random_0eU7XGYNErY5g6Js / trainer_state.json
cutelemonlili's picture
Add files using upload-large-folder tool
6373966 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 1124,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0017793594306049821,
"grad_norm": 1.2287051002957676,
"learning_rate": 9.99998046979289e-06,
"loss": 0.1431,
"step": 1
},
{
"epoch": 0.0035587188612099642,
"grad_norm": 1.328316661438433,
"learning_rate": 9.999921879324127e-06,
"loss": 0.1499,
"step": 2
},
{
"epoch": 0.005338078291814947,
"grad_norm": 1.0472423427240536,
"learning_rate": 9.999824229051425e-06,
"loss": 0.144,
"step": 3
},
{
"epoch": 0.0071174377224199285,
"grad_norm": 1.261616825002671,
"learning_rate": 9.999687519737639e-06,
"loss": 0.1511,
"step": 4
},
{
"epoch": 0.008896797153024912,
"grad_norm": 1.0760936144410918,
"learning_rate": 9.99951175245075e-06,
"loss": 0.1421,
"step": 5
},
{
"epoch": 0.010676156583629894,
"grad_norm": 1.341591597145216,
"learning_rate": 9.999296928563868e-06,
"loss": 0.182,
"step": 6
},
{
"epoch": 0.012455516014234875,
"grad_norm": 1.1677113662454504,
"learning_rate": 9.999043049755216e-06,
"loss": 0.1126,
"step": 7
},
{
"epoch": 0.014234875444839857,
"grad_norm": 0.6797240685117082,
"learning_rate": 9.998750118008117e-06,
"loss": 0.0747,
"step": 8
},
{
"epoch": 0.01601423487544484,
"grad_norm": 0.684016833760675,
"learning_rate": 9.998418135610974e-06,
"loss": 0.0765,
"step": 9
},
{
"epoch": 0.017793594306049824,
"grad_norm": 1.0994232826045491,
"learning_rate": 9.998047105157265e-06,
"loss": 0.145,
"step": 10
},
{
"epoch": 0.019572953736654804,
"grad_norm": 0.629761900665598,
"learning_rate": 9.997637029545509e-06,
"loss": 0.0686,
"step": 11
},
{
"epoch": 0.021352313167259787,
"grad_norm": 0.901143072741247,
"learning_rate": 9.997187911979252e-06,
"loss": 0.1061,
"step": 12
},
{
"epoch": 0.023131672597864767,
"grad_norm": 0.8336139643394676,
"learning_rate": 9.996699755967035e-06,
"loss": 0.0878,
"step": 13
},
{
"epoch": 0.02491103202846975,
"grad_norm": 0.7790317932551439,
"learning_rate": 9.996172565322375e-06,
"loss": 0.0887,
"step": 14
},
{
"epoch": 0.026690391459074734,
"grad_norm": 0.6724733060161117,
"learning_rate": 9.995606344163728e-06,
"loss": 0.088,
"step": 15
},
{
"epoch": 0.028469750889679714,
"grad_norm": 0.4997729059784594,
"learning_rate": 9.995001096914462e-06,
"loss": 0.0486,
"step": 16
},
{
"epoch": 0.030249110320284697,
"grad_norm": 0.7197000909568966,
"learning_rate": 9.994356828302818e-06,
"loss": 0.0781,
"step": 17
},
{
"epoch": 0.03202846975088968,
"grad_norm": 0.649712657161332,
"learning_rate": 9.993673543361874e-06,
"loss": 0.0748,
"step": 18
},
{
"epoch": 0.033807829181494664,
"grad_norm": 0.8760515158781976,
"learning_rate": 9.992951247429512e-06,
"loss": 0.1037,
"step": 19
},
{
"epoch": 0.03558718861209965,
"grad_norm": 0.5808585887352716,
"learning_rate": 9.992189946148366e-06,
"loss": 0.0682,
"step": 20
},
{
"epoch": 0.037366548042704624,
"grad_norm": 0.6094819491497003,
"learning_rate": 9.991389645465786e-06,
"loss": 0.0592,
"step": 21
},
{
"epoch": 0.03914590747330961,
"grad_norm": 0.8346538944863554,
"learning_rate": 9.990550351633784e-06,
"loss": 0.0963,
"step": 22
},
{
"epoch": 0.04092526690391459,
"grad_norm": 0.7072363959660027,
"learning_rate": 9.989672071208993e-06,
"loss": 0.0904,
"step": 23
},
{
"epoch": 0.042704626334519574,
"grad_norm": 0.6430528225431822,
"learning_rate": 9.988754811052616e-06,
"loss": 0.0697,
"step": 24
},
{
"epoch": 0.04448398576512456,
"grad_norm": 0.8821929221961957,
"learning_rate": 9.987798578330365e-06,
"loss": 0.0939,
"step": 25
},
{
"epoch": 0.046263345195729534,
