256400688e-06,
"loss": 2.3307,
"step": 451
},
{
"epoch": 0.32848837209302323,
"grad_norm": 1.0953213999810891,
"learning_rate": 7.908174196139907e-06,
"loss": 2.2721,
"step": 452
},
{
"epoch": 0.32921511627906974,
"grad_norm": 1.148401707752194,
"learning_rate": 7.898529811641393e-06,
"loss": 2.3272,
"step": 453
},
{
"epoch": 0.32994186046511625,
"grad_norm": 1.0408799055573608,
"learning_rate": 7.888869157041257e-06,
"loss": 2.3114,
"step": 454
},
{
"epoch": 0.33066860465116277,
"grad_norm": 1.1001301089410505,
"learning_rate": 7.879192286566929e-06,
"loss": 2.3111,
"step": 455
},
{
"epoch": 0.3313953488372093,
"grad_norm": 1.2045234323369642,
"learning_rate": 7.869499254536865e-06,
"loss": 2.3341,
"step": 456
},
{
"epoch": 0.3321220930232558,
"grad_norm": 1.1544365222104789,
"learning_rate": 7.859790115360243e-06,
"loss": 2.3371,
"step": 457
},
{
"epoch": 0.3328488372093023,
"grad_norm": 1.2552624406434518,
"learning_rate": 7.850064923536649e-06,
"loss": 2.3002,
"step": 458
},
{
"epoch": 0.3335755813953488,
"grad_norm": 1.293746161658906,
"learning_rate": 7.84032373365578e-06,
"loss": 2.3692,
"step": 459
},
{
"epoch": 0.33430232558139533,
"grad_norm": 1.333213925468466,
"learning_rate": 7.83056660039713e-06,
"loss": 2.2969,
"step": 460
},
{
"epoch": 0.33502906976744184,
"grad_norm": 1.492618082058401,
"learning_rate": 7.82079357852969e-06,
"loss": 2.2577,
"step": 461
},
{
"epoch": 0.33575581395348836,
"grad_norm": 1.15249654866146,
"learning_rate": 7.811004722911637e-06,
"loss": 2.3238,
"step": 462
},
{
"epoch": 0.33648255813953487,
"grad_norm": 1.2530632952138137,
"learning_rate": 7.801200088490026e-06,
"loss": 2.301,
"step": 463
},
{
"epoch": 0.3372093023255814,
"grad_norm": 1.2126256906203419,
"learning_rate": 7.791379730300476e-06,
"loss": 2.252,
"step": 464
},
{
"epoch": 0.3379360465116279,
"grad_norm": 1.1402545118499574,
"learning_rate": 7.781543703466881e-06,
"loss": 2.3447,
"step": 465
},
{
"epoch": 0.3386627906976744,
"grad_norm": 1.079289917582004,
"learning_rate": 7.771692063201072e-06,
"loss": 2.307,
"step": 466
},
{
"epoch": 0.3393895348837209,
"grad_norm": 1.0890060946616744,
"learning_rate": 7.76182486480253e-06,
"loss": 2.2897,
"step": 467
},
{
"epoch": 0.34011627906976744,
"grad_norm": 1.2191356307676027,
"learning_rate": 7.751942163658066e-06,
"loss": 2.2806,
"step": 468
},
{
"epoch": 0.34084302325581395,
"grad_norm": 1.1684818562278108,
"learning_rate": 7.742044015241508e-06,
"loss": 2.309,
"step": 469
},
{
"epoch": 0.34156976744186046,
"grad_norm": 1.1378897839340987,
"learning_rate": 7.7321304751134e-06,
"loss": 2.3417,
"step": 470
},
{
"epoch": 0.342296511627907,
"grad_norm": 1.0831185805529036,
"learning_rate": 7.722201598920673e-06,
"loss": 2.2842,
"step": 471
},
{
"epoch": 0.3430232558139535,
"grad_norm": 1.163651975848296,
"learning_rate": 7.712257442396355e-06,
"loss": 2.3188,
"step": 472
},
{
"epoch": 0.34375,
"grad_norm": 1.1173941040067659,
"learning_rate": 7.702298061359236e-06,
"loss": 2.29,
"step": 473
},
{
"epoch": 0.3444767441860465,
"grad_norm": 1.1553929055835899,
"learning_rate": 7.692323511713568e-06,
"loss": 2.3456,
"step": 474
},
{
"epoch": 0.345203488372093,
"grad_norm": 1.137786497705492,
"learning_rate": 7.682333849448749e-06,
"loss": 2.3446,
"step": 475
},
{
"epoch": 0.34593023255813954,
"grad_norm": 1.2480217073388282,
"learning_rate": 7.672329130639007e-06,
"loss": 2.3192,
"step": 476
},
{
"epoch": 0.34665697674418605,
"grad_norm": 1.1303930753175966,
"learning_rate": 7.662309411443084e-06,
"loss": 2.3791,
"step": 477
},
{
"epoch": 0.34738372093023256,
"grad_norm": 1.1988474595235474,
"learning_rate": 7.652274748103924e-06,
"loss": 2.3186,
"step": 478
},
{
"epoch": 0.3481104651162791,
"grad_norm": 1.1508981723474565,
"learning_rate": 7.642225196948357e-06,
"loss": 2.315,
"step": 479
},
{
"epoch": 0.3488372093023256,
"grad_norm": 1.138764144490926,
"learning_rate": 7.63216081438678e-06,
"loss": 2.2812,
"step": 480
},
{
"epoch": 0.3495639534883721,
"grad_norm": 1.2032052124329502,
"learning_rate": 7.622081656912842e-06,
"loss": 2.3625,
"step": 481
},
{
"epoch": 0.3502906976744186,
"grad_norm": 1.3016605768369958,
"learning_rate": 7.611987781103128e-06,
"loss": 2.3206,
"step": 482
},
{
"epoch": 0.35101744186046513,
"grad_norm": 1.0688947854232214,
"learning_rate": 7.601879243616838e-06,
"loss": 2.3483,
"step": 483
},
{
"epoch": 0.35174418604651164,
"grad_norm": 1.0249007592193522,
"learning_rate": 7.5917561011954755e-06,
"loss": 2.3421,
"step": 484
},
{
"epoch": 0.35247093023255816,
"grad_norm": 1.1709699355757717,
"learning_rate": 7.581618410662519e-06,
"loss": 2.3722,
"step": 485
},
{
"epoch": 0.35319767441860467,
"grad_norm": 1.0586329236128114,
"learning_rate": 7.571466228923115e-06,
"loss": 2.3478,
"step": 486
},
{
"epoch": 0.3539244186046512,
"grad_norm": 1.0813414465551134,
"learning_rate": 7.56129961296375e-06,
"loss": 2.3209,
"step": 487
},
{
"epoch": 0.3546511627906977,
"grad_norm": 1.0444974051946252,
"learning_rate": 7.551118619851929e-06,
"loss": 2.3114,
"step": 488
},
{
"epoch": 0.3553779069767442,
"grad_norm": 1.5815223503399907,
"learning_rate": 7.540923306735868e-06,
"loss": 2.2959,
"step": 489
},
{
"epoch": 0.3561046511627907,
"grad_norm": 1.1602847453567253,
"learning_rate": 7.530713730844153e-06,
"loss": 2.3397,
"step": 490
},
{
"epoch": 0.35683139534883723,
"grad_norm": 1.049120747351293,
"learning_rate": 7.5204899494854415e-06,
"loss": 2.2834,
"step": 491
},
{
"epoch": 0.35755813953488375,
"grad_norm": 1.2262144884196204,
"learning_rate": 7.510252020048121e-06,
"loss": 2.287,
"step": 492
},
{
"epoch": 0.35828488372093026,
"grad_norm": 1.2042548629565901,
"learning_rate": 7.500000000000001e-06,
"loss": 2.2654,
"step": 493
},
{
"epoch": 0.35901162790697677,
"grad_norm": 1.2530924104028403,
"learning_rate": 7.489733946887982e-06,
"loss": 2.3486,
"step": 494
},
{
"epoch": 0.35973837209302323,
"grad_norm": 1.2825193397772006,
"learning_rate": 7.479453918337733e-06,
"loss": 2.275,
"step": 495
},
{
"epoch": 0.36046511627906974,
"grad_norm": 1.3194033390036468,
"learning_rate": 7.469159972053377e-06,
"loss": 2.2908,
"step": 496
},
{
"epoch": 0.36119186046511625,
"grad_norm": 1.1152767805402615,
"learning_rate": 7.458852165817153e-06,
"loss": 2.3095,
"step": 497
},
{
"epoch": 0.36191860465116277,
"grad_norm": 1.1495601735419694,
"learning_rate": 7.448530557489105e-06,
"loss": 2.3252,
"step": 498
},
{
"epoch": 0.3626453488372093,
"grad_norm": 1.0736588885000224,
"learning_rate": 7.438195205006749e-06,
"loss": 2.2954,
"step": 499
},
{
"epoch": 0.3633720930232558,
"grad_norm": 1.207665825472445,
"learning_rate": 7.427846166384747e-06,
"loss": 2.2967,
"step": 500
},
{
"epoch": 0.3640988372093023,
"grad_norm": 1.18745268887086,
"learning_rate": 7.417483499714589e-06,
"loss": 2.2981,
"step": 501
},
{
"epoch": 0.3648255813953488,
"grad_norm": 1.564274356915172,
"learning_rate": 7.40710726316426e-06,
"loss": 2.2957,
"step": 502
},
{
"epoch": 0.36555232558139533,
"grad_norm": 1.290273123363002,
"learning_rate": 7.396717514977916e-06,
"loss": 2.3357,
"step": 503
},
{
"epoch": 0.36627906976744184,
"grad_norm": 1.2047088792360614,
"learning_rate": 7.386314313475557e-06,
"loss": 2.3721,
"step": 504
},
{
"epoch": 0.36700581395348836,
"grad_norm": 1.3246114123808277,
"learning_rate": 7.3758977170527e-06,
"loss": 2.3209,
"step": 505
},
{
"epoch": 0.36773255813953487,
"grad_norm": 1.1370471368395154,
"learning_rate": 7.365467784180051e-06,
"loss": 2.2803,
"step": 506
},
{
"epoch": 0.3684593023255814,
"grad_norm": 1.0769050085427025,
"learning_rate": 7.355024573403174e-06,
"loss": 2.3312,
"step": 507
},
{
"epoch": 0.3691860465116279,
"grad_norm": 1.3157313090863894,
"learning_rate": 7.3445681433421675e-06,
"loss": 2.3267,
"step": 508
},
{
"epoch": 0.3699127906976744,
"grad_norm": 1.1464812479744277,
"learning_rate": 7.3340985526913335e-06,
"loss": 2.3112,
"step": 509
},
{
"epoch": 0.3706395348837209,
"grad_norm": 1.1613820873773826,
"learning_rate": 7.323615860218844e-06,
"loss": 2.3007,
"step": 510
},
{
"epoch": 0.37136627906976744,
"grad_norm": 1.2091702257927293,
"learning_rate": 7.313120124766417e-06,
"loss": 2.316,
"step": 511
},
{
"epoch": 0.37209302325581395,
"grad_norm": 1.3421544244903119,
"learning_rate": 7.30261140524898e-06,
"loss": 2.2708,
"step": 512
},
{
"epoch": 0.37281976744186046,
"grad_norm": 1.39687115623401,
"learning_rate": 7.292089760654352e-06,
"loss": 2.3492,
"step": 513
},
{
"epoch": 0.373546511627907,
"grad_norm": 2.8218882467926356,
"learning_rate": 7.281555250042893e-06,
"loss": 2.2847,
"step": 514
},
{
"epoch": 0.3742732558139535,
"grad_norm": 1.088850072307723,
"learning_rate": 7.271007932547188e-06,
"loss": 2.3115,
"step": 515
},
{
"epoch": 0.375,
"grad_norm": 1.120826175014608,
"learning_rate": 7.2604478673717095e-06,
"loss": 2.3288,
"step": 516
},
{
"epoch": 0.3757267441860465,
"grad_norm": 1.145722078384872,
"learning_rate": 7.249875113792485e-06,
"loss": 2.3426,
"step": 517
},
{
"epoch": 0.376453488372093,
"grad_norm": 1.1602292605455282,
"learning_rate": 7.239289731156767e-06,
"loss": 2.2989,
"step": 518
},
{
"epoch": 0.37718023255813954,
"grad_norm": 1.1915736998674706,
"learning_rate": 7.2286917788826926e-06,
"loss": 2.3374,
"step": 519
},
{
"epoch": 0.37790697674418605,
"grad_norm": 1.13844351172199,
"learning_rate": 7.218081316458959e-06,
"loss": 2.332,
"step": 520
},
{
"epoch": 0.37863372093023256,
"grad_norm": 1.0975887317687405,
"learning_rate": 7.207458403444488e-06,
"loss": 2.3168,
"step": 521
},
{
"epoch": 0.3793604651162791,
"grad_norm": 1.3495614353404823,
"learning_rate": 7.196823099468084e-06,
"loss": 2.2743,
"step": 522
},
{
"epoch": 0.3800872093023256,
"grad_norm": 1.1407396853075507,
"learning_rate": 7.186175464228109e-06,
"loss": 2.2748,
"step": 523
},
{
"epoch": 0.3808139534883721,
"grad_norm": 1.083025323169463,
"learning_rate": 7.175515557492139e-06,
"loss": 2.2926,
"step": 524
},
{
"epoch": 0.3815406976744186,
"grad_norm": 1.2080116038697124,
"learning_rate": 7.1648434390966356e-06,
"loss": 2.3196,
"step": 525
},
{
"epoch": 0.38226744186046513,
"grad_norm": 2.066872793237152,
"learning_rate": 7.154159168946607e-06,
"loss": 2.2865,
"step": 526
},
{
"epoch": 0.38299418604651164,
"grad_norm": 1.2368945968737097,
"learning_rate": 7.143462807015271e-06,
"loss": 2.2789,
"step": 527
},
{
"epoch": 0.38372093023255816,
"grad_norm": 1.1743588332349175,
"learning_rate": 7.132754413343721e-06,
"loss": 2.3053,
"step": 528
},
{
"epoch": 0.38444767441860467,
"grad_norm": 1.4255712436197852,
"learning_rate": 7.122034048040586e-06,
"loss": 2.308,
"step": 529
},
{
"epoch": 0.3851744186046512,
"grad_norm": 1.1049951766621189,
"learning_rate": 7.111301771281692e-06,
"loss": 2.3054,
"step": 530
},
{
"epoch": 0.3859011627906977,
"grad_norm": 1.1038689017404644,
"learning_rate": 7.100557643309732e-06,
"loss": 2.2797,
"step": 531
},
{
"epoch": 0.3866279069767442,
"grad_norm": 1.0536128185639355,
"learning_rate": 7.089801724433918e-06,
"loss": 2.3071,
"step": 532
},
{
"epoch": 0.3873546511627907,
"grad_norm": 1.1449743039342242,
"learning_rate": 7.079034075029651e-06,
"loss": 2.2961,
"step": 533
},
{
"epoch": 0.38808139534883723,
"grad_norm": 1.1655567692704296,
"learning_rate": 7.0682547555381734e-06,
"loss": 2.329,
"step": 534
},
{
"epoch": 0.38880813953488375,
"grad_norm": 1.162251468256884,
"learning_rate": 7.057463826466235e-06,
"loss": 2.2895,
"step": 535
},
{
"epoch": 0.38953488372093026,
"grad_norm": 1.0896289667493249,
"learning_rate": 7.0466613483857615e-06,
"loss": 2.3105,
"step": 536
},
{
"epoch": 0.39026162790697677,
"grad_norm": 1.2880741445087427,
"learning_rate": 7.035847381933494e-06,
"loss": 2.3064,
"step": 537
},
{
"epoch": 0.39098837209302323,
"grad_norm": 1.2105896256495732,
"learning_rate": 7.025021987810664e-06,
"loss": 2.3694,
"step": 538
},
{
"epoch": 0.39171511627906974,
"grad_norm": 1.3065723003897216,
"learning_rate": 7.014185226782655e-06,
"loss": 2.3639,
"step": 539
},
{
"epoch": 0.39244186046511625,
"grad_norm": 1.1070630816606901,
"learning_rate": 7.003337159678649e-06,
"loss": 2.3083,
"step": 540
},
{
"epoch": 0.39316860465116277,
"grad_norm": 1.307708843946163,
"learning_rate": 6.992477847391292e-06,
"loss": 2.3399,
"step": 541
},
{
"epoch": 0.3938953488372093,
"grad_norm": 1.122196521717046,
"learning_rate": 6.981607350876357e-06,
"loss": 2.3313,
"step": 542
},
{
"epoch": 0.3946220930232558,
"grad_norm": 1.2186143505802445,
"learning_rate": 6.970725731152389e-06,
"loss": 2.3405,
"step": 543
},
{
"epoch": 0.3953488372093023,
"grad_norm": 1.167077189138343,
"learning_rate": 6.959833049300376e-06,
"loss": 2.29,
"step": 544
},
{
"epoch": 0.3960755813953488,
"grad_norm": 1.2944725806931758,
"learning_rate": 6.948929366463397e-06,
"loss": 2.3098,
"step": 545
},
{
"epoch": 0.39680232558139533,
"grad_norm": 12.31808967596737,
"learning_rate": 6.938014743846285e-06,
"loss": 2.2809,
"step": 546
},
{
"epoch": 0.39752906976744184,
"grad_norm": 1.2870952481596472,
"learning_rate": 6.927089242715277e-06,
"loss": 2.3301,
"step": 547
},
{
"epoch": 0.39825581395348836,
"grad_norm": 1.1514862581452583,
"learning_rate": 6.916152924397676e-06,
"loss": 2.3583,
"step": 548
},
{
"epoch": 0.39898255813953487,
"grad_norm": 1.100288574688579,
"learning_rate": 6.905205850281502e-06,
"loss": 2.2923,
"step": 549
},
{
"epoch": 0.3997093023255814,
"grad_norm": 1.050534628812702,
"learning_rate": 6.894248081815155e-06,
"loss": 2.3055,
"step": 550
},
{
"epoch": 0.4004360465116279,
"grad_norm": 1.3346961931360157,
"learning_rate": 6.883279680507057e-06,
"loss": 2.3039,
"step": 551
},
{
"epoch": 0.4011627906976744,
"grad_norm": 1.505626927488441,
"learning_rate": 6.872300707925319e-06,
"loss": 2.285,
"step": 552
},
{
"epoch": 0.4018895348837209,
"grad_norm": 1.1299060186104746,
"learning_rate": 6.861311225697392e-06,
"loss": 2.2638,
"step": 553
},
{
"epoch": 0.40261627906976744,
"grad_norm": 1.1060082205232702,
"learning_rate": 6.850311295509719e-06,
"loss": 2.3511,
"step": 554
},
{
"epoch": 0.40334302325581395,
"grad_norm": 1.1200104346671893,
"learning_rate": 6.8393009791073895e-06,
"loss": 2.329,
"step": 555
},
{
"epoch": 0.40406976744186046,
"grad_norm": 1.1476307981667215,
"learning_rate": 6.828280338293792e-06,
"loss": 2.3368,
"step": 556
},
{
"epoch": 0.404796511627907,
"grad_norm": 1.028644332291077,
"learning_rate": 6.817249434930267e-06,
"loss": 2.3036,
"step": 557
},
{
"epoch": 0.4055232558139535,
"grad_norm": 1.1390289832994775,
"learning_rate": 6.806208330935766e-06,
"loss": 2.3334,
"step": 558
},
{
"epoch": 0.40625,
"grad_norm": 1.216277772235329,
"learning_rate": 6.7951570882864944e-06,
"loss": 2.2741,
"step": 559
},
{
"epoch": 0.4069767441860465,
"grad_norm": 1.344668900236945,
"learning_rate": 6.784095769015573e-06,
"loss": 2.3205,
"step": 560
},
{
"epoch": 0.407703488372093,
"grad_norm": 1.070186991237952,
"learning_rate": 6.773024435212678e-06,
"loss": 2.307,
"step": 561
},
{
"epoch": 0.40843023255813954,
"grad_norm": 1.209196957642039,
"learning_rate": 6.761943149023706e-06,
"loss": 2.3112,
"step": 562
},
{
"epoch": 0.40915697674418605,
"grad_norm": 1.1432958784507203,
"learning_rate": 6.750851972650416e-06,
"loss": 2.2938,
"step": 563
},
{
"epoch": 0.40988372093023256,
"grad_norm": 1.2247587749897832,
"learning_rate": 6.739750968350081e-06,
"loss": 2.2909,
"step": 564
},
{
"epoch": 0.4106104651162791,
"grad_norm": 1.0268929904759363,
"learning_rate": 6.728640198435143e-06,
"loss": 2.2189,
"step": 565
},
{
"epoch": 0.4113372093023256,
"grad_norm": 1.6530106180630812,
"learning_rate": 6.717519725272859e-06,
"loss": 2.3005,
"step": 566
},
{
"epoch": 0.4120639534883721,
"grad_norm": 1.2858847661272295,
"learning_rate": 6.706389611284953e-06,
"loss": 2.3367,