"grad_norm": 0.6154864038444308,
"learning_rate": 9.986803380512406e-06,
"loss": 0.0695,
"step": 26
},
{
"epoch": 0.04804270462633452,
"grad_norm": 0.8304640202922415,
"learning_rate": 9.98576922537331e-06,
"loss": 0.0962,
"step": 27
},
{
"epoch": 0.0498220640569395,
"grad_norm": 0.6707019570597479,
"learning_rate": 9.984696120991979e-06,
"loss": 0.0729,
"step": 28
},
{
"epoch": 0.051601423487544484,
"grad_norm": 0.7814393283703666,
"learning_rate": 9.983584075751598e-06,
"loss": 0.0559,
"step": 29
},
{
"epoch": 0.05338078291814947,
"grad_norm": 0.6979222610087836,
"learning_rate": 9.982433098339553e-06,
"loss": 0.0745,
"step": 30
},
{
"epoch": 0.05516014234875445,
"grad_norm": 0.9292618744315236,
"learning_rate": 9.981243197747375e-06,
"loss": 0.094,
"step": 31
},
{
"epoch": 0.05693950177935943,
"grad_norm": 1.1205303969275626,
"learning_rate": 9.980014383270668e-06,
"loss": 0.1125,
"step": 32
},
{
"epoch": 0.05871886120996441,
"grad_norm": 0.7931321707520319,
"learning_rate": 9.978746664509032e-06,
"loss": 0.0824,
"step": 33
},
{
"epoch": 0.060498220640569395,
"grad_norm": 0.681388764098953,
"learning_rate": 9.97744005136599e-06,
"loss": 0.067,
"step": 34
},
{
"epoch": 0.06227758007117438,
"grad_norm": 0.5416885670676512,
"learning_rate": 9.976094554048912e-06,
"loss": 0.0588,
"step": 35
},
{
"epoch": 0.06405693950177936,
"grad_norm": 0.6591545654405103,
"learning_rate": 9.974710183068935e-06,
"loss": 0.0641,
"step": 36
},
{
"epoch": 0.06583629893238434,
"grad_norm": 0.557652746925499,
"learning_rate": 9.97328694924088e-06,
"loss": 0.0615,
"step": 37
},
{
"epoch": 0.06761565836298933,
"grad_norm": 0.6367172947738626,
"learning_rate": 9.971824863683168e-06,
"loss": 0.0737,
"step": 38
},
{
"epoch": 0.0693950177935943,
"grad_norm": 0.8183042059554975,
"learning_rate": 9.970323937817732e-06,
"loss": 0.0966,
"step": 39
},
{
"epoch": 0.0711743772241993,
"grad_norm": 0.6351184798705162,
"learning_rate": 9.968784183369929e-06,
"loss": 0.0628,
"step": 40
},
{
"epoch": 0.07295373665480427,
"grad_norm": 0.8305356177880099,
"learning_rate": 9.96720561236845e-06,
"loss": 0.0897,
"step": 41
},
{
"epoch": 0.07473309608540925,
"grad_norm": 0.8394365432483795,
"learning_rate": 9.965588237145219e-06,
"loss": 0.0895,
"step": 42
},
{
"epoch": 0.07651245551601424,
"grad_norm": 0.6377080342878184,
"learning_rate": 9.963932070335307e-06,
"loss": 0.0657,
"step": 43
},
{
"epoch": 0.07829181494661921,
"grad_norm": 0.8131629700188383,
"learning_rate": 9.962237124876828e-06,
"loss": 0.087,
"step": 44
},
{
"epoch": 0.0800711743772242,
"grad_norm": 0.6611276083558877,
"learning_rate": 9.960503414010833e-06,
"loss": 0.0717,
"step": 45
},
{
"epoch": 0.08185053380782918,
"grad_norm": 0.666799386562066,
"learning_rate": 9.958730951281218e-06,
"loss": 0.0598,
"step": 46
},
{
"epoch": 0.08362989323843416,
"grad_norm": 0.6430793425760815,
"learning_rate": 9.956919750534607e-06,
"loss": 0.0625,
"step": 47
},
{
"epoch": 0.08540925266903915,
"grad_norm": 1.0377863241495953,
"learning_rate": 9.955069825920249e-06,
"loss": 0.111,
"step": 48
},
{
"epoch": 0.08718861209964412,
"grad_norm": 0.7102353943084488,
"learning_rate": 9.953181191889913e-06,
"loss": 0.0719,
"step": 49
},
{
"epoch": 0.08896797153024912,
"grad_norm": 0.5744077840236071,
"learning_rate": 9.95125386319776e-06,
"loss": 0.0649,
"step": 50
},
{
"epoch": 0.09074733096085409,
"grad_norm": 0.8102234815316572,
"learning_rate": 9.949287854900243e-06,
"loss": 0.0681,
"step": 51
},
{
"epoch": 0.09252669039145907,
"grad_norm": 0.8727813269332798,
"learning_rate": 9.947283182355982e-06,
"loss": 0.0961,
"step": 52
},
{
"epoch": 0.09430604982206406,