"step": 567
},
{
"epoch": 0.4127906976744186,
"grad_norm": 1.0808913275520438,
"learning_rate": 6.6952499189472665e-06,
"loss": 2.2644,
"step": 568
},
{
"epoch": 0.41351744186046513,
"grad_norm": 1.0947135769666254,
"learning_rate": 6.684100710789405e-06,
"loss": 2.3109,
"step": 569
},
{
"epoch": 0.41424418604651164,
"grad_norm": 1.227905292543617,
"learning_rate": 6.6729420493943875e-06,
"loss": 2.3089,
"step": 570
},
{
"epoch": 0.41497093023255816,
"grad_norm": 1.1311925494343438,
"learning_rate": 6.6617739973982985e-06,
"loss": 2.3306,
"step": 571
},
{
"epoch": 0.41569767441860467,
"grad_norm": 1.1845320813818851,
"learning_rate": 6.6505966174899326e-06,
"loss": 2.2807,
"step": 572
},
{
"epoch": 0.4164244186046512,
"grad_norm": 1.192027080183658,
"learning_rate": 6.639409972410446e-06,
"loss": 2.2992,
"step": 573
},
{
"epoch": 0.4171511627906977,
"grad_norm": 1.1134220611408692,
"learning_rate": 6.628214124952999e-06,
"loss": 2.3198,
"step": 574
},
{
"epoch": 0.4178779069767442,
"grad_norm": 1.2137521582577167,
"learning_rate": 6.617009137962407e-06,
"loss": 2.2729,
"step": 575
},
{
"epoch": 0.4186046511627907,
"grad_norm": 1.171459631389888,
"learning_rate": 6.605795074334793e-06,
"loss": 2.2871,
"step": 576
},
{
"epoch": 0.41933139534883723,
"grad_norm": 1.1838961162503008,
"learning_rate": 6.594571997017224e-06,
"loss": 2.3488,
"step": 577
},
{
"epoch": 0.42005813953488375,
"grad_norm": 1.1791646717778388,
"learning_rate": 6.583339969007364e-06,
"loss": 2.3505,
"step": 578
},
{
"epoch": 0.42078488372093026,
"grad_norm": 1.1787043516107347,
"learning_rate": 6.57209905335312e-06,
"loss": 2.3469,
"step": 579
},
{
"epoch": 0.42151162790697677,
"grad_norm": 1.1000497157207225,
"learning_rate": 6.560849313152287e-06,
"loss": 2.2874,
"step": 580
},
{
"epoch": 0.42223837209302323,
"grad_norm": 1.145236146723751,
"learning_rate": 6.549590811552193e-06,
"loss": 2.2612,
"step": 581
},
{
"epoch": 0.42296511627906974,
"grad_norm": 1.2072587593279676,
"learning_rate": 6.538323611749351e-06,
"loss": 2.3195,
"step": 582
},
{
"epoch": 0.42369186046511625,
"grad_norm": 1.1421104121114019,
"learning_rate": 6.5270477769890906e-06,
"loss": 2.2864,
"step": 583
},
{
"epoch": 0.42441860465116277,
"grad_norm": 1.1733190168229655,
"learning_rate": 6.515763370565218e-06,
"loss": 2.3614,
"step": 584
},
{
"epoch": 0.4251453488372093,
"grad_norm": 1.101416582117341,
"learning_rate": 6.504470455819651e-06,
"loss": 2.3193,
"step": 585
},
{
"epoch": 0.4258720930232558,
"grad_norm": 1.1869346197022288,
"learning_rate": 6.493169096142068e-06,
"loss": 2.2734,
"step": 586
},
{
"epoch": 0.4265988372093023,
"grad_norm": 1.0719625987052026,
"learning_rate": 6.481859354969549e-06,
"loss": 2.2967,
"step": 587
},
{
"epoch": 0.4273255813953488,
"grad_norm": 1.2740863995681964,
"learning_rate": 6.470541295786222e-06,
"loss": 2.2977,
"step": 588
},
{
"epoch": 0.42805232558139533,
"grad_norm": 2.0881615265610054,
"learning_rate": 6.4592149821229064e-06,
"loss": 2.2487,
"step": 589
},
{
"epoch": 0.42877906976744184,
"grad_norm": 1.0918234165843568,
"learning_rate": 6.447880477556757e-06,
"loss": 2.2752,
"step": 590
},
{
"epoch": 0.42950581395348836,
"grad_norm": 1.0976921856780906,
"learning_rate": 6.436537845710904e-06,
"loss": 2.2932,
"step": 591
},
{
"epoch": 0.43023255813953487,
"grad_norm": 1.1941175241736834,
"learning_rate": 6.425187150254097e-06,
"loss": 2.3102,
"step": 592
},
{
"epoch": 0.4309593023255814,
"grad_norm": 1.0851777276937011,
"learning_rate": 6.413828454900351e-06,
"loss": 2.2651,
"step": 593
},
{
"epoch": 0.4316860465116279,
"grad_norm": 1.111314328918144,
"learning_rate": 6.402461823408584e-06,
"loss": 2.3142,
"step": 594
},
{
"epoch": 0.4324127906976744,
"grad_norm": 1.1107472084593226,
"learning_rate": 6.391087319582264e-06,
"loss": 2.3056,
"step": 595
},
{
"epoch": 0.4331395348837209,
"grad_norm": 1.110937134108463,
"learning_rate": 6.379705007269046e-06,
"loss": 2.3458,
"step": 596
},
{
"epoch": 0.43386627906976744,
"grad_norm": 1.1301775944171741,
"learning_rate": 6.368314950360416e-06,
"loss": 2.3003,
"step": 597
},
{
"epoch": 0.43459302325581395,
"grad_norm": 1.0782893388635726,
"learning_rate": 6.356917212791332e-06,
"loss": 2.3394,
"step": 598
},
{
"epoch": 0.43531976744186046,
"grad_norm": 1.1952070508071906,
"learning_rate": 6.3455118585398676e-06,
"loss": 2.3156,
"step": 599
},
{
"epoch": 0.436046511627907,
"grad_norm": 1.1855496510040404,
"learning_rate": 6.334098951626847e-06,
"loss": 2.3322,
"step": 600
},
{
"epoch": 0.4367732558139535,
"grad_norm": 1.1611150891834845,
"learning_rate": 6.3226785561154914e-06,
"loss": 2.3106,
"step": 601
},
{
"epoch": 0.4375,
"grad_norm": 1.129860856742074,
"learning_rate": 6.311250736111058e-06,
"loss": 2.2959,
"step": 602
},
{
"epoch": 0.4382267441860465,
"grad_norm": 1.1621204046893316,
"learning_rate": 6.299815555760478e-06,
"loss": 2.2952,
"step": 603
},
{
"epoch": 0.438953488372093,
"grad_norm": 1.0973622926564048,
"learning_rate": 6.288373079251996e-06,
"loss": 2.3415,
"step": 604
},
{
"epoch": 0.43968023255813954,
"grad_norm": 1.0808189653539233,
"learning_rate": 6.276923370814815e-06,
"loss": 2.2928,
"step": 605
},
{
"epoch": 0.44040697674418605,
"grad_norm": 1.1346845786046464,
"learning_rate": 6.265466494718731e-06,
"loss": 2.2576,
"step": 606
},
{
"epoch": 0.44113372093023256,
"grad_norm": 1.0957369183695937,
"learning_rate": 6.254002515273775e-06,
"loss": 2.2973,
"step": 607
},
{
"epoch": 0.4418604651162791,
"grad_norm": 1.1034874026636983,
"learning_rate": 6.242531496829848e-06,
"loss": 2.2827,
"step": 608
},
{
"epoch": 0.4425872093023256,
"grad_norm": 1.1697709003724293,
"learning_rate": 6.231053503776363e-06,
"loss": 2.3238,
"step": 609
},
{
"epoch": 0.4433139534883721,
"grad_norm": 1.040137222032865,
"learning_rate": 6.219568600541886e-06,
"loss": 2.2895,
"step": 610
},
{
"epoch": 0.4440406976744186,
"grad_norm": 1.3641221427787078,
"learning_rate": 6.208076851593768e-06,
"loss": 2.3187,
"step": 611
},
{
"epoch": 0.44476744186046513,
"grad_norm": 1.1013994361963013,
"learning_rate": 6.1965783214377895e-06,
"loss": 2.2899,
"step": 612
},
{
"epoch": 0.44549418604651164,
"grad_norm": 1.3087669498751793,
"learning_rate": 6.185073074617793e-06,
"loss": 2.2882,
"step": 613
},
{
"epoch": 0.44622093023255816,
"grad_norm": 1.1024012017074452,
"learning_rate": 6.173561175715323e-06,
"loss": 2.2886,
"step": 614
},
{
"epoch": 0.44694767441860467,
"grad_norm": 1.198664497940916,
"learning_rate": 6.1620426893492645e-06,
"loss": 2.3531,
"step": 615
},
{
"epoch": 0.4476744186046512,
"grad_norm": 1.1648828913400078,
"learning_rate": 6.150517680175482e-06,
"loss": 2.3145,
"step": 616
},
{
"epoch": 0.4484011627906977,
"grad_norm": 1.1859586096771217,
"learning_rate": 6.13898621288645e-06,
"loss": 2.2928,
"step": 617
},
{
"epoch": 0.4491279069767442,
"grad_norm": 1.074740378189526,
"learning_rate": 6.127448352210894e-06,
"loss": 2.3453,
"step": 618
},
{
"epoch": 0.4498546511627907,
"grad_norm": 1.1456985539298716,
"learning_rate": 6.115904162913431e-06,
"loss": 2.2879,
"step": 619
},
{
"epoch": 0.45058139534883723,
"grad_norm": 1.2351709659565,
"learning_rate": 6.1043537097941985e-06,
"loss": 2.2922,
"step": 620
},
{
"epoch": 0.45130813953488375,
"grad_norm": 1.1500337120485173,
"learning_rate": 6.092797057688496e-06,
"loss": 2.2747,
"step": 621
},
{
"epoch": 0.45203488372093026,
"grad_norm": 1.289050278474326,
"learning_rate": 6.081234271466416e-06,
"loss": 2.345,
"step": 622
},
{
"epoch": 0.45276162790697677,
"grad_norm": 1.279731812260382,
"learning_rate": 6.0696654160324875e-06,
"loss": 2.3395,
"step": 623
},
{
"epoch": 0.45348837209302323,
"grad_norm": 2.416509224462852,
"learning_rate": 6.058090556325305e-06,
"loss": 2.2773,
"step": 624
},
{
"epoch": 0.45421511627906974,
"grad_norm": 1.3184450912253118,
"learning_rate": 6.046509757317168e-06,
"loss": 2.2991,
"step": 625
},
{
"epoch": 0.45494186046511625,
"grad_norm": 1.20112258515423,
"learning_rate": 6.034923084013713e-06,
"loss": 2.2907,
"step": 626
},
{
"epoch": 0.45566860465116277,
"grad_norm": 1.2680453831355194,
"learning_rate": 6.0233306014535505e-06,
"loss": 2.3344,
"step": 627
},
{
"epoch": 0.4563953488372093,
"grad_norm": 1.077913847116916,
"learning_rate": 6.0117323747079e-06,
"loss": 2.2932,
"step": 628
},
{
"epoch": 0.4571220930232558,
"grad_norm": 1.2295931224219867,
"learning_rate": 6.000128468880223e-06,
"loss": 2.383,
"step": 629
},
{
"epoch": 0.4578488372093023,
"grad_norm": 1.1337008770515467,
"learning_rate": 5.988518949105862e-06,
"loss": 2.2843,
"step": 630
},
{
"epoch": 0.4585755813953488,
"grad_norm": 1.1357848209286197,
"learning_rate": 5.976903880551669e-06,
"loss": 2.2913,
"step": 631
},
{
"epoch": 0.45930232558139533,
"grad_norm": 1.164197176327958,
"learning_rate": 5.965283328415644e-06,
"loss": 2.3219,
"step": 632
},
{
"epoch": 0.46002906976744184,
"grad_norm": 1.261751325053521,
"learning_rate": 5.953657357926569e-06,
"loss": 2.2947,
"step": 633
},
{
"epoch": 0.46075581395348836,
"grad_norm": 1.4417326298084916,
"learning_rate": 5.942026034343636e-06,
"loss": 2.3586,
"step": 634
},
{
"epoch": 0.46148255813953487,
"grad_norm": 1.0410453025295299,
"learning_rate": 5.930389422956088e-06,
"loss": 2.2798,
"step": 635
},
{
"epoch": 0.4622093023255814,
"grad_norm": 3.232269670369028,
"learning_rate": 5.918747589082853e-06,
"loss": 2.3494,
"step": 636
},
{
"epoch": 0.4629360465116279,
"grad_norm": 1.3842857905764918,
"learning_rate": 5.907100598072166e-06,
"loss": 2.3548,
"step": 637
},
{
"epoch": 0.4636627906976744,
"grad_norm": 1.1648587635712846,
"learning_rate": 5.895448515301218e-06,
"loss": 2.3716,
"step": 638
},
{
"epoch": 0.4643895348837209,
"grad_norm": 1.284402759492195,
"learning_rate": 5.883791406175775e-06,
"loss": 2.2782,
"step": 639
},
{
"epoch": 0.46511627906976744,
"grad_norm": 1.4101311890951338,
"learning_rate": 5.872129336129821e-06,
"loss": 2.2666,
"step": 640
},
{
"epoch": 0.46584302325581395,
"grad_norm": 1.3443735682842992,
"learning_rate": 5.860462370625189e-06,
"loss": 2.2984,
"step": 641
},
{
"epoch": 0.46656976744186046,
"grad_norm": 1.3212822331398921,
"learning_rate": 5.848790575151181e-06,
"loss": 2.2868,
"step": 642
},
{
"epoch": 0.467296511627907,
"grad_norm": 1.0815673694776267,
"learning_rate": 5.837114015224223e-06,
"loss": 2.2562,
"step": 643
},
{
"epoch": 0.4680232558139535,
"grad_norm": 1.2786312590580617,
"learning_rate": 5.8254327563874794e-06,
"loss": 2.288,
"step": 644
},
{
"epoch": 0.46875,
"grad_norm": 1.055564931514476,
"learning_rate": 5.813746864210489e-06,
"loss": 2.2965,
"step": 645
},
{
"epoch": 0.4694767441860465,
"grad_norm": 1.2680090066437342,
"learning_rate": 5.8020564042888015e-06,
"loss": 2.2938,
"step": 646
},
{
"epoch": 0.470203488372093,
"grad_norm": 1.5648914252969863,
"learning_rate": 5.790361442243605e-06,
"loss": 2.307,
"step": 647
},
{
"epoch": 0.47093023255813954,
"grad_norm": 1.0883293797548825,
"learning_rate": 5.778662043721359e-06,
"loss": 2.3145,
"step": 648
},
{
"epoch": 0.47165697674418605,
"grad_norm": 1.1178636928212937,
"learning_rate": 5.766958274393428e-06,
"loss": 2.3307,
"step": 649
},
{
"epoch": 0.47238372093023256,
"grad_norm": 1.2368388872888687,
"learning_rate": 5.7552501999557065e-06,
"loss": 2.2741,
"step": 650
},
{
"epoch": 0.4731104651162791,
"grad_norm": 1.1358159404465558,
"learning_rate": 5.743537886128258e-06,
"loss": 2.2993,
"step": 651
},
{
"epoch": 0.4738372093023256,
"grad_norm": 1.1416099375776572,
"learning_rate": 5.731821398654944e-06,
"loss": 2.2632,
"step": 652
},
{
"epoch": 0.4745639534883721,
"grad_norm": 1.7153304480233373,
"learning_rate": 5.72010080330305e-06,
"loss": 2.2899,
"step": 653
},
{
"epoch": 0.4752906976744186,
"grad_norm": 1.1389469518056305,
"learning_rate": 5.708376165862921e-06,
"loss": 2.3123,
"step": 654
},
{
"epoch": 0.47601744186046513,
"grad_norm": 1.178285535354385,
"learning_rate": 5.696647552147589e-06,
"loss": 2.3086,
"step": 655
},
{
"epoch": 0.47674418604651164,
"grad_norm": 1.49285425667658,
"learning_rate": 5.684915027992415e-06,
"loss": 2.2843,
"step": 656
},
{
"epoch": 0.47747093023255816,
"grad_norm": 1.2217725848812606,
"learning_rate": 5.673178659254698e-06,
"loss": 2.3288,
"step": 657
},
{
"epoch": 0.47819767441860467,
"grad_norm": 1.1102124449809345,
"learning_rate": 5.661438511813324e-06,
"loss": 2.2965,
"step": 658
},
{
"epoch": 0.4789244186046512,
"grad_norm": 1.0889978851875848,
"learning_rate": 5.64969465156839e-06,
"loss": 2.2919,
"step": 659
},
{
"epoch": 0.4796511627906977,
"grad_norm": 1.137859070158329,
"learning_rate": 5.637947144440832e-06,
"loss": 2.307,
"step": 660
},
{
"epoch": 0.4803779069767442,
"grad_norm": 1.0732988436678494,
"learning_rate": 5.626196056372056e-06,
"loss": 2.322,
"step": 661
},
{
"epoch": 0.4811046511627907,
"grad_norm": 1.0814609091912941,
"learning_rate": 5.614441453323571e-06,
"loss": 2.3412,
"step": 662
},
{
"epoch": 0.48183139534883723,
"grad_norm": 1.1099866069205775,
"learning_rate": 5.6026834012766155e-06,
"loss": 2.296,
"step": 663
},
{
"epoch": 0.48255813953488375,
"grad_norm": 1.3837080687086778,
"learning_rate": 5.590921966231788e-06,
"loss": 2.3248,
"step": 664
},
{
"epoch": 0.48328488372093026,
"grad_norm": 1.0810546987007168,
"learning_rate": 5.579157214208675e-06,
"loss": 2.2786,
"step": 665
},
{
"epoch": 0.48401162790697677,
"grad_norm": 1.1320880266123603,
"learning_rate": 5.567389211245486e-06,
"loss": 2.2809,
"step": 666
},
{
"epoch": 0.48473837209302323,
"grad_norm": 1.0770925548918713,
"learning_rate": 5.555618023398671e-06,
"loss": 2.267,
"step": 667
},
{
"epoch": 0.48546511627906974,
"grad_norm": 1.1382037438982808,
"learning_rate": 5.5438437167425675e-06,
"loss": 2.3134,
"step": 668
},
{
"epoch": 0.48619186046511625,
"grad_norm": 1.0790789851177505,
"learning_rate": 5.532066357369012e-06,
"loss": 2.297,
"step": 669
},
{
"epoch": 0.48691860465116277,
"grad_norm": 1.1191002552884677,
"learning_rate": 5.52028601138698e-06,
"loss": 2.2478,
"step": 670
},
{
"epoch": 0.4876453488372093,
"grad_norm": 1.2027681208143934,
"learning_rate": 5.508502744922212e-06,
"loss": 2.2869,
"step": 671
},
{
"epoch": 0.4883720930232558,
"grad_norm": 1.0310047577721608,
"learning_rate": 5.496716624116836e-06,
"loss": 2.3195,
"step": 672
},
{
"epoch": 0.4890988372093023,
"grad_norm": 1.335901274130401,
"learning_rate": 5.484927715129011e-06,
"loss": 2.328,
"step": 673
},
{
"epoch": 0.4898255813953488,
"grad_norm": 1.1541404345602513,
"learning_rate": 5.4731360841325405e-06,
"loss": 2.2637,
"step": 674
},
{
"epoch": 0.49055232558139533,
"grad_norm": 1.3074779848860199,
"learning_rate": 5.46134179731651e-06,
"loss": 2.313,
"step": 675
},
{
"epoch": 0.49127906976744184,
"grad_norm": 1.1990590115578161,
"learning_rate": 5.449544920884912e-06,
"loss": 2.2838,
"step": 676
},
{
"epoch": 0.49200581395348836,
"grad_norm": 1.1453336133464966,
"learning_rate": 5.437745521056272e-06,
"loss": 2.309,
"step": 677
},
{
"epoch": 0.49273255813953487,
"grad_norm": 1.2139105440204063,
"learning_rate": 5.425943664063284e-06,
"loss": 2.3336,
"step": 678
},
{
"epoch": 0.4934593023255814,
"grad_norm": 1.2636132251369618,
"learning_rate": 5.414139416152435e-06,
"loss": 2.328,
"step": 679
},
{
"epoch": 0.4941860465116279,
"grad_norm": 1.208004638217351,
"learning_rate": 5.402332843583631e-06,
"loss": 2.2732,
"step": 680
},
{
"epoch": 0.4949127906976744,
"grad_norm": 1.18632348684088,
"learning_rate": 5.390524012629824e-06,
"loss": 2.3023,
"step": 681
},
{
"epoch": 0.4956395348837209,
"grad_norm": 1.73169268404447,
"learning_rate": 5.3787129895766484e-06,
"loss": 2.3595,
"step": 682
},
{
"epoch": 0.49636627906976744,
"grad_norm": 1.22781474125336,
"learning_rate": 5.3668998407220385e-06,
"loss": 2.3425,
"step": 683
},
{
"epoch": 0.49709302325581395,
"grad_norm": 1.4004816370059154,
"learning_rate": 5.3550846323758666e-06,
"loss": 2.2719,
"step": 684
},
{
"epoch": 0.49781976744186046,
"grad_norm": 1.175122817240152,
"learning_rate": 5.343267430859559e-06,
"loss": 2.2725,
"step": 685
},
{
"epoch": 0.498546511627907,
"grad_norm": 1.2765509033619677,
"learning_rate": 5.331448302505736e-06,
"loss": 2.2324,
"step": 686
},
{
"epoch": 0.4992732558139535,
"grad_norm": 1.134665763858464,
"learning_rate": 5.319627313657829e-06,
"loss": 2.295,
"step": 687
},
{
"epoch": 0.5,
"grad_norm": 1.2622065199475687,
"learning_rate": 5.3078045306697154e-06,
"loss": 2.3023,
"step": 688
},
{
"epoch": 0.5007267441860465,
"grad_norm": 1.103313759369295,
"learning_rate": 5.295980019905342e-06,
"loss": 2.2914,
"step": 689
},
{
"epoch": 0.501453488372093,
"grad_norm": 1.0658556313608603,
"learning_rate": 5.284153847738356e-06,
"loss": 2.2972,
"step": 690
},
{
"epoch": 0.5021802325581395,
"grad_norm": 1.1540767104639644,
"learning_rate": 5.272326080551729e-06,
"loss": 2.3199,
"step": 691
},
{
"epoch": 0.502906976744186,
"grad_norm": 1.0685698322062964,
"learning_rate": 5.260496784737386e-06,
"loss": 2.3586,
"step": 692
},
{
"epoch": 0.5036337209302325,
"grad_norm": 1.075137770401511,
"learning_rate": 5.248666026695835e-06,
"loss": 2.3327,
"step": 693
},
{
"epoch": 0.5043604651162791,
"grad_norm": 1.2485248045497954,
"learning_rate": 5.236833872835785e-06,
"loss": 2.2846,
"step": 694
},
{
"epoch": 0.5050872093023255,
"grad_norm": 1.1866909415018816,
"learning_rate": 5.2250003895737865e-06,
"loss": 2.347,
"step": 695
},
{
"epoch": 0.5058139534883721,
"grad_norm": 1.1733792890442774,
"learning_rate": 5.213165643333851e-06,
"loss": 2.2621,
"step": 696
},
{
"epoch": 0.5065406976744186,
"grad_norm": 1.0957320197546991,
"learning_rate": 5.201329700547077e-06,
"loss": 2.319,
"step": 697
},
{
"epoch": 0.5072674418604651,
"grad_norm": 1.783025447548493,
"learning_rate": 5.1894926276512824e-06,
"loss": 2.331,
"step": 698
},
{
"epoch": 0.5079941860465116,
"grad_norm": 1.1016261702586139,
"learning_rate": 5.177654491090627e-06,
"loss": 2.2892,
"step": 699
},
{
"epoch": 0.5087209302325582,
"grad_norm": 1.3255266643870636,
"learning_rate": 5.1658153573152405e-06,
"loss": 2.3513,
"step": 700
},
{
"epoch": 0.5094476744186046,
"grad_norm": 1.130548269660744,
"learning_rate": 5.153975292780852e-06,
"loss": 2.3004,
"step": 701
},
{
"epoch": 0.5101744186046512,
"grad_norm": 1.2122532789242535,
"learning_rate": 5.1421343639484165e-06,
"loss": 2.2811,
"step": 702
},
{
"epoch": 0.5109011627906976,
"grad_norm": 1.1880143344840974,
"learning_rate": 5.130292637283735e-06,
"loss": 2.3121,
"step": 703
},
{
"epoch": 0.5116279069767442,
"grad_norm": 1.4083680998178665,
"learning_rate": 5.118450179257091e-06,
"loss": 2.2639,
"step": 704
},
{
"epoch": 0.5123546511627907,
"grad_norm": 1.2695136745113604,
"learning_rate": 5.1066070563428736e-06,
"loss": 2.2975,
"step": 705
},
{
"epoch": 0.5130813953488372,
"grad_norm": 1.1487564964506674,
"learning_rate": 5.0947633350192035e-06,
"loss": 2.313,
"step": 706
},
{
"epoch": 0.5138081395348837,
"grad_norm": 1.2133446995418098,
"learning_rate": 5.082919081767558e-06,
"loss": 2.2599,
"step": 707
},
{
"epoch": 0.5145348837209303,
"grad_norm": 1.060658390712619,
"learning_rate": 5.071074363072403e-06,
"loss": 2.3123,
"step": 708
},
{
"epoch": 0.5152616279069767,
"grad_norm": 1.2238495475049125,
"learning_rate": 5.059229245420819e-06,
"loss": 2.307,
"step": 709
},
{
"epoch": 0.5159883720930233,
"grad_norm": 1.2300813441464897,
"learning_rate": 5.047383795302119e-06,
"loss": 2.3029,
"step": 710
},
{
"epoch": 0.5167151162790697,
"grad_norm": 1.0721690467315328,
"learning_rate": 5.035538079207488e-06,
"loss": 2.2998,
"step": 711
},
{
"epoch": 0.5174418604651163,
"grad_norm": 1.0965555474718804,
"learning_rate": 5.023692163629603e-06,
"loss": 2.3067,
"step": 712
},
{
"epoch": 0.5181686046511628,
"grad_norm": 1.0914497068226068,
"learning_rate": 5.01184611506226e-06,
"loss": 2.3203,
"step": 713
},
{
"epoch": 0.5188953488372093,
"grad_norm": 1.0861900942041762,
"learning_rate": 5e-06,
"loss": 2.3539,
"step": 714
},
{
"epoch": 0.5196220930232558,
"grad_norm": 1.1817757691113775,
"learning_rate": 4.988153884937742e-06,
"loss": 2.323,
"step": 715
},
{
"epoch": 0.5203488372093024,
"grad_norm": 1.285821525788281,
"learning_rate": 4.9763078363703975e-06,
"loss": 2.2366,
"step": 716
},
{
"epoch": 0.5210755813953488,
"grad_norm": 1.2634506605635316,
"learning_rate": 4.964461920792512e-06,
"loss": 2.3212,
"step": 717
},
{
"epoch": 0.5218023255813954,
"grad_norm": 1.0448269993418493,
"learning_rate": 4.952616204697882e-06,
"loss": 2.3133,
"step": 718
},
{
"epoch": 0.5225290697674418,
"grad_norm": 1.0529467546593827,
"learning_rate": 4.940770754579183e-06,
"loss": 2.2669,
"step": 719
},
{
"epoch": 0.5232558139534884,
"grad_norm": 1.0748992392149725,
"learning_rate": 4.928925636927597e-06,
"loss": 2.2731,
"step": 720
},
{
"epoch": 0.5239825581395349,
"grad_norm": 1.2467135751557892,
"learning_rate": 4.917080918232444e-06,
"loss": 2.3027,
"step": 721
},
{
"epoch": 0.5247093023255814,
"grad_norm": 1.3441280421823978,
"learning_rate": 4.905236664980797e-06,
"loss": 2.2939,
"step": 722
},
{
"epoch": 0.5254360465116279,
"grad_norm": 1.148440473548479,
"learning_rate": 4.893392943657127e-06,
"loss": 2.2725,
"step": 723
},
{
"epoch": 0.5261627906976745,
"grad_norm": 1.0956915790093777,
"learning_rate": 4.88154982074291e-06,
"loss": 2.278,
"step": 724
},
{
"epoch": 0.5268895348837209,
"grad_norm": 1.0794100545360528,
"learning_rate": 4.8697073627162675e-06,
"loss": 2.344,
"step": 725
},
{
"epoch": 0.5276162790697675,
"grad_norm": 1.0353531761976227,
"learning_rate": 4.857865636051586e-06,
"loss": 2.3008,
"step": 726
},
{
"epoch": 0.528343023255814,
"grad_norm": 1.2156865801913177,
"learning_rate": 4.846024707219149e-06,
"loss": 2.3188,
"step": 727
},
{
"epoch": 0.5290697674418605,
"grad_norm": 1.112920842833822,
"learning_rate": 4.834184642684762e-06,
"loss": 2.3114,
"step": 728
},
{
"epoch": 0.529796511627907,
"grad_norm": 1.0831654219932778,
"learning_rate": 4.822345508909376e-06,
"loss": 2.2565,
"step": 729
},
{
"epoch": 0.5305232558139535,
"grad_norm": 1.0230080869234226,
"learning_rate": 4.810507372348721e-06,
"loss": 2.2941,
"step": 730
},
{
"epoch": 0.53125,
"grad_norm": 1.1561917177674794,
"learning_rate": 4.798670299452926e-06,
"loss": 2.3085,
"step": 731
},
{
"epoch": 0.5319767441860465,
"grad_norm": 1.0618213129594278,
"learning_rate": 4.786834356666153e-06,
"loss": 2.3032,
"step": 732
},
{
"epoch": 0.532703488372093,
"grad_norm": 1.1594693610124356,
"learning_rate": 4.774999610426216e-06,
"loss": 2.341,
"step": 733
},
{
"epoch": 0.5334302325581395,
"grad_norm": 1.169649137664396,
"learning_rate": 4.7631661271642185e-06,
"loss": 2.2809,
"step": 734
},
{
"epoch": 0.534156976744186,
"grad_norm": 1.0549049328053892,
"learning_rate": 4.751333973304166e-06,
"loss": 2.333,
"step": 735
},
{
"epoch": 0.5348837209302325,
"grad_norm": 1.0038525318185807,
"learning_rate": 4.739503215262614e-06,
"loss": 2.3116,
"step": 736
},
{
"epoch": 0.5356104651162791,
"grad_norm": 1.31997708555104,
"learning_rate": 4.727673919448271e-06,
"loss": 2.2741,
"step": 737
},
{
"epoch": 0.5363372093023255,
"grad_norm": 2.627288102246105,
"learning_rate": 4.715846152261645e-06,
"loss": 2.3146,
"step": 738
},
{
"epoch": 0.5370639534883721,
"grad_norm": 1.2734425519555057,
"learning_rate": 4.704019980094659e-06,
"loss": 2.2962,
"step": 739
},
{
"epoch": 0.5377906976744186,
"grad_norm": 1.0854352730012788,
"learning_rate": 4.692195469330286e-06,
"loss": 2.3271,
"step": 740
},
{
"epoch": 0.5385174418604651,
"grad_norm": 1.2912493479896627,
"learning_rate": 4.680372686342173e-06,
"loss": 2.3492,
"step": 741
},
{
"epoch": 0.5392441860465116,
"grad_norm": 1.1007088516087424,
"learning_rate": 4.668551697494265e-06,
"loss": 2.2956,
"step": 742
},
{
"epoch": 0.5399709302325582,
"grad_norm": 1.2218535001454547,
"learning_rate": 4.656732569140441e-06,
"loss": 2.3164,
"step": 743
},
{
"epoch": 0.5406976744186046,
"grad_norm": 1.1620697376080404,
"learning_rate": 4.644915367624134e-06,
"loss": 2.3121,
"step": 744
},
{
"epoch": 0.5414244186046512,
"grad_norm": 1.0853929215895797,
"learning_rate": 4.6331001592779615e-06,
"loss": 2.2714,
"step": 745
},
{
"epoch": 0.5421511627906976,
"grad_norm": 1.2845929065378492,
"learning_rate": 4.621287010423353e-06,
"loss": 2.2752,
"step": 746
},
{
"epoch": 0.5428779069767442,
"grad_norm": 1.7105298123808783,
"learning_rate": 4.609475987370177e-06,
"loss": 2.226,
"step": 747
},
{
"epoch": 0.5436046511627907,
"grad_norm": 1.089059131220579,
"learning_rate": 4.597667156416371e-06,
"loss": 2.2778,
"step": 748
},
{
"epoch": 0.5443313953488372,
"grad_norm": 1.16904595629462,
"learning_rate": 4.585860583847566e-06,
"loss": 2.2746,
"step": 749
},
{
"epoch": 0.5450581395348837,
"grad_norm": 1.0790766039012891,
"learning_rate": 4.5740563359367164e-06,
"loss": 2.3374,
"step": 750
},
{
"epoch": 0.5457848837209303,
"grad_norm": 1.1928040971714162,
"learning_rate": 4.562254478943729e-06,
"loss": 2.295,
"step": 751
},
{
"epoch": 0.5465116279069767,
"grad_norm": 1.0780584166585039,
"learning_rate": 4.550455079115091e-06,
"loss": 2.3019,
"step": 752
},
{
"epoch": 0.5472383720930233,
"grad_norm": 1.2031883158838685,
"learning_rate": 4.53865820268349e-06,
"loss": 2.3214,
"step": 753
},
{
"epoch": 0.5479651162790697,
"grad_norm": 1.165919792008399,
"learning_rate": 4.52686391586746e-06,
"loss": 2.3162,
"step": 754
},
{
"epoch": 0.5486918604651163,
"grad_norm": 1.0614325726147842,
"learning_rate": 4.51507228487099e-06,
"loss": 2.3383,
"step": 755
},
{
"epoch": 0.5494186046511628,
"grad_norm": 1.6562096024577802,
"learning_rate": 4.503283375883165e-06,
"loss": 2.2749,
"step": 756
},
{
"epoch": 0.5501453488372093,
"grad_norm": 1.032325987733994,
"learning_rate": 4.49149725507779e-06,
"loss": 2.3248,
"step": 757
},
{
"epoch": 0.5508720930232558,
"grad_norm": 1.1761618418955344,
"learning_rate": 4.479713988613021e-06,
"loss": 2.3352,
"step": 758
},
{
"epoch": 0.5515988372093024,
"grad_norm": 1.4407183902299339,
"learning_rate": 4.467933642630989e-06,
"loss": 2.3129,
"step": 759
},
{
"epoch": 0.5523255813953488,
"grad_norm": 1.7485899959244522,
"learning_rate": 4.456156283257433e-06,
"loss": 2.2731,
"step": 760
},
{
"epoch": 0.5530523255813954,
"grad_norm": 1.067156350753222,
"learning_rate": 4.44438197660133e-06,
"loss": 2.2612,
"step": 761
},
{
"epoch": 0.5537790697674418,
"grad_norm": 1.1944089605557895,
"learning_rate": 4.432610788754517e-06,
"loss": 2.3038,
"step": 762
},
{
"epoch": 0.5545058139534884,
"grad_norm": 1.0693481833494738,
"learning_rate": 4.420842785791326e-06,
"loss": 2.3026,
"step": 763
},
{
"epoch": 0.5552325581395349,
"grad_norm": 1.038170071261918,
"learning_rate": 4.409078033768214e-06,
"loss": 2.3048,
"step": 764
},
{
"epoch": 0.5559593023255814,
"grad_norm": 1.0195295401284605,
"learning_rate": 4.397316598723385e-06,
"loss": 2.3144,
"step": 765
},
{
"epoch": 0.5566860465116279,
"grad_norm": 1.0481555681045303,
"learning_rate": 4.3855585466764305e-06,
"loss": 2.3215,
"step": 766
},
{
"epoch": 0.5574127906976745,
"grad_norm": 1.0610663400067764,
"learning_rate": 4.373803943627946e-06,
"loss": 2.2499,
"step": 767
},
{
"epoch": 0.5581395348837209,
"grad_norm": 1.3765746707741413,
"learning_rate": 4.362052855559171e-06,
"loss": 2.3205,
"step": 768
},
{
"epoch": 0.5588662790697675,
"grad_norm": 1.3701617218217217,
"learning_rate": 4.350305348431612e-06,
"loss": 2.3042,
"step": 769
},
{
"epoch": 0.559593023255814,
"grad_norm": 1.4655827010991576,
"learning_rate": 4.338561488186678e-06,
"loss": 2.2732,
"step": 770
},
{
"epoch": 0.5603197674418605,
"grad_norm": 1.269480197875066,
"learning_rate": 4.326821340745304e-06,
"loss": 2.3294,
"step": 771
},
{
"epoch": 0.561046511627907,
"grad_norm": 1.2015230300920634,
"learning_rate": 4.315084972007587e-06,
"loss": 2.3379,
"step": 772
},
{
"epoch": 0.5617732558139535,
"grad_norm": 1.6683359319886466,
"learning_rate": 4.303352447852412e-06,
"loss": 2.2795,
"step": 773
},
{
"epoch": 0.5625,
"grad_norm": 1.145915220611224,
"learning_rate": 4.291623834137082e-06,
"loss": 2.3492,
"step": 774
},
{
"epoch": 0.5632267441860465,
"grad_norm": 1.0525245783287247,
"learning_rate": 4.279899196696953e-06,
"loss": 2.2973,
"step": 775
},
{
"epoch": 0.563953488372093,
"grad_norm": 1.0792534481764795,
"learning_rate": 4.268178601345057e-06,
"loss": 2.3456,
"step": 776
},
{
"epoch": 0.5646802325581395,
"grad_norm": 1.0722001996118922,
"learning_rate": 4.256462113871741e-06,
"loss": 2.3002,
"step": 777
},
{
"epoch": 0.565406976744186,
"grad_norm": 1.1005919176139587,
"learning_rate": 4.2447498000442935e-06,
"loss": 2.2957,
"step": 778
},
{
"epoch": 0.5661337209302325,
"grad_norm": 1.0155813164835708,
"learning_rate": 4.233041725606573e-06,
"loss": 2.3771,
"step": 779
},
{
"epoch": 0.5668604651162791,
"grad_norm": 1.1740941879042222,
"learning_rate": 4.2213379562786406e-06,
"loss": 2.3246,
"step": 780
},
{
"epoch": 0.5675872093023255,
"grad_norm": 1.0820477727645355,
"learning_rate": 4.209638557756396e-06,
"loss": 2.306,
"step": 781
},
{
"epoch": 0.5683139534883721,
"grad_norm": 1.184123545414512,
"learning_rate": 4.1979435957111984e-06,
"loss": 2.3486,
"step": 782
},
{
"epoch": 0.5690406976744186,
"grad_norm": 1.2273527031302494,
"learning_rate": 4.186253135789511e-06,
"loss": 2.3706,
"step": 783
},
{
"epoch": 0.5697674418604651,
"grad_norm": 2.5274481337597683,
"learning_rate": 4.1745672436125205e-06,
"loss": 2.2692,
"step": 784
},
{
"epoch": 0.5704941860465116,
"grad_norm": 1.1293062589962266,
"learning_rate": 4.162885984775777e-06,
"loss": 2.3382,
"step": 785
},
{
"epoch": 0.5712209302325582,
"grad_norm": 1.1904762121908412,
"learning_rate": 4.15120942484882e-06,
"loss": 2.2992,
"step": 786
},
{
"epoch": 0.5719476744186046,
"grad_norm": 1.0604891438581503,
"learning_rate": 4.139537629374814e-06,
"loss": 2.2331,
"step": 787
},
{
"epoch": 0.5726744186046512,
"grad_norm": 1.9696331643707805,
"learning_rate": 4.12787066387018e-06,
"loss": 2.2814,
"step": 788
},
{
"epoch": 0.5734011627906976,
"grad_norm": 1.101600496243351,
"learning_rate": 4.116208593824227e-06,
"loss": 2.2422,
"step": 789
},
{
"epoch": 0.5741279069767442,
"grad_norm": 1.116699022328913,
"learning_rate": 4.104551484698785e-06,
"loss": 2.2757,
"step": 790
},
{
"epoch": 0.5748546511627907,
"grad_norm": 1.1199526384176086,
"learning_rate": 4.092899401927836e-06,
"loss": 2.3383,
"step": 791
},
{
"epoch": 0.5755813953488372,
"grad_norm": 1.093552386009082,
"learning_rate": 4.081252410917148e-06,
"loss": 2.2934,
"step": 792
},
{
"epoch": 0.5763081395348837,
"grad_norm": 1.0487756756149464,
"learning_rate": 4.069610577043912e-06,
"loss": 2.2438,
"step": 793
},
{
"epoch": 0.5770348837209303,
"grad_norm": 1.1051452936093074,
"learning_rate": 4.057973965656365e-06,
"loss": 2.2804,
"step": 794
},
{
"epoch": 0.5777616279069767,
"grad_norm": 1.1164624106499657,
"learning_rate": 4.046342642073433e-06,
"loss": 2.282,
"step": 795
},
{
"epoch": 0.5784883720930233,
"grad_norm": 0.9908772046288108,
"learning_rate": 4.034716671584357e-06,
"loss": 2.2756,
"step": 796
},
{
"epoch": 0.5792151162790697,
"grad_norm": 1.0456373042936773,
"learning_rate": 4.0230961194483325e-06,
"loss": 2.3089,
"step": 797
},
{
"epoch": 0.5799418604651163,
"grad_norm": 1.0795700462534907,
"learning_rate": 4.01148105089414e-06,
"loss": 2.3303,
"step": 798
},
{
"epoch": 0.5806686046511628,
"grad_norm": 1.1053174271707678,
"learning_rate": 3.999871531119779e-06,
"loss": 2.3436,
"step": 799
},
{
"epoch": 0.5813953488372093,
"grad_norm": 1.1301776518164668,
"learning_rate": 3.988267625292102e-06,
"loss": 2.2661,
"step": 800
},
{
"epoch": 0.5821220930232558,
"grad_norm": 1.172930982684841,
"learning_rate": 3.976669398546451e-06,
"loss": 2.345,
"step": 801
},
{
"epoch": 0.5828488372093024,
"grad_norm": 1.0890559009206928,
"learning_rate": 3.9650769159862875e-06,
"loss": 2.2633,
"step": 802
},
{
"epoch": 0.5835755813953488,
"grad_norm": 1.0775106721417829,
"learning_rate": 3.9534902426828325e-06,
"loss": 2.2898,
"step": 803
},
{
"epoch": 0.5843023255813954,
"grad_norm": 1.1586454869442395,
"learning_rate": 3.941909443674696e-06,
"loss": 2.2771,
"step": 804
},
{
"epoch": 0.5850290697674418,
"grad_norm": 1.176709888292826,
"learning_rate": 3.930334583967514e-06,
"loss": 2.2887,
"step": 805
},
{
"epoch": 0.5857558139534884,
"grad_norm": 1.0930311463479425,
"learning_rate": 3.918765728533586e-06,
"loss": 2.2662,
"step": 806
},
{
"epoch": 0.5864825581395349,
"grad_norm": 0.9869447591684988,
"learning_rate": 3.907202942311506e-06,
"loss": 2.2685,
"step": 807
},
{
"epoch": 0.5872093023255814,
"grad_norm": 1.1593303017123573,
"learning_rate": 3.895646290205803e-06,
"loss": 2.3404,
"step": 808
},
{
"epoch": 0.5879360465116279,
"grad_norm": 1.3958293581245043,
"learning_rate": 3.884095837086571e-06,
"loss": 2.3046,
"step": 809
},
{
"epoch": 0.5886627906976745,
"grad_norm": 1.0971451664444143,
"learning_rate": 3.872551647789108e-06,
"loss": 2.2717,
"step": 810
},
{
"epoch": 0.5893895348837209,
"grad_norm": 3.09870534328071,
"learning_rate": 3.861013787113553e-06,
"loss": 2.3797,
"step": 811
},
{
"epoch": 0.5901162790697675,
"grad_norm": 1.141704824594259,
"learning_rate": 3.849482319824521e-06,
"loss": 2.296,
"step": 812
},
{
"epoch": 0.590843023255814,
"grad_norm": 0.9859445781082979,
"learning_rate": 3.837957310650738e-06,
"loss": 2.3019,
"step": 813
},
{
"epoch": 0.5915697674418605,
"grad_norm": 1.0950233921797288,
"learning_rate": 3.82643882428468e-06,
"loss": 2.3029,
"step": 814
},
{
"epoch": 0.592296511627907,
"grad_norm": 1.0700692250622799,
"learning_rate": 3.81492692538221e-06,
"loss": 2.3141,
"step": 815
},
{
"epoch": 0.5930232558139535,
"grad_norm": 1.091662316980483,
"learning_rate": 3.803421678562213e-06,
"loss": 2.2995,
"step": 816
},
{
"epoch": 0.59375,
"grad_norm": 1.204369219463071,
"learning_rate": 3.7919231484062334e-06,
"loss": 2.2585,
"step": 817
},
{
"epoch": 0.5944767441860465,
"grad_norm": 1.0764274721400988,
"learning_rate": 3.7804313994581143e-06,
"loss": 2.2878,
"step": 818
},
{
"epoch": 0.595203488372093,
"grad_norm": 1.1663852917982485,
"learning_rate": 3.7689464962236367e-06,
"loss": 2.2805,
"step": 819
},
{
"epoch": 0.5959302325581395,
"grad_norm": 1.0119449535199592,
"learning_rate": 3.757468503170153e-06,
"loss": 2.3124,
"step": 820
},
{
"epoch": 0.596656976744186,
"grad_norm": 0.9879631073022755,
"learning_rate": 3.7459974847262253e-06,
"loss": 2.2971,
"step": 821
},
{
"epoch": 0.5973837209302325,
"grad_norm": 1.0980204844106818,
"learning_rate": 3.734533505281269e-06,
"loss": 2.3215,
"step": 822
},
{
"epoch": 0.5981104651162791,
"grad_norm": 1.1727982952700535,
"learning_rate": 3.723076629185186e-06,
"loss": 2.3073,
"step": 823
},
{
"epoch": 0.5988372093023255,
"grad_norm": 1.095773058831185,
"learning_rate": 3.7116269207480055e-06,
"loss": 2.3034,
"step": 824
},
{
"epoch": 0.5995639534883721,
"grad_norm": 1.1102423289772532,
"learning_rate": 3.700184444239524e-06,
"loss": 2.3286,
"step": 825
},
{
"epoch": 0.6002906976744186,
"grad_norm": 1.2168841679733824,
"learning_rate": 3.6887492638889433e-06,
"loss": 2.3142,
"step": 826
},
{
"epoch": 0.6010174418604651,
"grad_norm": 1.1239188965273867,
"learning_rate": 3.677321443884509e-06,
"loss": 2.3021,
"step": 827
},
{
"epoch": 0.6017441860465116,
"grad_norm": 1.0061388225584935,
"learning_rate": 3.6659010483731543e-06,
"loss": 2.2644,
"step": 828
},
{
"epoch": 0.6024709302325582,
"grad_norm": 1.0745214690142126,
"learning_rate": 3.654488141460134e-06,
"loss": 2.3161,
"step": 829
},
{
"epoch": 0.6031976744186046,
"grad_norm": 1.066456873356294,
"learning_rate": 3.6430827872086694e-06,
"loss": 2.3044,
"step": 830
},
{
"epoch": 0.6039244186046512,
"grad_norm": 1.0201857786513382,
"learning_rate": 3.6316850496395863e-06,
"loss": 2.344,
"step": 831
},
{
"epoch": 0.6046511627906976,
"grad_norm": 1.3732462363703541,
"learning_rate": 3.6202949927309555e-06,
"loss": 2.3352,
"step": 832
},
{
"epoch": 0.6053779069767442,
"grad_norm": 1.0703274758634678,
"learning_rate": 3.6089126804177373e-06,
"loss": 2.283,
"step": 833
},
{
"epoch": 0.6061046511627907,
"grad_norm": 1.183872355394252,
"learning_rate": 3.597538176591417e-06,
"loss": 2.2443,
"step": 834
},
{
"epoch": 0.6068313953488372,
"grad_norm": 2.828409917772514,
"learning_rate": 3.5861715450996505e-06,
"loss": 2.3027,
"step": 835
},
{
"epoch": 0.6075581395348837,
"grad_norm": 1.0453232535385137,
"learning_rate": 3.5748128497459044e-06,
"loss": 2.2397,
"step": 836
},
{
"epoch": 0.6082848837209303,
"grad_norm": 1.0869450979441266,
"learning_rate": 3.563462154289098e-06,
"loss": 2.2687,
"step": 837
},
{
"epoch": 0.6090116279069767,
"grad_norm": 1.0621257836663385,
"learning_rate": 3.5521195224432436e-06,
"loss": 2.2794,
"step": 838
},
{
"epoch": 0.6097383720930233,
"grad_norm": 1.0092312203378675,
"learning_rate": 3.5407850178770944e-06,
"loss": 2.334,
"step": 839
},
{
"epoch": 0.6104651162790697,
"grad_norm": 1.105493758604535,
"learning_rate": 3.5294587042137796e-06,
"loss": 2.318,
"step": 840
},
{
"epoch": 0.6111918604651163,
"grad_norm": 1.1993001659775373,
"learning_rate": 3.5181406450304536e-06,
"loss": 2.2729,
"step": 841
},
{
"epoch": 0.6119186046511628,
"grad_norm": 1.0688907809651846,
"learning_rate": 3.506830903857933e-06,
"loss": 2.3048,
"step": 842
},
{
"epoch": 0.6126453488372093,
"grad_norm": 1.2797486907093922,
"learning_rate": 3.49552954418035e-06,
"loss": 2.3112,
"step": 843
},
{
"epoch": 0.6133720930232558,
"grad_norm": 1.0551236479337645,
"learning_rate": 3.484236629434783e-06,
"loss": 2.2843,
"step": 844
},
{
"epoch": 0.6140988372093024,
"grad_norm": 1.0368752476104846,
"learning_rate": 3.4729522230109103e-06,
"loss": 2.275,
"step": 845
},
{
"epoch": 0.6148255813953488,
"grad_norm": 1.2212850537214994,
"learning_rate": 3.461676388250651e-06,
"loss": 2.3211,
"step": 846
},
{
"epoch": 0.6155523255813954,
"grad_norm": 1.1497485865405142,
"learning_rate": 3.4504091884478076e-06,
"loss": 2.2402,
"step": 847
},
{
"epoch": 0.6162790697674418,
"grad_norm": 1.0906549367951974,
"learning_rate": 3.4391506868477153e-06,
"loss": 2.2838,
"step": 848
},
{
"epoch": 0.6170058139534884,
"grad_norm": 1.1230941195693127,
"learning_rate": 3.4279009466468825e-06,
"loss": 2.3212,
"step": 849
},
{
"epoch": 0.6177325581395349,
"grad_norm": 1.4108103209541318,
"learning_rate": 3.416660030992639e-06,
"loss": 2.2671,
"step": 850
},
{
"epoch": 0.6184593023255814,
"grad_norm": 3.170948816296744,
"learning_rate": 3.405428002982779e-06,
"loss": 2.3123,
"step": 851
},
{
"epoch": 0.6191860465116279,
"grad_norm": 1.1804134414107927,
"learning_rate": 3.3942049256652093e-06,
"loss": 2.3191,
"step": 852
},
{
"epoch": 0.6199127906976745,
"grad_norm": 1.0937922010266585,
"learning_rate": 3.3829908620375953e-06,
"loss": 2.2751,
"step": 853
},
{
"epoch": 0.6206395348837209,
"grad_norm": 1.1172481321734316,
"learning_rate": 3.3717858750470046e-06,
"loss": 2.2579,
"step": 854
},
{
"epoch": 0.6213662790697675,
"grad_norm": 1.1200023639589576,
"learning_rate": 3.3605900275895565e-06,
"loss": 2.2352,
"step": 855
},
{
"epoch": 0.622093023255814,
"grad_norm": 1.0708024845639486,
"learning_rate": 3.349403382510068e-06,
"loss": 2.2787,
"step": 856
},
{
"epoch": 0.6228197674418605,
"grad_norm": 1.084888989250318,
"learning_rate": 3.3382260026017027e-06,
"loss": 2.232,
"step": 857
},
{
"epoch": 0.623546511627907,
"grad_norm": 1.268520820863664,
"learning_rate": 3.3270579506056146e-06,
"loss": 2.2854,
"step": 858
},
{
"epoch": 0.6242732558139535,
"grad_norm": 1.0779562508013507,
"learning_rate": 3.3158992892105975e-06,
"loss": 2.3121,
"step": 859
},
{
"epoch": 0.625,
"grad_norm": 1.152571981786872,
"learning_rate": 3.3047500810527343e-06,
"loss": 2.2762,
"step": 860
},
{
"epoch": 0.6257267441860465,
"grad_norm": 1.1127590575536852,
"learning_rate": 3.2936103887150484e-06,
"loss": 2.3208,
"step": 861
},
{
"epoch": 0.626453488372093,
"grad_norm": 1.1100878449841611,
"learning_rate": 3.2824802747271424e-06,
"loss": 2.3413,
"step": 862
},
{
"epoch": 0.6271802325581395,
"grad_norm": 1.0795636271079854,
"learning_rate": 3.271359801564858e-06,
"loss": 2.2705,
"step": 863
},
{
"epoch": 0.627906976744186,
"grad_norm": 1.0960995727635903,
"learning_rate": 3.2602490316499197e-06,
"loss": 2.2201,
"step": 864
},
{
"epoch": 0.6286337209302325,
"grad_norm": 2.9347541178607575,
"learning_rate": 3.2491480273495847e-06,
"loss": 2.3146,
"step": 865
},
{
"epoch": 0.6293604651162791,
"grad_norm": 1.0617998945563392,
"learning_rate": 3.2380568509762935e-06,
"loss": 2.3156,
"step": 866
},
{
"epoch": 0.6300872093023255,
"grad_norm": 1.0271455481558849,
"learning_rate": 3.226975564787322e-06,
"loss": 2.277,
"step": 867
},
{
"epoch": 0.6308139534883721,
"grad_norm": 1.3002586280292399,
"learning_rate": 3.215904230984428e-06,
"loss": 2.3272,
"step": 868
},
{
"epoch": 0.6315406976744186,
"grad_norm": 1.5574297719183472,
"learning_rate": 3.204842911713506e-06,
"loss": 2.3306,
"step": 869
},
{
"epoch": 0.6322674418604651,
"grad_norm": 1.1428523129407389,
"learning_rate": 3.1937916690642356e-06,
"loss": 2.3228,
"step": 870
},
{
"epoch": 0.6329941860465116,
"grad_norm": 1.089861769979664,
"learning_rate": 3.182750565069735e-06,
"loss": 2.2114,
"step": 871
},
{
"epoch": 0.6337209302325582,
"grad_norm": 1.1291307856061896,
"learning_rate": 3.171719661706211e-06,
"loss": 2.2489,
"step": 872
},
{
"epoch": 0.6344476744186046,
"grad_norm": 1.101392448358028,
"learning_rate": 3.1606990208926125e-06,
"loss": 2.32,
"step": 873
},
{
"epoch": 0.6351744186046512,
"grad_norm": 5.592420183479199,
"learning_rate": 3.1496887044902815e-06,
"loss": 2.293,
"step": 874
},
{
"epoch": 0.6359011627906976,
"grad_norm": 1.2099630065498752,
"learning_rate": 3.1386887743026083e-06,
"loss": 2.3239,
"step": 875
},
{
"epoch": 0.6366279069767442,
"grad_norm": 1.165045093955686,
"learning_rate": 3.127699292074683e-06,
"loss": 2.2441,
"step": 876
},
{
"epoch": 0.6373546511627907,
"grad_norm": 1.088722434436471,
"learning_rate": 3.1167203194929447e-06,
"loss": 2.2861,
"step": 877
},
{
"epoch": 0.6380813953488372,
"grad_norm": 1.2492892755052187,
"learning_rate": 3.1057519181848474e-06,
"loss": 2.3617,
"step": 878
},
{
"epoch": 0.6388081395348837,
"grad_norm": 1.1658961078716712,
"learning_rate": 3.0947941497184985e-06,
"loss": 2.2811,
"step": 879
},
{
"epoch": 0.6395348837209303,
"grad_norm": 1.4870278439353821,
"learning_rate": 3.0838470756023253e-06,
"loss": 2.2888,
"step": 880
},
{
"epoch": 0.6402616279069767,
"grad_norm": 1.103064716569564,
"learning_rate": 3.0729107572847244e-06,
"loss": 2.2916,
"step": 881
},
{
"epoch": 0.6409883720930233,
"grad_norm": 2.4560218782566583,
"learning_rate": 3.0619852561537165e-06,
"loss": 2.3084,
"step": 882
},
{
"epoch": 0.6417151162790697,
"grad_norm": 1.1742459127551743,
"learning_rate": 3.0510706335366034e-06,
"loss": 2.297,
"step": 883
},
{
"epoch": 0.6424418604651163,
"grad_norm": 1.0987221705498311,
"learning_rate": 3.040166950699626e-06,
"loss": 2.297,
"step": 884
},
{
"epoch": 0.6431686046511628,
"grad_norm": 1.071287309340485,
"learning_rate": 3.0292742688476125e-06,
"loss": 2.3008,
"step": 885
},
{
"epoch": 0.6438953488372093,
"grad_norm": 1.0886795715194904,
"learning_rate": 3.018392649123645e-06,
"loss": 2.2998,
"step": 886
},
{
"epoch": 0.6446220930232558,
"grad_norm": 1.0354258695320315,
"learning_rate": 3.0075221526087083e-06,
"loss": 2.2977,
"step": 887
},
{
"epoch": 0.6453488372093024,
"grad_norm": 1.2586098674126731,
"learning_rate": 2.9966628403213528e-06,
"loss": 2.3182,
"step": 888
},
{
"epoch": 0.6460755813953488,
"grad_norm": 1.0639891337581244,
"learning_rate": 2.985814773217346e-06,
"loss": 2.342,
"step": 889
},
{
"epoch": 0.6468023255813954,
"grad_norm": 1.0142278093049604,
"learning_rate": 2.9749780121893366e-06,
"loss": 2.3273,
"step": 890
},
{
"epoch": 0.6475290697674418,
"grad_norm": 1.1300528337470748,
"learning_rate": 2.964152618066508e-06,
"loss": 2.2904,
"step": 891
},
{
"epoch": 0.6482558139534884,
"grad_norm": 1.0454947250521127,
"learning_rate": 2.9533386516142402e-06,
"loss": 2.2952,
"step": 892
},
{
"epoch": 0.6489825581395349,
"grad_norm": 1.0526378472042244,
"learning_rate": 2.9425361735337655e-06,
"loss": 2.2735,
"step": 893
},
{
"epoch": 0.6497093023255814,
"grad_norm": 1.080243091060958,
"learning_rate": 2.93174524446183e-06,
"loss": 2.2815,
"step": 894
},
{
"epoch": 0.6504360465116279,
"grad_norm": 1.088488914886554,
"learning_rate": 2.920965924970352e-06,
"loss": 2.2592,
"step": 895
},
{
"epoch": 0.6511627906976745,
"grad_norm": 1.1122754897361344,
"learning_rate": 2.910198275566085e-06,
"loss": 2.2884,
"step": 896
},
{
"epoch": 0.6518895348837209,
"grad_norm": 1.0528404554543893,
"learning_rate": 2.899442356690271e-06,
"loss": 2.3596,
"step": 897
},
{
"epoch": 0.6526162790697675,
"grad_norm": 1.0174816158118893,
"learning_rate": 2.8886982287183092e-06,
"loss": 2.3454,
"step": 898
},
{
"epoch": 0.653343023255814,
"grad_norm": 1.064816839778644,
"learning_rate": 2.8779659519594173e-06,
"loss": 2.2798,
"step": 899
},
{
"epoch": 0.6540697674418605,
"grad_norm": 1.1220550056919638,
"learning_rate": 2.8672455866562797e-06,
"loss": 2.3059,
"step": 900
},
{
"epoch": 0.654796511627907,
"grad_norm": 2.0980951298709227,
"learning_rate": 2.8565371929847286e-06,
"loss": 2.3817,
"step": 901
},
{
"epoch": 0.6555232558139535,
"grad_norm": 1.081746293600907,
"learning_rate": 2.8458408310533948e-06,
"loss": 2.3264,
"step": 902
},
{
"epoch": 0.65625,
"grad_norm": 1.7403806177324872,
"learning_rate": 2.835156560903365e-06,
"loss": 2.2865,
"step": 903
},
{
"epoch": 0.6569767441860465,
"grad_norm": 1.2218329122403868,
"learning_rate": 2.824484442507863e-06,
"loss": 2.3367,
"step": 904
},
{
"epoch": 0.657703488372093,
"grad_norm": 1.0595512053583265,
"learning_rate": 2.813824535771892e-06,
"loss": 2.3019,
"step": 905
},
{
"epoch": 0.6584302325581395,
"grad_norm": 1.071380698452199,
"learning_rate": 2.803176900531915e-06,
"loss": 2.3019,
"step": 906
},
{
"epoch": 0.659156976744186,
"grad_norm": 1.0825188857789525,
"learning_rate": 2.7925415965555126e-06,
"loss": 2.2516,
"step": 907
},
{
"epoch": 0.6598837209302325,
"grad_norm": 2.920822764065476,
"learning_rate": 2.78191868354104e-06,
"loss": 2.3098,
"step": 908
},
{
"epoch": 0.6606104651162791,
"grad_norm": 1.0283483488585925,
"learning_rate": 2.771308221117309e-06,
"loss": 2.2796,
"step": 909
},
{
"epoch": 0.6613372093023255,
"grad_norm": 1.108342894881162,
"learning_rate": 2.760710268843234e-06,
"loss": 2.2898,
"step": 910
},
{
"epoch": 0.6620639534883721,
"grad_norm": 1.2927271073592739,
"learning_rate": 2.7501248862075163e-06,
"loss": 2.314,
"step": 911
},
{
"epoch": 0.6627906976744186,
"grad_norm": 1.0730365777748851,
"learning_rate": 2.7395521326282913e-06,
"loss": 2.2943,
"step": 912
},
{
"epoch": 0.6635174418604651,
"grad_norm": 1.053779281550719,
"learning_rate": 2.7289920674528142e-06,
"loss": 2.3405,
"step": 913
},
{
"epoch": 0.6642441860465116,
"grad_norm": 1.0399607086713838,
"learning_rate": 2.718444749957109e-06,
"loss": 2.3139,
"step": 914
},
{
"epoch": 0.6649709302325582,
"grad_norm": 1.0368906197250551,
"learning_rate": 2.7079102393456503e-06,
"loss": 2.3257,
"step": 915
},
{
"epoch": 0.6656976744186046,
"grad_norm": 0.9905272281268529,
"learning_rate": 2.69738859475102e-06,
"loss": 2.2996,
"step": 916
},
{
"epoch": 0.6664244186046512,
"grad_norm": 1.068690583432205,
"learning_rate": 2.6868798752335867e-06,
"loss": 2.3056,
"step": 917
},
{
"epoch": 0.6671511627906976,
"grad_norm": 1.0595936340940884,
"learning_rate": 2.6763841397811576e-06,
"loss": 2.2933,
"step": 918
},
{
"epoch": 0.6678779069767442,
"grad_norm": 1.0457724244809747,
"learning_rate": 2.6659014473086665e-06,
"loss": 2.2715,
"step": 919
},
{
"epoch": 0.6686046511627907,
"grad_norm": 1.063579439322238,
"learning_rate": 2.655431856657833e-06,
"loss": 2.3071,
"step": 920
},
{
"epoch": 0.6693313953488372,
"grad_norm": 1.1221698319624531,
"learning_rate": 2.6449754265968263e-06,
"loss": 2.3282,
"step": 921
},
{
"epoch": 0.6700581395348837,
"grad_norm": 1.2349259940820547,
"learning_rate": 2.6345322158199503e-06,
"loss": 2.2832,
"step": 922
},
{
"epoch": 0.6707848837209303,
"grad_norm": 1.2893272128889726,
"learning_rate": 2.6241022829473e-06,
"loss": 2.2892,
"step": 923
},
{
"epoch": 0.6715116279069767,
"grad_norm": 1.0094392612404577,
"learning_rate": 2.6136856865244443e-06,
"loss": 2.2967,
"step": 924
},
{
"epoch": 0.6722383720930233,
"grad_norm": 1.0773647473259351,
"learning_rate": 2.603282485022085e-06,
"loss": 2.2939,
"step": 925
},
{
"epoch": 0.6729651162790697,
"grad_norm": 1.1004366282951852,
"learning_rate": 2.592892736835742e-06,
"loss": 2.3022,
"step": 926
},
{
"epoch": 0.6736918604651163,
"grad_norm": 1.209320122969423,
"learning_rate": 2.5825165002854124e-06,
"loss": 2.3231,
"step": 927
},
{
"epoch": 0.6744186046511628,
"grad_norm": 1.1212132712601037,
"learning_rate": 2.5721538336152553e-06,
"loss": 2.307,
"step": 928
},
{
"epoch": 0.6751453488372093,
"grad_norm": 1.091673096600694,
"learning_rate": 2.5618047949932524e-06,
"loss": 2.2216,
"step": 929
},
{
"epoch": 0.6758720930232558,
"grad_norm": 1.0773898030071996,
"learning_rate": 2.5514694425108968e-06,
"loss": 2.2849,
"step": 930
},
{
"epoch": 0.6765988372093024,
"grad_norm": 1.1276874114176263,
"learning_rate": 2.5411478341828475e-06,
"loss": 2.2728,
"step": 931
},
{
"epoch": 0.6773255813953488,
"grad_norm": 1.0562814394201008,
"learning_rate": 2.5308400279466262e-06,
"loss": 2.2255,
"step": 932
},
{
"epoch": 0.6780523255813954,
"grad_norm": 1.2901606524754075,
"learning_rate": 2.5205460816622684e-06,
"loss": 2.3445,
"step": 933
},
{
"epoch": 0.6787790697674418,
"grad_norm": 1.353015885378886,
"learning_rate": 2.5102660531120204e-06,
"loss": 2.2881,
"step": 934
},
{
"epoch": 0.6795058139534884,
"grad_norm": 1.032764934565106,
"learning_rate": 2.5000000000000015e-06,
"loss": 2.252,
"step": 935
},
{
"epoch": 0.6802325581395349,
"grad_norm": 1.0298405135241169,
"learning_rate": 2.4897479799518797e-06,
"loss": 2.2612,
"step": 936
},
{
"epoch": 0.6809593023255814,
"grad_norm": 1.0157695794323445,
"learning_rate": 2.479510050514561e-06,
"loss": 2.3077,
"step": 937
},
{
"epoch": 0.6816860465116279,
"grad_norm": 0.9944505259481613,
"learning_rate": 2.469286269155848e-06,
"loss": 2.2568,
"step": 938
},
{
"epoch": 0.6824127906976745,
"grad_norm": 1.2455482005978737,
"learning_rate": 2.4590766932641353e-06,
"loss": 2.2888,
"step": 939
},
{
"epoch": 0.6831395348837209,
"grad_norm": 1.0148580442463242,
"learning_rate": 2.4488813801480717e-06,
"loss": 2.2671,
"step": 940
},
{
"epoch": 0.6838662790697675,
"grad_norm": 1.0647336115989892,
"learning_rate": 2.438700387036253e-06,
"loss": 2.2511,
"step": 941
},
{
"epoch": 0.684593023255814,
"grad_norm": 1.1906818174033162,
"learning_rate": 2.4285337710768843e-06,
"loss": 2.3056,
"step": 942
},
{
"epoch": 0.6853197674418605,
"grad_norm": 1.045245906467335,
"learning_rate": 2.4183815893374817e-06,
"loss": 2.2748,
"step": 943
},
{
"epoch": 0.686046511627907,
"grad_norm": 1.2179174742303633,
"learning_rate": 2.4082438988045253e-06,
"loss": 2.3691,
"step": 944
},
{
"epoch": 0.6867732558139535,
"grad_norm": 1.0939754996774316,
"learning_rate": 2.3981207563831633e-06,
"loss": 2.2668,
"step": 945
},
{
"epoch": 0.6875,
"grad_norm": 1.0410688427836987,
"learning_rate": 2.388012218896873e-06,
"loss": 2.275,
"step": 946
},
{
"epoch": 0.6882267441860465,
"grad_norm": 1.0845983862848876,
"learning_rate": 2.3779183430871596e-06,
"loss": 2.2835,
"step": 947
},
{
"epoch": 0.688953488372093,
"grad_norm": 1.1420315061262278,
"learning_rate": 2.3678391856132203e-06,
"loss": 2.3258,
"step": 948
},
{
"epoch": 0.6896802325581395,
"grad_norm": 1.1645336497292325,
"learning_rate": 2.3577748030516443e-06,
"loss": 2.3346,
"step": 949
},
{
"epoch": 0.690406976744186,
"grad_norm": 1.1103520886088816,
"learning_rate": 2.3477252518960764e-06,
"loss": 2.2921,
"step": 950
},
{
"epoch": 0.6911337209302325,
"grad_norm": 1.0138661713171222,
"learning_rate": 2.3376905885569185e-06,
"loss": 2.2863,
"step": 951
},
{
"epoch": 0.6918604651162791,
"grad_norm": 1.7747130578282735,
"learning_rate": 2.3276708693609947e-06,
"loss": 2.3005,
"step": 952
},
{
"epoch": 0.6925872093023255,
"grad_norm": 1.2120454663077505,
"learning_rate": 2.3176661505512534e-06,
"loss": 2.2774,
"step": 953
},
{
"epoch": 0.6933139534883721,
"grad_norm": 1.139318798147674,
"learning_rate": 2.3076764882864333e-06,
"loss": 2.3056,
"step": 954
},
{
"epoch": 0.6940406976744186,
"grad_norm": 1.0685661289451227,
"learning_rate": 2.2977019386407653e-06,
"loss": 2.2886,
"step": 955
},
{
"epoch": 0.6947674418604651,
"grad_norm": 1.1453862770943386,
"learning_rate": 2.2877425576036467e-06,
"loss": 2.3176,
"step": 956
},
{
"epoch": 0.6954941860465116,
"grad_norm": 1.0783921305254762,
"learning_rate": 2.2777984010793264e-06,
"loss": 2.4089,
"step": 957
},
{
"epoch": 0.6962209302325582,
"grad_norm": 1.071086466220267,
"learning_rate": 2.267869524886603e-06,
"loss": 2.2986,
"step": 958
},
{
"epoch": 0.6969476744186046,
"grad_norm": 1.0942023435068833,
"learning_rate": 2.2579559847584924e-06,
"loss": 2.2728,
"step": 959
},
{
"epoch": 0.6976744186046512,
"grad_norm": 1.2455437847883444,
"learning_rate": 2.2480578363419363e-06,
"loss": 2.2967,
"step": 960
},
{
"epoch": 0.6984011627906976,
"grad_norm": 0.9971129904259833,
"learning_rate": 2.238175135197471e-06,
"loss": 2.3308,
"step": 961
},
{
"epoch": 0.6991279069767442,
"grad_norm": 1.070897723373824,
"learning_rate": 2.2283079367989303e-06,
"loss": 2.2437,
"step": 962
},
{
"epoch": 0.6998546511627907,
"grad_norm": 1.0863749709759165,
"learning_rate": 2.2184562965331203e-06,
"loss": 2.303,
"step": 963
},
{
"epoch": 0.7005813953488372,
"grad_norm": 1.0311502311512875,
"learning_rate": 2.2086202696995248e-06,
"loss": 2.2846,
"step": 964
},
{
"epoch": 0.7013081395348837,
"grad_norm": 1.0623408936048355,
"learning_rate": 2.1987999115099763e-06,
"loss": 2.2579,
"step": 965
},
{
"epoch": 0.7020348837209303,
"grad_norm": 1.0153193816029904,
"learning_rate": 2.1889952770883644e-06,
"loss": 2.2474,
"step": 966
},
{
"epoch": 0.7027616279069767,
"grad_norm": 1.0440507740384226,
"learning_rate": 2.17920642147031e-06,
"loss": 2.2733,
"step": 967
},
{
"epoch": 0.7034883720930233,
"grad_norm": 1.1511064044424957,
"learning_rate": 2.169433399602872e-06,
"loss": 2.3142,
"step": 968
},
{
"epoch": 0.7042151162790697,
"grad_norm": 1.053669842381643,
"learning_rate": 2.159676266344222e-06,
"loss": 2.3098,
"step": 969
},
{
"epoch": 0.7049418604651163,
"grad_norm": 1.0471770020331883,
"learning_rate": 2.1499350764633513e-06,
"loss": 2.2895,
"step": 970
},
{
"epoch": 0.7056686046511628,
"grad_norm": 1.4571732825727595,
"learning_rate": 2.140209884639759e-06,
"loss": 2.2841,
"step": 971
},
{
"epoch": 0.7063953488372093,
"grad_norm": 1.0880769462399786,
"learning_rate": 2.130500745463136e-06,
"loss": 2.2991,
"step": 972
},
{
"epoch": 0.7071220930232558,
"grad_norm": 1.037540633256132,
"learning_rate": 2.120807713433074e-06,
"loss": 2.2681,
"step": 973
},
{
"epoch": 0.7078488372093024,
"grad_norm": 1.0696892208948963,
"learning_rate": 2.1111308429587446e-06,
"loss": 2.2959,
"step": 974
},
{
"epoch": 0.7085755813953488,
"grad_norm": 1.0208444816267834,
"learning_rate": 2.1014701883586087e-06,
"loss": 2.2748,
"step": 975
},
{
"epoch": 0.7093023255813954,
"grad_norm": 1.0418009858188437,
"learning_rate": 2.091825803860095e-06,
"loss": 2.2743,
"step": 976
},
{
"epoch": 0.7100290697674418,
"grad_norm": 1.0236164793641587,
"learning_rate": 2.082197743599314e-06,
"loss": 2.3051,
"step": 977
},
{
"epoch": 0.7107558139534884,
"grad_norm": 1.208585374069975,
"learning_rate": 2.072586061620735e-06,
"loss": 2.2571,
"step": 978
},
{
"epoch": 0.7114825581395349,
"grad_norm": 1.123554743123603,
"learning_rate": 2.0629908118769004e-06,
"loss": 2.2774,
"step": 979
},
{
"epoch": 0.7122093023255814,
"grad_norm": 1.070269970918954,
"learning_rate": 2.0534120482281087e-06,
"loss": 2.3054,
"step": 980
},
{
"epoch": 0.7129360465116279,
"grad_norm": 1.0520946805260354,
"learning_rate": 2.043849824442124e-06,
"loss": 2.3158,
"step": 981
},
{
"epoch": 0.7136627906976745,
"grad_norm": 1.0634540989868737,
"learning_rate": 2.034304194193861e-06,
"loss": 2.2118,
"step": 982
},
{
"epoch": 0.7143895348837209,
"grad_norm": 1.0132991218437926,
"learning_rate": 2.024775211065098e-06,
"loss": 2.2696,
"step": 983
},
{
"epoch": 0.7151162790697675,
"grad_norm": 1.018246298526723,
"learning_rate": 2.0152629285441668e-06,
"loss": 2.248,
"step": 984
},
{
"epoch": 0.715843023255814,
"grad_norm": 1.1097222649464207,
"learning_rate": 2.0057674000256556e-06,
"loss": 2.3706,
"step": 985
},
{
"epoch": 0.7165697674418605,
"grad_norm": 1.1636566819216903,
"learning_rate": 1.996288678810105e-06,
"loss": 2.3096,
"step": 986
},
{
"epoch": 0.717296511627907,
"grad_norm": 1.1627061516985389,
"learning_rate": 1.9868268181037186e-06,
"loss": 2.3061,
"step": 987
},
{
"epoch": 0.7180232558139535,
"grad_norm": 1.0306901524108303,
"learning_rate": 1.9773818710180514e-06,
"loss": 2.323,
"step": 988
},
{
"epoch": 0.71875,
"grad_norm": 1.0526076940743654,
"learning_rate": 1.967953890569723e-06,
"loss": 2.2851,
"step": 989
},
{
"epoch": 0.7194767441860465,
"grad_norm": 1.8419498783617658,
"learning_rate": 1.958542929680117e-06,
"loss": 2.2927,
"step": 990
},
{
"epoch": 0.720203488372093,
"grad_norm": 1.1109141166503629,
"learning_rate": 1.9491490411750745e-06,
"loss": 2.2877,
"step": 991
},
{
"epoch": 0.7209302325581395,
"grad_norm": 1.0479816810432738,
"learning_rate": 1.9397722777846153e-06,
"loss": 2.291,
"step": 992
},
{
"epoch": 0.721656976744186,
"grad_norm": 1.0459670031607213,
"learning_rate": 1.9304126921426235e-06,
"loss": 2.2751,
"step": 993
},
{
"epoch": 0.7223837209302325,
"grad_norm": 1.0525843890419861,
"learning_rate": 1.921070336786568e-06,
"loss": 2.3021,
"step": 994
},
{
"epoch": 0.7231104651162791,
"grad_norm": 1.0601706121307002,
"learning_rate": 1.9117452641571934e-06,
"loss": 2.3253,
"step": 995
},
{
"epoch": 0.7238372093023255,
"grad_norm": 1.2357758164161554,
"learning_rate": 1.9024375265982386e-06,
"loss": 2.2932,
"step": 996
},
{
"epoch": 0.7245639534883721,
"grad_norm": 1.191064713641231,
"learning_rate": 1.893147176356131e-06,
"loss": 2.3273,
"step": 997
},
{
"epoch": 0.7252906976744186,
"grad_norm": 1.2241505412066669,
"learning_rate": 1.8838742655797053e-06,
"loss": 2.2872,
"step": 998
},
{
"epoch": 0.7260174418604651,
"grad_norm": 1.1096350635181376,
"learning_rate": 1.8746188463198983e-06,
"loss": 2.2533,
"step": 999
},
{
"epoch": 0.7267441860465116,
"grad_norm": 1.0810442272478031,
"learning_rate": 1.865380970529469e-06,
"loss": 2.2439,
"step": 1000
},
{
"epoch": 0.7274709302325582,
"grad_norm": 1.0611734339943502,
"learning_rate": 1.8561606900626938e-06,
"loss": 2.3247,
"step": 1001
},
{
"epoch": 0.7281976744186046,
"grad_norm": 1.339243616520072,
"learning_rate": 1.8469580566750911e-06,
"loss": 2.2827,
"step": 1002
},
{
"epoch": 0.7289244186046512,
"grad_norm": 1.1081621018474208,
"learning_rate": 1.8377731220231144e-06,
"loss": 2.3357,
"step": 1003
},
{
"epoch": 0.7296511627906976,
"grad_norm": 1.0403890016606065,
"learning_rate": 1.8286059376638748e-06,
"loss": 2.283,
"step": 1004
},
{
"epoch": 0.7303779069767442,
"grad_norm": 1.0630454548978983,
"learning_rate": 1.8194565550548477e-06,
"loss": 2.283,
"step": 1005
},
{
"epoch": 0.7311046511627907,
"grad_norm": 1.00735081057976,
"learning_rate": 1.810325025553578e-06,
"loss": 2.2705,
"step": 1006
},
{
"epoch": 0.7318313953488372,
"grad_norm": 1.0768984635152028,
"learning_rate": 1.8012114004174048e-06,
"loss": 2.2943,
"step": 1007
},
{
"epoch": 0.7325581395348837,
"grad_norm": 1.126696129099283,
"learning_rate": 1.7921157308031567e-06,
"loss": 2.3179,
"step": 1008
},
{
"epoch": 0.7332848837209303,
"grad_norm": 1.0160803840046975,
"learning_rate": 1.7830380677668836e-06,
"loss": 2.3228,
"step": 1009
},
{
"epoch": 0.7340116279069767,
"grad_norm": 1.0796523647513225,
"learning_rate": 1.7739784622635514e-06,
"loss": 2.2307,
"step": 1010
},
{
"epoch": 0.7347383720930233,
"grad_norm": 1.0462192247450826,
"learning_rate": 1.764936965146773e-06,
"loss": 2.2933,
"step": 1011
},
{
"epoch": 0.7354651162790697,
"grad_norm": 1.0331053296317911,
"learning_rate": 1.7559136271685079e-06,
"loss": 2.2922,
"step": 1012
},
{
"epoch": 0.7361918604651163,
"grad_norm": 0.9519678790569631,
"learning_rate": 1.746908498978791e-06,
"loss": 2.2542,
"step": 1013
},
{
"epoch": 0.7369186046511628,
"grad_norm": 1.0833415039732,
"learning_rate": 1.7379216311254339e-06,
"loss": 2.2388,
"step": 1014
},
{
"epoch": 0.7376453488372093,
"grad_norm": 1.0619106386944168,
"learning_rate": 1.7289530740537569e-06,
"loss": 2.2501,
"step": 1015
},
{
"epoch": 0.7383720930232558,
"grad_norm": 1.076815203497985,
"learning_rate": 1.72000287810629e-06,
"loss": 2.329,
"step": 1016
},
{
"epoch": 0.7390988372093024,
"grad_norm": 1.1080069971444473,
"learning_rate": 1.7110710935225055e-06,
"loss": 2.2878,
"step": 1017
},
{
"epoch": 0.7398255813953488,
"grad_norm": 1.1611753091102224,
"learning_rate": 1.7021577704385218e-06,
"loss": 2.2986,
"step": 1018
},
{
"epoch": 0.7405523255813954,
"grad_norm": 1.017226365288529,
"learning_rate": 1.6932629588868332e-06,
"loss": 2.3679,
"step": 1019
},
{
"epoch": 0.7412790697674418,
"grad_norm": 1.1889012837812845,
"learning_rate": 1.6843867087960252e-06,
"loss": 2.2629,
"step": 1020
},
{
"epoch": 0.7420058139534884,
"grad_norm": 1.059279124205967,
"learning_rate": 1.6755290699904881e-06,
"loss": 2.3013,
"step": 1021
},
{
"epoch": 0.7427325581395349,
"grad_norm": 1.0182359613614196,
"learning_rate": 1.6666900921901497e-06,
"loss": 2.3037,
"step": 1022
},
{
"epoch": 0.7434593023255814,
"grad_norm": 1.0912666761581755,
"learning_rate": 1.6578698250101828e-06,
"loss": 2.3042,
"step": 1023
},
{
"epoch": 0.7441860465116279,
"grad_norm": 1.2496008211453598,
"learning_rate": 1.6490683179607403e-06,
"loss": 2.2539,
"step": 1024
},
{
"epoch": 0.7449127906976745,
"grad_norm": 1.0449183892310232,
"learning_rate": 1.6402856204466611e-06,
"loss": 2.2656,
"step": 1025
},
{
"epoch": 0.7456395348837209,
"grad_norm": 1.0109361865487376,
"learning_rate": 1.6315217817672142e-06,
"loss": 2.3545,
"step": 1026
},
{
"epoch": 0.7463662790697675,
"grad_norm": 1.9222203824349149,
"learning_rate": 1.6227768511157976e-06,
"loss": 2.2807,
"step": 1027
},
{
"epoch": 0.747093023255814,
"grad_norm": 1.015484674656179,
"learning_rate": 1.6140508775796832e-06,
"loss": 2.2515,
"step": 1028
},
{
"epoch": 0.7478197674418605,
"grad_norm": 1.0079500670175678,
"learning_rate": 1.6053439101397257e-06,
"loss": 2.2943,
"step": 1029
},
{
"epoch": 0.748546511627907,
"grad_norm": 1.0661148978290145,
"learning_rate": 1.5966559976701e-06,
"loss": 2.3096,
"step": 1030
},
{
"epoch": 0.7492732558139535,
"grad_norm": 1.0568451868453557,
"learning_rate": 1.5879871889380155e-06,
"loss": 2.3256,
"step": 1031
},
{
"epoch": 0.75,
"grad_norm": 1.05938722775654,
"learning_rate": 1.5793375326034539e-06,
"loss": 2.3017,
"step": 1032
},
{
"epoch": 0.7507267441860465,
"grad_norm": 1.0107621685449764,
"learning_rate": 1.5707070772188843e-06,
"loss": 2.3019,
"step": 1033
},
{
"epoch": 0.751453488372093,
"grad_norm": 1.157820058838249,
"learning_rate": 1.5620958712290023e-06,
"loss": 2.245,
"step": 1034
},
{
"epoch": 0.7521802325581395,
"grad_norm": 0.9901772476736681,
"learning_rate": 1.5535039629704467e-06,
"loss": 2.2744,
"step": 1035
},
{
"epoch": 0.752906976744186,
"grad_norm": 1.4571372969257075,
"learning_rate": 1.5449314006715394e-06,
"loss": 2.3006,
"step": 1036
},
{
"epoch": 0.7536337209302325,
"grad_norm": 1.0727364544563056,
"learning_rate": 1.5363782324520033e-06,
"loss": 2.3023,
"step": 1037
},
{
"epoch": 0.7543604651162791,
"grad_norm": 1.051295726249027,
"learning_rate": 1.5278445063227038e-06,
"loss": 2.2896,
"step": 1038
},
{
"epoch": 0.7550872093023255,
"grad_norm": 1.1272815731891954,
"learning_rate": 1.5193302701853674e-06,
"loss": 2.2988,
"step": 1039
},
{
"epoch": 0.7558139534883721,
"grad_norm": 1.0502184723821115,
"learning_rate": 1.5108355718323236e-06,
"loss": 2.275,
"step": 1040
},
{
"epoch": 0.7565406976744186,
"grad_norm": 1.0583070742342378,
"learning_rate": 1.502360458946232e-06,
"loss": 2.2462,
"step": 1041
},
{
"epoch": 0.7572674418604651,
"grad_norm": 1.0060059414695133,
"learning_rate": 1.4939049790998095e-06,
"loss": 2.3153,
"step": 1042
},
{
"epoch": 0.7579941860465116,
"grad_norm": 1.158848162270058,
"learning_rate": 1.4854691797555753e-06,
"loss": 2.3023,
"step": 1043
},
{
"epoch": 0.7587209302325582,
"grad_norm": 1.0656483573077637,
"learning_rate": 1.4770531082655704e-06,
"loss": 2.3101,
"step": 1044
},
{
"epoch": 0.7594476744186046,
"grad_norm": 1.2449865044376767,
"learning_rate": 1.4686568118711054e-06,
"loss": 2.3007,
"step": 1045
},
{
"epoch": 0.7601744186046512,
"grad_norm": 1.0262544542085497,
"learning_rate": 1.4602803377024833e-06,
"loss": 2.3016,
"step": 1046
},
{
"epoch": 0.7609011627906976,
"grad_norm": 1.060293437802111,
"learning_rate": 1.451923732778745e-06,
"loss": 2.3011,
"step": 1047
},
{
"epoch": 0.7616279069767442,
"grad_norm": 1.28456037123509,
"learning_rate": 1.4435870440073968e-06,
"loss": 2.2614,
"step": 1048
},
{
"epoch": 0.7623546511627907,
"grad_norm": 1.0443395729679117,
"learning_rate": 1.435270318184156e-06,
"loss": 2.2908,
"step": 1049
},
{
"epoch": 0.7630813953488372,
"grad_norm": 1.0490096593491272,
"learning_rate": 1.4269736019926778e-06,
"loss": 2.2691,
"step": 1050
},
{
"epoch": 0.7638081395348837,
"grad_norm": 1.4580699017767131,
"learning_rate": 1.418696942004304e-06,
"loss": 2.276,
"step": 1051
},
{
"epoch": 0.7645348837209303,
"grad_norm": 1.0554824961663374,
"learning_rate": 1.410440384677791e-06,
"loss": 2.2815,
"step": 1052
},
{
"epoch": 0.7652616279069767,
"grad_norm": 1.0445976304458051,
"learning_rate": 1.4022039763590595e-06,
"loss": 2.3219,
"step": 1053
},
{
"epoch": 0.7659883720930233,
"grad_norm": 3.0454676377565466,
"learning_rate": 1.3939877632809279e-06,
"loss": 2.3021,
"step": 1054
},
{
"epoch": 0.7667151162790697,
"grad_norm": 1.0944528530649524,
"learning_rate": 1.3857917915628516e-06,
"loss": 2.2589,
"step": 1055
},
{
"epoch": 0.7674418604651163,
"grad_norm": 1.4188459634174018,
"learning_rate": 1.3776161072106703e-06,
"loss": 2.3215,
"step": 1056
},
{
"epoch": 0.7681686046511628,
"grad_norm": 1.0431343233534842,
"learning_rate": 1.369460756116342e-06,
"loss": 2.2651,
"step": 1057
},
{
"epoch": 0.7688953488372093,
"grad_norm": 1.1271722033081277,
"learning_rate": 1.3613257840576954e-06,
"loss": 2.2626,
"step": 1058
},
{
"epoch": 0.7696220930232558,
"grad_norm": 0.9807905284115528,
"learning_rate": 1.3532112366981598e-06,
"loss": 2.2429,
"step": 1059
},
{
"epoch": 0.7703488372093024,
"grad_norm": 1.0792539906176082,
"learning_rate": 1.3451171595865226e-06,
"loss": 2.2441,
"step": 1060
},
{
"epoch": 0.7710755813953488,
"grad_norm": 1.014599603945714,
"learning_rate": 1.3370435981566622e-06,
"loss": 2.3012,
"step": 1061
},
{
"epoch": 0.7718023255813954,
"grad_norm": 1.039382145166407,
"learning_rate": 1.3289905977273027e-06,
"loss": 2.2614,
"step": 1062
},
{
"epoch": 0.7725290697674418,
"grad_norm": 1.110162802251219,
"learning_rate": 1.3209582035017487e-06,
"loss": 2.3009,
"step": 1063
},
{
"epoch": 0.7732558139534884,
"grad_norm": 1.0268137021721464,
"learning_rate": 1.312946460567644e-06,
"loss": 2.2988,
"step": 1064
},
{
"epoch": 0.7739825581395349,
"grad_norm": 0.9991099915776216,
"learning_rate": 1.3049554138967052e-06,
"loss": 2.2574,
"step": 1065
},
{
"epoch": 0.7747093023255814,
"grad_norm": 1.096594932836595,
"learning_rate": 1.2969851083444834e-06,
"loss": 2.3029,
"step": 1066
},
{
"epoch": 0.7754360465116279,
"grad_norm": 1.0331471723360561,
"learning_rate": 1.2890355886500971e-06,
"loss": 2.3349,
"step": 1067
},
{
"epoch": 0.7761627906976745,
"grad_norm": 0.9989200321985331,
"learning_rate": 1.2811068994359992e-06,
"loss": 2.2743,
"step": 1068
},
{
"epoch": 0.7768895348837209,
"grad_norm": 1.0334842215623463,
"learning_rate": 1.273199085207706e-06,
"loss": 2.2318,
"step": 1069
},
{
"epoch": 0.7776162790697675,
"grad_norm": 1.1166669421482618,
"learning_rate": 1.2653121903535653e-06,
"loss": 2.3044,
"step": 1070
},
{
"epoch": 0.778343023255814,
"grad_norm": 1.0544628135207288,
"learning_rate": 1.257446259144494e-06,
"loss": 2.3597,
"step": 1071
},
{
"epoch": 0.7790697674418605,
"grad_norm": 1.0101121441444254,
"learning_rate": 1.2496013357337416e-06,
"loss": 2.2348,
"step": 1072
},
{
"epoch": 0.779796511627907,
"grad_norm": 1.0856552266206294,
"learning_rate": 1.2417774641566298e-06,
"loss": 2.2915,
"step": 1073
},
{
"epoch": 0.7805232558139535,
"grad_norm": 1.008999510135277,
"learning_rate": 1.233974688330315e-06,
"loss": 2.3178,
"step": 1074
},
{
"epoch": 0.78125,
"grad_norm": 0.9934991466078096,
"learning_rate": 1.2261930520535403e-06,
"loss": 2.2697,
"step": 1075
},
{
"epoch": 0.7819767441860465,
"grad_norm": 1.1302730062371027,
"learning_rate": 1.2184325990063822e-06,
"loss": 2.313,
"step": 1076
},
{
"epoch": 0.782703488372093,
"grad_norm": 1.0051323138300596,
"learning_rate": 1.210693372750017e-06,
"loss": 2.3108,
"step": 1077
},
{
"epoch": 0.7834302325581395,
"grad_norm": 1.3727245547649582,
"learning_rate": 1.202975416726464e-06,
"loss": 2.2932,
"step": 1078
},
{
"epoch": 0.784156976744186,
"grad_norm": 0.9962127663435182,
"learning_rate": 1.1952787742583549e-06,
"loss": 2.2928,
"step": 1079
},
{
"epoch": 0.7848837209302325,
"grad_norm": 2.918506835943175,
"learning_rate": 1.1876034885486764e-06,
"loss": 2.2867,
"step": 1080
},
{
"epoch": 0.7856104651162791,
"grad_norm": 1.002517901009031,
"learning_rate": 1.1799496026805413e-06,
"loss": 2.3079,
"step": 1081
},
{
"epoch": 0.7863372093023255,
"grad_norm": 1.0608837488625182,
"learning_rate": 1.1723171596169353e-06,
"loss": 2.2952,
"step": 1082
},
{
"epoch": 0.7870639534883721,
"grad_norm": 1.0495070164322962,
"learning_rate": 1.1647062022004845e-06,
"loss": 2.2838,
"step": 1083
},
{
"epoch": 0.7877906976744186,
"grad_norm": 0.9747665315176195,
"learning_rate": 1.157116773153208e-06,
"loss": 2.3577,
"step": 1084
},
{
"epoch": 0.7885174418604651,
"grad_norm": 0.995930671706204,
"learning_rate": 1.1495489150762851e-06,
"loss": 2.3267,
"step": 1085
},
{
"epoch": 0.7892441860465116,
"grad_norm": 1.0298574764276371,
"learning_rate": 1.1420026704498077e-06,
"loss": 2.2693,
"step": 1086
},
{
"epoch": 0.7899709302325582,
"grad_norm": 1.3353782375384569,
"learning_rate": 1.1344780816325512e-06,
"loss": 2.2941,
"step": 1087
},
{
"epoch": 0.7906976744186046,
"grad_norm": 1.0042512833949695,
"learning_rate": 1.1269751908617277e-06,
"loss": 2.2895,
"step": 1088
},
{
"epoch": 0.7914244186046512,
"grad_norm": 1.0159794876634873,
"learning_rate": 1.1194940402527566e-06,
"loss": 2.3337,
"step": 1089
},
{
"epoch": 0.7921511627906976,
"grad_norm": 1.7170986311408605,
"learning_rate": 1.112034671799025e-06,
"loss": 2.2687,
"step": 1090
},
{
"epoch": 0.7928779069767442,
"grad_norm": 1.0720092333346907,
"learning_rate": 1.1045971273716476e-06,
"loss": 2.2731,
"step": 1091
},
{
"epoch": 0.7936046511627907,
"grad_norm": 1.0901546781468832,
"learning_rate": 1.0971814487192429e-06,
"loss": 2.2649,
"step": 1092
},
{
"epoch": 0.7943313953488372,
"grad_norm": 1.0068271176988854,
"learning_rate": 1.089787677467683e-06,
"loss": 2.3544,
"step": 1093
},
{
"epoch": 0.7950581395348837,
"grad_norm": 1.143554056516148,
"learning_rate": 1.0824158551198783e-06,
"loss": 2.3312,
"step": 1094
},
{
"epoch": 0.7957848837209303,
"grad_norm": 1.0257503226390932,
"learning_rate": 1.075066023055527e-06,
"loss": 2.3307,
"step": 1095
},
{
"epoch": 0.7965116279069767,
"grad_norm": 1.0864265176648018,
"learning_rate": 1.0677382225308969e-06,
"loss": 2.3023,
"step": 1096
},
{
"epoch": 0.7972383720930233,
"grad_norm": 1.4336976428817494,
"learning_rate": 1.0604324946785826e-06,
"loss": 2.2618,
"step": 1097
},
{
"epoch": 0.7979651162790697,
"grad_norm": 1.0165625672732432,
"learning_rate": 1.0531488805072848e-06,
"loss": 2.2923,
"step": 1098
},
{
"epoch": 0.7986918604651163,
"grad_norm": 2.8438555195834705,
"learning_rate": 1.0458874209015708e-06,
"loss": 2.2758,
"step": 1099
},
{
"epoch": 0.7994186046511628,
"grad_norm": 1.011889649770043,
"learning_rate": 1.0386481566216532e-06,
"loss": 2.2674,
"step": 1100
},
{
"epoch": 0.8001453488372093,
"grad_norm": 0.9698911337581984,
"learning_rate": 1.0314311283031531e-06,
"loss": 2.261,
"step": 1101
},
{
"epoch": 0.8008720930232558,
"grad_norm": 1.0090528280127107,
"learning_rate": 1.0242363764568808e-06,
"loss": 2.2956,
"step": 1102
},
{
"epoch": 0.8015988372093024,
"grad_norm": 1.213795376548453,
"learning_rate": 1.0170639414685985e-06,
"loss": 2.2651,
"step": 1103
},
{
"epoch": 0.8023255813953488,
"grad_norm": 1.0633109792614825,
"learning_rate": 1.0099138635988026e-06,
"loss": 2.2556,
"step": 1104
},
{
"epoch": 0.8030523255813954,
"grad_norm": 1.1236794950217257,
"learning_rate": 1.0027861829824953e-06,
"loss": 2.3028,
"step": 1105
},
{
"epoch": 0.8037790697674418,
"grad_norm": 1.0179145252465904,
"learning_rate": 9.956809396289519e-07,
"loss": 2.2882,
"step": 1106
},
{
"epoch": 0.8045058139534884,
"grad_norm": 1.0440708463771915,
"learning_rate": 9.885981734215094e-07,
"loss": 2.2323,
"step": 1107
},
{
"epoch": 0.8052325581395349,
"grad_norm": 1.001838305148324,
"learning_rate": 9.815379241173295e-07,
"loss": 2.2875,
"step": 1108
},
{
"epoch": 0.8059593023255814,
"grad_norm": 1.0012314774080968,
"learning_rate": 9.745002313471847e-07,
"loss": 2.2784,
"step": 1109
},
{
"epoch": 0.8066860465116279,
"grad_norm": 1.3260411675767552,
"learning_rate": 9.67485134615232e-07,
"loss": 2.2714,
"step": 1110
},
{
"epoch": 0.8074127906976745,
"grad_norm": 0.9786969103850816,
"learning_rate": 9.60492673298794e-07,
"loss": 2.2664,
"step": 1111
},
{
"epoch": 0.8081395348837209,
"grad_norm": 1.5073197015028201,
"learning_rate": 9.535228866481295e-07,
"loss": 2.2279,
"step": 1112
},
{
"epoch": 0.8088662790697675,
"grad_norm": 1.056086008875671,
"learning_rate": 9.465758137862264e-07,
"loss": 2.2881,
"step": 1113
},
{
"epoch": 0.809593023255814,
"grad_norm": 1.0809504924319808,
"learning_rate": 9.396514937085682e-07,
"loss": 2.3032,
"step": 1114
},
{
"epoch": 0.8103197674418605,
"grad_norm": 1.2486681539195614,
"learning_rate": 9.327499652829292e-07,
"loss": 2.3155,
"step": 1115
},
{
"epoch": 0.811046511627907,
"grad_norm": 1.0213292959789695,
"learning_rate": 9.258712672491416e-07,
"loss": 2.2899,
"step": 1116
},
{
"epoch": 0.8117732558139535,
"grad_norm": 1.0461554742483217,
"learning_rate": 9.190154382188921e-07,
"loss": 2.3477,
"step": 1117
},
{
"epoch": 0.8125,
"grad_norm": 1.0675949468870904,
"learning_rate": 9.121825166754927e-07,
"loss": 2.2543,
"step": 1118
},
{
"epoch": 0.8132267441860465,
"grad_norm": 2.530653282570627,
"learning_rate": 9.053725409736752e-07,
"loss": 2.3301,
"step": 1119
},
{
"epoch": 0.813953488372093,
"grad_norm": 0.9984512358752262,
"learning_rate": 8.98585549339368e-07,
"loss": 2.3042,
"step": 1120
},
{
"epoch": 0.8146802325581395,
"grad_norm": 1.0213185627968009,
"learning_rate": 8.918215798694879e-07,
"loss": 2.2799,
"step": 1121
},
{
"epoch": 0.815406976744186,
"grad_norm": 1.0613752567095163,
"learning_rate": 8.850806705317183e-07,
"loss": 2.2999,
"step": 1122
},
{
"epoch": 0.8161337209302325,
"grad_norm": 1.0310489044310347,
"learning_rate": 8.783628591643056e-07,
"loss": 2.3235,
"step": 1123
},
{
"epoch": 0.8168604651162791,
"grad_norm": 4.283745359968558,
"learning_rate": 8.716681834758411e-07,
"loss": 2.3211,
"step": 1124
},
{
"epoch": 0.8175872093023255,
"grad_norm": 1.331778348277442,
"learning_rate": 8.649966810450472e-07,
"loss": 2.341,
"step": 1125
},
{
"epoch": 0.8183139534883721,
"grad_norm": 1.0415010268575644,
"learning_rate": 8.583483893205746e-07,
"loss": 2.2896,
"step": 1126
},
{
"epoch": 0.8190406976744186,
"grad_norm": 1.006456220083455,
"learning_rate": 8.517233456207819e-07,
"loss": 2.3029,
"step": 1127
},
{
"epoch": 0.8197674418604651,
"grad_norm": 1.6510497568308151,
"learning_rate": 8.451215871335355e-07,
"loss": 2.308,
"step": 1128
},
{
"epoch": 0.8204941860465116,
"grad_norm": 1.0156351391245906,
"learning_rate": 8.38543150915993e-07,
"loss": 2.3161,
"step": 1129
},
{
"epoch": 0.8212209302325582,
"grad_norm": 1.0120097967166009,
"learning_rate": 8.31988073894403e-07,
"loss": 2.2806,
"step": 1130
},
{
"epoch": 0.8219476744186046,
"grad_norm": 0.9976328982359449,
"learning_rate": 8.254563928638892e-07,
"loss": 2.356,
"step": 1131
},
{
"epoch": 0.8226744186046512,
"grad_norm": 0.9964751712960173,
"learning_rate": 8.189481444882524e-07,
"loss": 2.248,
"step": 1132
},
{
"epoch": 0.8234011627906976,
"grad_norm": 1.233431757610957,
"learning_rate": 8.124633652997571e-07,
"loss": 2.2668,
"step": 1133
},
{
"epoch": 0.8241279069767442,
"grad_norm": 1.0780894801654444,
"learning_rate": 8.060020916989331e-07,
"loss": 2.303,
"step": 1134
},
{
"epoch": 0.8248546511627907,
"grad_norm": 1.122308399785914,
"learning_rate": 7.995643599543645e-07,
"loss": 2.3352,
"step": 1135
},
{
"epoch": 0.8255813953488372,
"grad_norm": 1.0627182539952464,
"learning_rate": 7.931502062024949e-07,
"loss": 2.275,
"step": 1136
},
{
"epoch": 0.8263081395348837,
"grad_norm": 1.126076955118742,
"learning_rate": 7.86759666447412e-07,
"loss": 2.2813,
"step": 1137
},
{
"epoch": 0.8270348837209303,
"grad_norm": 1.4854164415590436,
"learning_rate": 7.803927765606595e-07,
"loss": 2.3282,
"step": 1138
},
{
"epoch": 0.8277616279069767,
"grad_norm": 1.0570569678350674,
"learning_rate": 7.740495722810271e-07,
"loss": 2.3106,
"step": 1139
},
{
"epoch": 0.8284883720930233,
"grad_norm": 0.9347752540270663,
"learning_rate": 7.677300892143485e-07,
"loss": 2.2337,
"step": 1140
},
{
"epoch": 0.8292151162790697,
"grad_norm": 0.9579877274387915,
"learning_rate": 7.614343628333104e-07,
"loss": 2.2958,
"step": 1141
},
{
"epoch": 0.8299418604651163,
"grad_norm": 1.1249688280208123,
"learning_rate": 7.55162428477243e-07,
"loss": 2.2656,
"step": 1142
},
{
"epoch": 0.8306686046511628,
"grad_norm": 0.9756167229918767,
"learning_rate": 7.489143213519301e-07,
"loss": 2.2924,
"step": 1143
},
{
"epoch": 0.8313953488372093,
"grad_norm": 1.044572083475489,
"learning_rate": 7.426900765294043e-07,
"loss": 2.2677,
"step": 1144
},
{
"epoch": 0.8321220930232558,
"grad_norm": 1.379796567891184,
"learning_rate": 7.364897289477585e-07,
"loss": 2.3265,
"step": 1145
},
{
"epoch": 0.8328488372093024,
"grad_norm": 1.0695146040268584,
"learning_rate": 7.303133134109391e-07,
"loss": 2.3288,
"step": 1146
},
{
"epoch": 0.8335755813953488,
"grad_norm": 1.016987223129461,
"learning_rate": 7.241608645885629e-07,
"loss": 2.2986,
"step": 1147
},
{
"epoch": 0.8343023255813954,
"grad_norm": 1.1916192934614125,
"learning_rate": 7.180324170157094e-07,
"loss": 2.2874,
"step": 1148
},
{
"epoch": 0.8350290697674418,
"grad_norm": 1.0839138661442358,
"learning_rate": 7.119280050927407e-07,
"loss": 2.265,
"step": 1149
},
{
"epoch": 0.8357558139534884,
"grad_norm": 1.0304987400941559,
"learning_rate": 7.058476630850935e-07,
"loss": 2.2685,
"step": 1150
},
{
"epoch": 0.8364825581395349,
"grad_norm": 1.0321949762195342,
"learning_rate": 6.997914251231036e-07,
"loss": 2.3048,
"step": 1151
},
{
"epoch": 0.8372093023255814,
"grad_norm": 1.0154567754879218,
"learning_rate": 6.937593252017983e-07,
"loss": 2.2805,
"step": 1152
},
{
"epoch": 0.8379360465116279,
"grad_norm": 1.0838324246811337,
"learning_rate": 6.87751397180716e-07,
"loss": 2.2884,
"step": 1153
},
{
"epoch": 0.8386627906976745,
"grad_norm": 1.0897256752726174,
"learning_rate": 6.817676747837104e-07,
"loss": 2.3718,
"step": 1154
},
{
"epoch": 0.8393895348837209,
"grad_norm": 1.2095702202073741,
"learning_rate": 6.758081915987669e-07,
"loss": 2.2704,
"step": 1155
},
{
"epoch": 0.8401162790697675,
"grad_norm": 1.0032802281153688,
"learning_rate": 6.698729810778065e-07,
"loss": 2.3228,
"step": 1156
},
{
"epoch": 0.840843023255814,
"grad_norm": 1.2006946309581232,
"learning_rate": 6.639620765365074e-07,
"loss": 2.3023,
"step": 1157
},
{
"epoch": 0.8415697674418605,
"grad_norm": 1.09168549344349,
"learning_rate": 6.580755111541076e-07,
"loss": 2.346,
"step": 1158
},
{
"epoch": 0.842296511627907,
"grad_norm": 1.1606076057102177,
"learning_rate": 6.522133179732271e-07,
"loss": 2.371,
"step": 1159
},
{
"epoch": 0.8430232558139535,
"grad_norm": 0.9745185257832802,
"learning_rate": 6.463755298996799e-07,
"loss": 2.241,
"step": 1160
},
{
"epoch": 0.84375,
"grad_norm": 0.9926775523213874,
"learning_rate": 6.405621797022848e-07,
"loss": 2.2971,
"step": 1161
},
{
"epoch": 0.8444767441860465,
"grad_norm": 1.0589270188129587,
"learning_rate": 6.347733000126899e-07,
"loss": 2.2864,
"step": 1162
},
{
"epoch": 0.845203488372093,
"grad_norm": 1.1986749871451796,
"learning_rate": 6.290089233251811e-07,
"loss": 2.2959,
"step": 1163
},
{
"epoch": 0.8459302325581395,
"grad_norm": 0.9904095687889044,
"learning_rate": 6.232690819965065e-07,
"loss": 2.2937,
"step": 1164
},
{
"epoch": 0.846656976744186,
"grad_norm": 1.000092587187146,
"learning_rate": 6.175538082456883e-07,
"loss": 2.2807,
"step": 1165
},
{
"epoch": 0.8473837209302325,
"grad_norm": 1.6370550141375801,
"learning_rate": 6.118631341538489e-07,
"loss": 2.2839,
"step": 1166
},
{
"epoch": 0.8481104651162791,
"grad_norm": 1.007731213927358,
"learning_rate": 6.061970916640236e-07,
"loss": 2.2792,
"step": 1167
},
{
"epoch": 0.8488372093023255,
"grad_norm": 1.052835634775134,
"learning_rate": 6.005557125809896e-07,
"loss": 2.2941,
"step": 1168
},
{
"epoch": 0.8495639534883721,
"grad_norm": 1.2895324602216356,
"learning_rate": 5.949390285710777e-07,
"loss": 2.3294,
"step": 1169
},
{
"epoch": 0.8502906976744186,
"grad_norm": 1.1015498862150495,
"learning_rate": 5.893470711620036e-07,
"loss": 2.2882,
"step": 1170
},
{
"epoch": 0.8510174418604651,
"grad_norm": 1.0003788574406323,
"learning_rate": 5.837798717426846e-07,
"loss": 2.2764,
"step": 1171
},
{
"epoch": 0.8517441860465116,
"grad_norm": 1.024968554027419,
"learning_rate": 5.782374615630682e-07,
"loss": 2.228,
"step": 1172
},
{
"epoch": 0.8524709302325582,
"grad_norm": 1.0094305384280167,
"learning_rate": 5.727198717339511e-07,
"loss": 2.2914,
"step": 1173
},
{
"epoch": 0.8531976744186046,
"grad_norm": 1.0110519070659092,
"learning_rate": 5.672271332268098e-07,
"loss": 2.3431,
"step": 1174
},
{
"epoch": 0.8539244186046512,
"grad_norm": 1.0385439058435382,
"learning_rate": 5.617592768736269e-07,
"loss": 2.3103,
"step": 1175
},
{
"epoch": 0.8546511627906976,
"grad_norm": 1.0511462342322628,
"learning_rate": 5.563163333667098e-07,
"loss": 2.2852,
"step": 1176
},
{
"epoch": 0.8553779069767442,
"grad_norm": 1.0132897307989803,
"learning_rate": 5.508983332585316e-07,
"loss": 2.2611,
"step": 1177
},
{
"epoch": 0.8561046511627907,
"grad_norm": 0.9985229317455512,
"learning_rate": 5.455053069615456e-07,
"loss": 2.3396,
"step": 1178
},
{
"epoch": 0.8568313953488372,
"grad_norm": 1.2719782958329457,
"learning_rate": 5.401372847480285e-07,
"loss": 2.2947,
"step": 1179
},
{
"epoch": 0.8575581395348837,
"grad_norm": 1.1276835003016656,
"learning_rate": 5.347942967498965e-07,
"loss": 2.3052,
"step": 1180
},
{
"epoch": 0.8582848837209303,
"grad_norm": 4.384284810043298,
"learning_rate": 5.294763729585484e-07,
"loss": 2.2757,
"step": 1181
},
{
"epoch": 0.8590116279069767,
"grad_norm": 1.2941565270254776,
"learning_rate": 5.241835432246888e-07,
"loss": 2.2744,
"step": 1182
},
{
"epoch": 0.8597383720930233,
"grad_norm": 0.9811407369818785,
"learning_rate": 5.18915837258166e-07,
"loss": 2.2824,
"step": 1183
},
{
"epoch": 0.8604651162790697,
"grad_norm": 0.978937747602062,
"learning_rate": 5.136732846278003e-07,
"loss": 2.2379,
"step": 1184
},
{
"epoch": 0.8611918604651163,
"grad_norm": 1.1398240380059452,
"learning_rate": 5.084559147612244e-07,
"loss": 2.3125,
"step": 1185
},
{
"epoch": 0.8619186046511628,
"grad_norm": 0.9982485781718631,
"learning_rate": 5.032637569447091e-07,
"loss": 2.3173,
"step": 1186
},
{
"epoch": 0.8626453488372093,
"grad_norm": 1.1061089964328414,
"learning_rate": 4.980968403230097e-07,
"loss": 2.301,
"step": 1187
},
{
"epoch": 0.8633720930232558,
"grad_norm": 1.0805494456177815,
"learning_rate": 4.929551938991945e-07,
"loss": 2.2888,
"step": 1188
},
{
"epoch": 0.8640988372093024,
"grad_norm": 1.1308532612444147,
"learning_rate": 4.87838846534483e-07,
"loss": 2.2963,
"step": 1189
},
{
"epoch": 0.8648255813953488,
"grad_norm": 1.1810215565297821,
"learning_rate": 4.827478269480895e-07,
"loss": 2.2896,
"step": 1190
},
{
"epoch": 0.8655523255813954,
"grad_norm": 1.0037351932727745,
"learning_rate": 4.776821637170525e-07,
"loss": 2.3178,
"step": 1191
},
{
"epoch": 0.8662790697674418,
"grad_norm": 1.1742864207537889,
"learning_rate": 4.726418852760839e-07,
"loss": 2.3462,
"step": 1192
},
{
"epoch": 0.8670058139534884,
"grad_norm": 1.0390414759278597,
"learning_rate": 4.6762701991740434e-07,
"loss": 2.3178,
"step": 1193
},
{
"epoch": 0.8677325581395349,
"grad_norm": 1.0152117439648225,
"learning_rate": 4.626375957905821e-07,
"loss": 2.3091,
"step": 1194
},
{
"epoch": 0.8684593023255814,
"grad_norm": 1.0125484137726066,
"learning_rate": 4.576736409023813e-07,
"loss": 2.295,
"step": 1195
},
{
"epoch": 0.8691860465116279,
"grad_norm": 1.0648811208027753,
"learning_rate": 4.5273518311660103e-07,
"loss": 2.311,
"step": 1196
},
{
"epoch": 0.8699127906976745,
"grad_norm": 0.9969625476223827,
"learning_rate": 4.4782225015391754e-07,
"loss": 2.3062,
"step": 1197
},
{
"epoch": 0.8706395348837209,
"grad_norm": 1.191266036217783,
"learning_rate": 4.429348695917329e-07,
"loss": 2.2864,
"step": 1198
},
{
"epoch": 0.8713662790697675,
"grad_norm": 1.054591606203741,
"learning_rate": 4.3807306886401555e-07,
"loss": 2.2381,
"step": 1199
},
{
"epoch": 0.872093023255814,
"grad_norm": 1.0919505023664373,
"learning_rate": 4.3323687526115045e-07,
"loss": 2.258,
"step": 1200
},
{
"epoch": 0.8728197674418605,
"grad_norm": 1.087148546984066,
"learning_rate": 4.284263159297819e-07,
"loss": 2.3222,
"step": 1201
},
{
"epoch": 0.873546511627907,
"grad_norm": 6.580439059174034,
"learning_rate": 4.2364141787266613e-07,
"loss": 2.2518,
"step": 1202
},
{
"epoch": 0.8742732558139535,
"grad_norm": 1.0455521942127206,
"learning_rate": 4.1888220794851386e-07,
"loss": 2.3222,
"step": 1203
},
{
"epoch": 0.875,
"grad_norm": 1.0342987091911544,
"learning_rate": 4.141487128718452e-07,
"loss": 2.2385,
"step": 1204
},
{
"epoch": 0.8757267441860465,
"grad_norm": 1.1782647809260987,
"learning_rate": 4.0944095921283347e-07,
"loss": 2.3029,
"step": 1205
},
{
"epoch": 0.876453488372093,
"grad_norm": 1.0105311825679908,
"learning_rate": 4.0475897339716466e-07,
"loss": 2.2689,
"step": 1206
},
{
"epoch": 0.8771802325581395,
"grad_norm": 1.1756077441855846,
"learning_rate": 4.001027817058789e-07,
"loss": 2.2578,
"step": 1207
},
{
"epoch": 0.877906976744186,
"grad_norm": 1.1994916473407944,
"learning_rate": 3.9547241027523164e-07,
"loss": 2.359,
"step": 1208
},
{
"epoch": 0.8786337209302325,
"grad_norm": 1.0065299853456233,
"learning_rate": 3.908678850965425e-07,
"loss": 2.2853,
"step": 1209
},
{
"epoch": 0.8793604651162791,
"grad_norm": 0.995293307148518,
"learning_rate": 3.862892320160483e-07,
"loss": 2.2858,
"step": 1210
},
{
"epoch": 0.8800872093023255,
"grad_norm": 1.0258935554168493,
"learning_rate": 3.8173647673476366e-07,
"loss": 2.3548,
"step": 1211
},
{
"epoch": 0.8808139534883721,
"grad_norm": 0.9956512980655323,
"learning_rate": 3.7720964480832847e-07,
"loss": 2.2886,
"step": 1212
},
{
"epoch": 0.8815406976744186,
"grad_norm": 1.0464019346352542,
"learning_rate": 3.727087616468739e-07,
"loss": 2.2975,
"step": 1213
},
{
"epoch": 0.8822674418604651,
"grad_norm": 1.0261367720153103,
"learning_rate": 3.682338525148699e-07,
"loss": 2.2455,
"step": 1214
},
{
"epoch": 0.8829941860465116,
"grad_norm": 0.9784631495637669,
"learning_rate": 3.6378494253099307e-07,
"loss": 2.2994,
"step": 1215
},
{
"epoch": 0.8837209302325582,
"grad_norm": 1.0561243830461962,
"learning_rate": 3.5936205666797675e-07,
"loss": 2.3172,
"step": 1216
},
{
"epoch": 0.8844476744186046,
"grad_norm": 0.9866167353173949,
"learning_rate": 3.549652197524783e-07,
"loss": 2.2568,
"step": 1217
},
{
"epoch": 0.8851744186046512,
"grad_norm": 1.7694336587594428,
"learning_rate": 3.505944564649344e-07,
"loss": 2.3109,
"step": 1218
},
{
"epoch": 0.8859011627906976,
"grad_norm": 1.0237927443574126,
"learning_rate": 3.462497913394258e-07,
"loss": 2.2819,
"step": 1219
},
{
"epoch": 0.8866279069767442,
"grad_norm": 1.0797737719641145,
"learning_rate": 3.419312487635362e-07,
"loss": 2.2423,
"step": 1220
},
{
"epoch": 0.8873546511627907,
"grad_norm": 1.0965412032340838,
"learning_rate": 3.3763885297822153e-07,
"loss": 2.294,
"step": 1221
},
{
"epoch": 0.8880813953488372,
"grad_norm": 1.1707762135194628,
"learning_rate": 3.333726280776656e-07,
"loss": 2.3118,
"step": 1222
},
{
"epoch": 0.8888081395348837,
"grad_norm": 1.312947277182204,
"learning_rate": 3.2913259800915196e-07,
"loss": 2.259,
"step": 1223
},
{
"epoch": 0.8895348837209303,
"grad_norm": 1.0501531220783618,
"learning_rate": 3.2491878657292643e-07,
"loss": 2.2494,
"step": 1224
},
{
"epoch": 0.8902616279069767,
"grad_norm": 0.9978642932087061,
"learning_rate": 3.2073121742206117e-07,
"loss": 2.2676,
"step": 1225
},
{
"epoch": 0.8909883720930233,
"grad_norm": 1.0219935225349466,
"learning_rate": 3.165699140623285e-07,
"loss": 2.2609,
"step": 1226
},
{
"epoch": 0.8917151162790697,
"grad_norm": 1.0271766170690235,
"learning_rate": 3.1243489985206097e-07,
"loss": 2.2444,
"step": 1227
},
{
"epoch": 0.8924418604651163,
"grad_norm": 1.2162058838576948,
"learning_rate": 3.0832619800202746e-07,
"loss": 2.3607,
"step": 1228
},
{
"epoch": 0.8931686046511628,
"grad_norm": 0.9647661066514175,
"learning_rate": 3.0424383157529716e-07,
"loss": 2.2938,
"step": 1229
},
{
"epoch": 0.8938953488372093,
"grad_norm": 1.0359666318241527,
"learning_rate": 3.001878234871147e-07,
"loss": 2.2933,
"step": 1230
},
{
"epoch": 0.8946220930232558,
"grad_norm": 1.0260708734909725,
"learning_rate": 2.961581965047672e-07,
"loss": 2.2944,
"step": 1231
},
{
"epoch": 0.8953488372093024,
"grad_norm": 1.9596558246449935,
"learning_rate": 2.921549732474599e-07,
"loss": 2.3171,
"step": 1232
},
{
"epoch": 0.8960755813953488,
"grad_norm": 1.1149831837749224,
"learning_rate": 2.8817817618618846e-07,
"loss": 2.3445,
"step": 1233
},
{
"epoch": 0.8968023255813954,
"grad_norm": 1.0325460985272938,
"learning_rate": 2.842278276436128e-07,
"loss": 2.2966,
"step": 1234
},
{
"epoch": 0.8975290697674418,
"grad_norm": 1.0390227024813596,
"learning_rate": 2.803039497939281e-07,
"loss": 2.26,
"step": 1235
},
{
"epoch": 0.8982558139534884,
"grad_norm": 1.091918329127377,
"learning_rate": 2.7640656466274785e-07,
"loss": 2.3169,
"step": 1236
},
{
"epoch": 0.8989825581395349,
"grad_norm": 1.0489517864380382,
"learning_rate": 2.7253569412697244e-07,
"loss": 2.2924,
"step": 1237
},
{
"epoch": 0.8997093023255814,
"grad_norm": 1.1270113218465792,
"learning_rate": 2.686913599146723e-07,
"loss": 2.305,
"step": 1238
},
{
"epoch": 0.9004360465116279,
"grad_norm": 1.0077746037473514,
"learning_rate": 2.648735836049615e-07,
"loss": 2.2416,
"step": 1239
},
{
"epoch": 0.9011627906976745,
"grad_norm": 1.0534517963417607,
"learning_rate": 2.6108238662788057e-07,
"loss": 2.2928,
"step": 1240
},
{
"epoch": 0.9018895348837209,
"grad_norm": 1.2397150258831577,
"learning_rate": 2.573177902642726e-07,
"loss": 2.3184,
"step": 1241
},
{
"epoch": 0.9026162790697675,
"grad_norm": 1.1152532284797432,
"learning_rate": 2.5357981564566647e-07,
"loss": 2.3111,
"step": 1242
},
{
"epoch": 0.903343023255814,
"grad_norm": 1.015507101048782,
"learning_rate": 2.4986848375415653e-07,
"loss": 2.2436,
"step": 1243
},
{
"epoch": 0.9040697674418605,
"grad_norm": 1.0734114420916194,
"learning_rate": 2.4618381542228565e-07,
"loss": 2.3737,
"step": 1244
},
{
"epoch": 0.904796511627907,
"grad_norm": 1.346656207028202,
"learning_rate": 2.4252583133292927e-07,
"loss": 2.2639,
"step": 1245
},
{
"epoch": 0.9055232558139535,
"grad_norm": 1.0354149154983616,
"learning_rate": 2.3889455201917655e-07,
"loss": 2.3014,
"step": 1246
},
{
"epoch": 0.90625,
"grad_norm": 1.0294255073427006,
"learning_rate": 2.3528999786421758e-07,
"loss": 2.329,
"step": 1247
},
{
"epoch": 0.9069767441860465,
"grad_norm": 1.0697056028935716,
"learning_rate": 2.3171218910122695e-07,
"loss": 2.3277,
"step": 1248
},
{
"epoch": 0.907703488372093,
"grad_norm": 1.0427212290454575,
"learning_rate": 2.2816114581325377e-07,
"loss": 2.3078,
"step": 1249
},
{
"epoch": 0.9084302325581395,
"grad_norm": 1.0376797552205965,
"learning_rate": 2.2463688793310345e-07,
"loss": 2.3154,
"step": 1250
},
{
"epoch": 0.909156976744186,
"grad_norm": 1.0936357024502716,
"learning_rate": 2.2113943524323167e-07,
"loss": 2.3418,
"step": 1251
},
{
"epoch": 0.9098837209302325,
"grad_norm": 1.0037611206384802,
"learning_rate": 2.1766880737562833e-07,
"loss": 2.2743,
"step": 1252
},
{
"epoch": 0.9106104651162791,
"grad_norm": 1.086635907292096,
"learning_rate": 2.1422502381171163e-07,
"loss": 2.3162,
"step": 1253
},
{
"epoch": 0.9113372093023255,
"grad_norm": 1.1145643232811908,
"learning_rate": 2.1080810388221406e-07,
"loss": 2.2704,
"step": 1254
},
{
"epoch": 0.9120639534883721,
"grad_norm": 1.0598433068070068,
"learning_rate": 2.0741806676707887e-07,
"loss": 2.3319,
"step": 1255
},
{
"epoch": 0.9127906976744186,
"grad_norm": 1.3127144244086706,
"learning_rate": 2.0405493149534828e-07,
"loss": 2.3116,
"step": 1256
},
{
"epoch": 0.9135174418604651,
"grad_norm": 1.269702023739857,
"learning_rate": 2.007187169450603e-07,
"loss": 2.2136,
"step": 1257
},
{
"epoch": 0.9142441860465116,
"grad_norm": 0.9803876917909695,
"learning_rate": 1.9740944184313882e-07,
"loss": 2.2767,
"step": 1258
},
{
"epoch": 0.9149709302325582,
"grad_norm": 0.9509347439890067,
"learning_rate": 1.941271247652915e-07,
"loss": 2.2829,
"step": 1259
},
{
"epoch": 0.9156976744186046,
"grad_norm": 0.9886083066522969,
"learning_rate": 1.908717841359048e-07,
"loss": 2.2567,
"step": 1260
},
{
"epoch": 0.9164244186046512,
"grad_norm": 1.128556701776484,
"learning_rate": 1.8764343822793962e-07,
"loss": 2.3189,
"step": 1261
},
{
"epoch": 0.9171511627906976,
"grad_norm": 1.0214841774757928,
"learning_rate": 1.8444210516283035e-07,
"loss": 2.2686,
"step": 1262
},
{
"epoch": 0.9178779069767442,
"grad_norm": 1.0384165232113651,
"learning_rate": 1.8126780291038037e-07,
"loss": 2.2953,
"step": 1263
},
{
"epoch": 0.9186046511627907,
"grad_norm": 1.0239892896324345,
"learning_rate": 1.7812054928866617e-07,
"loss": 2.2896,
"step": 1264
},
{
"epoch": 0.9193313953488372,
"grad_norm": 1.0049155639014276,
"learning_rate": 1.7500036196392956e-07,
"loss": 2.2505,
"step": 1265
},
{
"epoch": 0.9200581395348837,
"grad_norm": 1.0166829246084337,
"learning_rate": 1.7190725845048827e-07,
"loss": 2.2876,
"step": 1266
},
{
"epoch": 0.9207848837209303,
"grad_norm": 1.0121063072615553,
"learning_rate": 1.688412561106284e-07,
"loss": 2.2638,
"step": 1267
},
{
"epoch": 0.9215116279069767,
"grad_norm": 1.0325016280932235,
"learning_rate": 1.6580237215451378e-07,
"loss": 2.2858,
"step": 1268
},
{
"epoch": 0.9222383720930233,
"grad_norm": 1.1015868462586509,
"learning_rate": 1.6279062364008446e-07,
"loss": 2.3058,
"step": 1269
},
{
"epoch": 0.9229651162790697,
"grad_norm": 1.0513722612177225,
"learning_rate": 1.5980602747296513e-07,
"loss": 2.2703,
"step": 1270
},
{
"epoch": 0.9236918604651163,
"grad_norm": 1.0097455230074164,
"learning_rate": 1.5684860040636573e-07,
"loss": 2.3194,
"step": 1271
},
{
"epoch": 0.9244186046511628,
"grad_norm": 1.0351656666175164,
"learning_rate": 1.5391835904099316e-07,
"loss": 2.3516,
"step": 1272
},
{
"epoch": 0.9251453488372093,
"grad_norm": 1.006940226866894,
"learning_rate": 1.510153198249531e-07,
"loss": 2.2536,
"step": 1273
},
{
"epoch": 0.9258720930232558,
"grad_norm": 1.0120832517149825,
"learning_rate": 1.4813949905365833e-07,
"loss": 2.2966,
"step": 1274
},
{
"epoch": 0.9265988372093024,
"grad_norm": 1.04037739337114,
"learning_rate": 1.4529091286973994e-07,
"loss": 2.3211,
"step": 1275
},
{
"epoch": 0.9273255813953488,
"grad_norm": 1.0923715057949517,
"learning_rate": 1.424695772629553e-07,
"loss": 2.286,
"step": 1276
},
{
"epoch": 0.9280523255813954,
"grad_norm": 1.0745241746743175,
"learning_rate": 1.3967550807009677e-07,
"loss": 2.3102,
"step": 1277
},
{
"epoch": 0.9287790697674418,
"grad_norm": 1.2680690659745402,
"learning_rate": 1.3690872097490481e-07,
"loss": 2.3097,
"step": 1278
},
{
"epoch": 0.9295058139534884,
"grad_norm": 0.98738554433098,
"learning_rate": 1.3416923150798123e-07,
"loss": 2.3177,
"step": 1279
},
{
"epoch": 0.9302325581395349,
"grad_norm": 1.7610982138078959,
"learning_rate": 1.3145705504669592e-07,
"loss": 2.2221,
"step": 1280
},
{
"epoch": 0.9309593023255814,
"grad_norm": 1.0476925479595647,
"learning_rate": 1.2877220681510927e-07,
"loss": 2.273,
"step": 1281
},
{
"epoch": 0.9316860465116279,
"grad_norm": 1.0107799113760905,
"learning_rate": 1.261147018838782e-07,
"loss": 2.2488,
"step": 1282
},
{
"epoch": 0.9324127906976745,
"grad_norm": 1.1192751459025567,
"learning_rate": 1.2348455517017855e-07,
"loss": 2.2751,
"step": 1283
},
{
"epoch": 0.9331395348837209,
"grad_norm": 0.9712048469205136,
"learning_rate": 1.208817814376162e-07,
"loss": 2.2985,
"step": 1284
},
{
"epoch": 0.9338662790697675,
"grad_norm": 0.9681812041910587,
"learning_rate": 1.1830639529614774e-07,
"loss": 2.2898,
"step": 1285
},
{
"epoch": 0.934593023255814,
"grad_norm": 1.031175408747568,
"learning_rate": 1.157584112019966e-07,
"loss": 2.2973,
"step": 1286
},
{
"epoch": 0.9353197674418605,
"grad_norm": 1.080255118494038,
"learning_rate": 1.1323784345757205e-07,
"loss": 2.2881,
"step": 1287
},
{
"epoch": 0.936046511627907,
"grad_norm": 1.1584490527865852,
"learning_rate": 1.1074470621138866e-07,
"loss": 2.2674,
"step": 1288
},
{
"epoch": 0.9367732558139535,
"grad_norm": 0.9893069974580888,
"learning_rate": 1.0827901345798919e-07,
"loss": 2.3274,
"step": 1289
},
{
"epoch": 0.9375,
"grad_norm": 1.1609278902238676,
"learning_rate": 1.0584077903786238e-07,
"loss": 2.2954,
"step": 1290
},
{
"epoch": 0.9382267441860465,
"grad_norm": 1.1214652879688172,
"learning_rate": 1.0343001663736918e-07,
"loss": 2.2437,
"step": 1291
},
{
"epoch": 0.938953488372093,
"grad_norm": 1.0314479692365157,
"learning_rate": 1.0104673978866164e-07,
"loss": 2.3459,
"step": 1292
},
{
"epoch": 0.9396802325581395,
"grad_norm": 1.0202246090098164,
"learning_rate": 9.869096186961025e-08,
"loss": 2.3238,
"step": 1293
},
{
"epoch": 0.940406976744186,
"grad_norm": 1.735524228594828,
"learning_rate": 9.636269610372895e-08,
"loss": 2.2979,
"step": 1294
},
{
"epoch": 0.9411337209302325,
"grad_norm": 1.0231462789990604,
"learning_rate": 9.406195556009745e-08,
"loss": 2.3135,
"step": 1295
},
{
"epoch": 0.9418604651162791,
"grad_norm": 1.0387601752172189,
"learning_rate": 9.178875315329183e-08,
"loss": 2.2836,
"step": 1296
},
{
"epoch": 0.9425872093023255,
"grad_norm": 1.0455127776642161,
"learning_rate": 8.954310164331015e-08,
"loss": 2.3477,
"step": 1297
},
{
"epoch": 0.9433139534883721,
"grad_norm": 1.5004649471635758,
"learning_rate": 8.732501363550028e-08,
"loss": 2.31,
"step": 1298
},
{
"epoch": 0.9440406976744186,
"grad_norm": 1.0296843011644148,
"learning_rate": 8.513450158049109e-08,
"loss": 2.3116,
"step": 1299
},
{
"epoch": 0.9447674418604651,
"grad_norm": 0.9799375953016403,
"learning_rate": 8.29715777741208e-08,
"loss": 2.2626,
"step": 1300
},
{
"epoch": 0.9454941860465116,
"grad_norm": 0.9922942907902778,
"learning_rate": 8.08362543573682e-08,
"loss": 2.2388,
"step": 1301
},
{
"epoch": 0.9462209302325582,
"grad_norm": 2.6395611468739446,
"learning_rate": 7.872854331628599e-08,
"loss": 2.3387,
"step": 1302
},
{
"epoch": 0.9469476744186046,
"grad_norm": 1.3509341255307576,
"learning_rate": 7.664845648193087e-08,
"loss": 2.2566,
"step": 1303
},
{
"epoch": 0.9476744186046512,
"grad_norm": 1.081975825106032,
"learning_rate": 7.459600553029966e-08,
"loss": 2.2852,
"step": 1304
},
{
"epoch": 0.9484011627906976,
"grad_norm": 1.0477686545335492,
"learning_rate": 7.257120198226219e-08,
"loss": 2.2782,
"step": 1305
},
{
"epoch": 0.9491279069767442,
"grad_norm": 1.0147183696685629,
"learning_rate": 7.057405720349853e-08,
"loss": 2.2711,
"step": 1306
},
{
"epoch": 0.9498546511627907,
"grad_norm": 0.9804356237239087,
"learning_rate": 6.860458240443179e-08,
"loss": 2.3472,
"step": 1307
},
{
"epoch": 0.9505813953488372,
"grad_norm": 1.0073244406187998,
"learning_rate": 6.666278864016884e-08,
"loss": 2.2972,
"step": 1308
},
{
"epoch": 0.9513081395348837,
"grad_norm": 1.0325826072792375,
"learning_rate": 6.474868681043578e-08,
"loss": 2.2552,
"step": 1309
},
{
"epoch": 0.9520348837209303,
"grad_norm": 1.0796140386264204,
"learning_rate": 6.286228765951807e-08,
"loss": 2.2278,
"step": 1310
},
{
"epoch": 0.9527616279069767,
"grad_norm": 0.9903943158949445,
"learning_rate": 6.100360177619946e-08,
"loss": 2.3095,
"step": 1311
},
{
"epoch": 0.9534883720930233,
"grad_norm": 1.0300867433621375,
"learning_rate": 5.917263959370312e-08,
"loss": 2.3375,
"step": 1312
},
{
"epoch": 0.9542151162790697,
"grad_norm": 1.076918105492359,
"learning_rate": 5.736941138963281e-08,
"loss": 2.3038,
"step": 1313
},
{
"epoch": 0.9549418604651163,
"grad_norm": 1.0114088652079647,
"learning_rate": 5.559392728591517e-08,
"loss": 2.3148,
"step": 1314
},
{
"epoch": 0.9556686046511628,
"grad_norm": 1.1906333105918063,
"learning_rate": 5.384619724874307e-08,
"loss": 2.2539,
"step": 1315
},
{
"epoch": 0.9563953488372093,
"grad_norm": 1.134623849472849,
"learning_rate": 5.2126231088519e-08,
"loss": 2.2944,
"step": 1316
},
{
"epoch": 0.9571220930232558,
"grad_norm": 0.9992883943690529,
"learning_rate": 5.0434038459801213e-08,
"loss": 2.2755,
"step": 1317
},
{
"epoch": 0.9578488372093024,
"grad_norm": 1.1453760819725112,
"learning_rate": 4.876962886124936e-08,
"loss": 2.3314,
"step": 1318
},
{
"epoch": 0.9585755813953488,
"grad_norm": 1.0711339360870722,
"learning_rate": 4.713301163556894e-08,
"loss": 2.3663,
"step": 1319
},
{
"epoch": 0.9593023255813954,
"grad_norm": 1.193135985144878,
"learning_rate": 4.5524195969461895e-08,
"loss": 2.3048,
"step": 1320
},
{
"epoch": 0.9600290697674418,
"grad_norm": 1.1033385441868588,
"learning_rate": 4.394319089357335e-08,
"loss": 2.3037,
"step": 1321
},
{
"epoch": 0.9607558139534884,
"grad_norm": 1.0425755305320366,
"learning_rate": 4.239000528244164e-08,
"loss": 2.2908,
"step": 1322
},
{
"epoch": 0.9614825581395349,
"grad_norm": 1.0043149458946938,
"learning_rate": 4.086464785444777e-08,
"loss": 2.3172,
"step": 1323
},
{
"epoch": 0.9622093023255814,
"grad_norm": 0.9510526219043042,
"learning_rate": 3.936712717176716e-08,
"loss": 2.3454,
"step": 1324
},
{
"epoch": 0.9629360465116279,
"grad_norm": 1.0242871931784736,
"learning_rate": 3.7897451640321326e-08,
"loss": 2.255,
"step": 1325
},
{
"epoch": 0.9636627906976745,
"grad_norm": 1.032293949458567,
"learning_rate": 3.645562950973014e-08,
"loss": 2.2723,
"step": 1326
},
{
"epoch": 0.9643895348837209,
"grad_norm": 1.0085818746357065,
"learning_rate": 3.504166887326688e-08,
"loss": 2.28,
"step": 1327
},
{
"epoch": 0.9651162790697675,
"grad_norm": 1.0094863113312103,
"learning_rate": 3.365557766781047e-08,
"loss": 2.3334,
"step": 1328
},
{
"epoch": 0.965843023255814,
"grad_norm": 1.0044401474819298,
"learning_rate": 3.229736367380498e-08,
"loss": 2.3234,
"step": 1329
},
{
"epoch": 0.9665697674418605,
"grad_norm": 1.1918541717117412,
"learning_rate": 3.0967034515211323e-08,
"loss": 2.3161,
"step": 1330
},
{
"epoch": 0.967296511627907,
"grad_norm": 0.9889890714491774,
"learning_rate": 2.966459765946672e-08,
"loss": 2.3153,
"step": 1331
},
{
"epoch": 0.9680232558139535,
"grad_norm": 1.064404134607704,
"learning_rate": 2.8390060417443632e-08,
"loss": 2.2949,
"step": 1332
},
{
"epoch": 0.96875,
"grad_norm": 0.9820533577158319,
"learning_rate": 2.714342994340646e-08,
"loss": 2.3176,
"step": 1333
},
{
"epoch": 0.9694767441860465,
"grad_norm": 0.9869071666343803,
"learning_rate": 2.592471323497381e-08,
"loss": 2.2175,
"step": 1334
},
{
"epoch": 0.970203488372093,
"grad_norm": 1.024509197686126,
"learning_rate": 2.4733917133077378e-08,
"loss": 2.3161,
"step": 1335
},
{
"epoch": 0.9709302325581395,
"grad_norm": 1.0089119887260893,
"learning_rate": 2.3571048321923694e-08,
"loss": 2.2915,
"step": 1336
},
{
"epoch": 0.971656976744186,
"grad_norm": 0.9995707439981476,
"learning_rate": 2.2436113328958565e-08,
"loss": 2.3017,
"step": 1337
},
{
"epoch": 0.9723837209302325,
"grad_norm": 1.2909850946179122,
"learning_rate": 2.1329118524827662e-08,
"loss": 2.2576,
"step": 1338
},
{
"epoch": 0.9731104651162791,
"grad_norm": 0.9839313923654567,
"learning_rate": 2.0250070123342124e-08,
"loss": 2.3013,
"step": 1339
},
{
"epoch": 0.9738372093023255,
"grad_norm": 1.0289650966110175,
"learning_rate": 1.9198974181444675e-08,
"loss": 2.3018,
"step": 1340
},
{
"epoch": 0.9745639534883721,
"grad_norm": 1.2914117926481417,
"learning_rate": 1.8175836599173545e-08,
"loss": 2.3079,
"step": 1341
},
{
"epoch": 0.9752906976744186,
"grad_norm": 1.0055706662475437,
"learning_rate": 1.7180663119630846e-08,
"loss": 2.2506,
"step": 1342
},
{
"epoch": 0.9760174418604651,
"grad_norm": 0.9952235589773648,
"learning_rate": 1.6213459328950355e-08,
"loss": 2.2793,
"step": 1343
},
{
"epoch": 0.9767441860465116,
"grad_norm": 1.1363778138841705,
"learning_rate": 1.5274230656263656e-08,
"loss": 2.3271,
"step": 1344
},
{
"epoch": 0.9774709302325582,
"grad_norm": 1.0689170237371675,
"learning_rate": 1.4362982373675171e-08,
"loss": 2.3187,
"step": 1345
},
{
"epoch": 0.9781976744186046,
"grad_norm": 1.0815995913125536,
"learning_rate": 1.347971959622496e-08,
"loss": 2.3334,
"step": 1346
},
{
"epoch": 0.9789244186046512,
"grad_norm": 1.035545163434121,
"learning_rate": 1.2624447281867625e-08,
"loss": 2.3098,
"step": 1347
},
{
"epoch": 0.9796511627906976,
"grad_norm": 1.0948187248544319,
"learning_rate": 1.1797170231439004e-08,
"loss": 2.3251,
"step": 1348
},
{
"epoch": 0.9803779069767442,
"grad_norm": 1.046292294927679,
"learning_rate": 1.0997893088632306e-08,
"loss": 2.3078,
"step": 1349
},
{
"epoch": 0.9811046511627907,
"grad_norm": 1.0300319397600337,
"learning_rate": 1.0226620339969795e-08,
"loss": 2.3033,
"step": 1350
},
{
"epoch": 0.9818313953488372,
"grad_norm": 0.9767087046357332,
"learning_rate": 9.48335631477948e-09,
"loss": 2.3278,
"step": 1351
},
{
"epoch": 0.9825581395348837,
"grad_norm": 1.198857506546238,
"learning_rate": 8.768105185170683e-09,
"loss": 2.2652,
"step": 1352
},
{
"epoch": 0.9832848837209303,
"grad_norm": 0.9863221166831615,
"learning_rate": 8.080870966008513e-09,
"loss": 2.3035,
"step": 1353
},
{
"epoch": 0.9840116279069767,
"grad_norm": 0.9996635167228194,
"learning_rate": 7.421657514893321e-09,
"loss": 2.3307,
"step": 1354
},
{
"epoch": 0.9847383720930233,
"grad_norm": 0.9743277323285772,
"learning_rate": 6.79046853214016e-09,
"loss": 2.2822,
"step": 1355
},
{
"epoch": 0.9854651162790697,
"grad_norm": 1.025580657364137,
"learning_rate": 6.187307560754363e-09,
"loss": 2.3112,
"step": 1356
},
{
"epoch": 0.9861918604651163,
"grad_norm": 1.0871552124899244,
"learning_rate": 5.612177986414891e-09,
"loss": 2.2338,
"step": 1357
},
{
"epoch": 0.9869186046511628,
"grad_norm": 1.2139378980093414,
"learning_rate": 5.065083037453234e-09,
"loss": 2.2583,
"step": 1358
},
{
"epoch": 0.9876453488372093,
"grad_norm": 1.1655655831929417,
"learning_rate": 4.546025784837316e-09,
"loss": 2.3364,
"step": 1359
},
{
"epoch": 0.9883720930232558,
"grad_norm": 1.0020916790797725,
"learning_rate": 4.055009142152066e-09,
"loss": 2.258,
"step": 1360
},
{
"epoch": 0.9890988372093024,
"grad_norm": 1.0161785758267667,
"learning_rate": 3.5920358655844312e-09,
"loss": 2.2484,
"step": 1361
},
{
"epoch": 0.9898255813953488,
"grad_norm": 0.9954194415297898,
"learning_rate": 3.1571085539089384e-09,
"loss": 2.2928,
"step": 1362
},
{
"epoch": 0.9905523255813954,
"grad_norm": 0.9865483653721157,
"learning_rate": 2.7502296484699374e-09,
"loss": 2.2793,
"step": 1363
},
{
"epoch": 0.9912790697674418,
"grad_norm": 1.0946412791341318,
"learning_rate": 2.371401433170495e-09,
"loss": 2.2972,
"step": 1364
},
{
"epoch": 0.9920058139534884,
"grad_norm": 1.004980651503775,
"learning_rate": 2.0206260344590724e-09,
"loss": 2.3216,
"step": 1365
},
{
"epoch": 0.9927325581395349,
"grad_norm": 1.0316396359422022,
"learning_rate": 1.6979054213173141e-09,
"loss": 2.2782,
"step": 1366
},
{
"epoch": 0.9934593023255814,
"grad_norm": 1.3085920072304564,
"learning_rate": 1.4032414052478348e-09,
"loss": 2.2255,
"step": 1367
},
{
"epoch": 0.9941860465116279,
"grad_norm": 1.045228547510724,
"learning_rate": 1.136635640267003e-09,
"loss": 2.3237,
"step": 1368
},
{
"epoch": 0.9949127906976745,
"grad_norm": 1.3155145718130687,
"learning_rate": 8.980896228932834e-10,
"loss": 2.2275,
"step": 1369
},
{
"epoch": 0.9956395348837209,
"grad_norm": 0.9921059811114996,
"learning_rate": 6.876046921389102e-10,
"loss": 2.2629,
"step": 1370
},
{
"epoch": 0.9963662790697675,
"grad_norm": 1.2788236393841776,
"learning_rate": 5.051820295032262e-10,
"loss": 2.2405,
"step": 1371
},
{
"epoch": 0.997093023255814,
"grad_norm": 1.0501847646368694,
"learning_rate": 3.508226589660213e-10,
"loss": 2.3305,
"step": 1372
},
{
"epoch": 0.9978197674418605,
"grad_norm": 1.1446929377734676,
"learning_rate": 2.2452744698087114e-10,
"loss": 2.3071,
"step": 1373
},
{
"epoch": 0.998546511627907,
"grad_norm": 0.9948071557609979,
"learning_rate": 1.2629710247180626e-10,
"loss": 2.2705,
"step": 1374
},
{
"epoch": 0.9992732558139535,
"grad_norm": 1.0555106775673517,
"learning_rate": 5.613217682720606e-11,
"loss": 2.2112,
"step": 1375
},
{
"epoch": 1.0,
"grad_norm": 0.9820602357954474,
"learning_rate": 1.4033063899243637e-11,
"loss": 2.3398,
"step": 1376
}
],
"logging_steps": 1,
"max_steps": 1376,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 688,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.0909863237176525e+18,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}