"grad_norm": 0.5196503142416636,
"learning_rate": 9.945239861225644e-06,
"loss": 0.0476,
"step": 53
},
{
"epoch": 0.09608540925266904,
"grad_norm": 0.9071690533795475,
"learning_rate": 9.943157907471825e-06,
"loss": 0.1021,
"step": 54
},
{
"epoch": 0.09786476868327403,
"grad_norm": 0.6103948746916095,
"learning_rate": 9.941037337358918e-06,
"loss": 0.0651,
"step": 55
},
{
"epoch": 0.099644128113879,
"grad_norm": 0.8961003886367809,
"learning_rate": 9.938878167452991e-06,
"loss": 0.0793,
"step": 56
},
{
"epoch": 0.10142348754448399,
"grad_norm": 1.4032273001061504,
"learning_rate": 9.936680414621663e-06,
"loss": 0.1137,
"step": 57
},
{
"epoch": 0.10320284697508897,
"grad_norm": 0.6823614309008581,
"learning_rate": 9.934444096033958e-06,
"loss": 0.0666,
"step": 58
},
{
"epoch": 0.10498220640569395,
"grad_norm": 0.6193031856915844,
"learning_rate": 9.932169229160183e-06,
"loss": 0.0658,
"step": 59
},
{
"epoch": 0.10676156583629894,
"grad_norm": 0.7200709860401112,
"learning_rate": 9.929855831771787e-06,
"loss": 0.0757,
"step": 60
},
{
"epoch": 0.10854092526690391,
"grad_norm": 0.6477842866085564,
"learning_rate": 9.927503921941218e-06,
"loss": 0.0683,
"step": 61
},
{
"epoch": 0.1103202846975089,
"grad_norm": 0.6704151101293453,
"learning_rate": 9.925113518041796e-06,
"loss": 0.0707,
"step": 62
},
{
"epoch": 0.11209964412811388,
"grad_norm": 0.7350205291327735,
"learning_rate": 9.922684638747551e-06,
"loss": 0.0675,
"step": 63
},
{
"epoch": 0.11387900355871886,
"grad_norm": 0.6076787442954522,
"learning_rate": 9.920217303033091e-06,
"loss": 0.0649,
"step": 64
},
{
"epoch": 0.11565836298932385,
"grad_norm": 0.5022521732479575,
"learning_rate": 9.917711530173444e-06,
"loss": 0.0579,
"step": 65
},
{
"epoch": 0.11743772241992882,
"grad_norm": 0.7489416901899917,
"learning_rate": 9.91516733974392e-06,
"loss": 0.0738,
"step": 66
},
{
"epoch": 0.11921708185053381,
"grad_norm": 0.7128649256271236,
"learning_rate": 9.912584751619943e-06,
"loss": 0.068,
"step": 67
},
{
"epoch": 0.12099644128113879,
"grad_norm": 0.6893148679877266,
"learning_rate": 9.909963785976902e-06,
"loss": 0.076,
"step": 68
},
{
"epoch": 0.12277580071174377,
"grad_norm": 0.735751233586422,
"learning_rate": 9.907304463290004e-06,
"loss": 0.066,
"step": 69
},
{
"epoch": 0.12455516014234876,
"grad_norm": 0.629010664700668,
"learning_rate": 9.904606804334094e-06,
"loss": 0.0585,
"step": 70
},
{
"epoch": 0.12633451957295375,
"grad_norm": 0.7278080611451638,
"learning_rate": 9.901870830183506e-06,
"loss": 0.0768,
"step": 71
},
{
"epoch": 0.12811387900355872,
"grad_norm": 0.5986424438743305,
"learning_rate": 9.899096562211902e-06,
"loss": 0.0632,
"step": 72
},
{
"epoch": 0.1298932384341637,
"grad_norm": 0.8502696649785917,
"learning_rate": 9.896284022092088e-06,
"loss": 0.0888,
"step": 73
},
{
"epoch": 0.13167259786476868,
"grad_norm": 0.9201998395880486,
"learning_rate": 9.893433231795864e-06,
"loss": 0.0914,
"step": 74
},
{
"epoch": 0.13345195729537365,
"grad_norm": 0.9568142935182258,
"learning_rate": 9.890544213593838e-06,
"loss": 0.0971,
"step": 75
},
{
"epoch": 0.13523131672597866,
"grad_norm": 0.7222500495202485,
"learning_rate": 9.887616990055262e-06,
"loss": 0.0834,
"step": 76
},
{
"epoch": 0.13701067615658363,
"grad_norm": 0.7382842221072208,
"learning_rate": 9.884651584047845e-06,
"loss": 0.0865,
"step": 77
},
{
"epoch": 0.1387900355871886,
"grad_norm": 0.633330083736123,
"learning_rate": 9.881648018737587e-06,
"loss": 0.0697,
"step": 78
},
{
"epoch": 0.14056939501779359,
"grad_norm": 0.6769102252255559,
"learning_rate": 9.878606317588588e-06,
"loss": 0.0572,
"step": 79
},
{
"epoch": 0.1423487544483986,
"grad_norm": 1.1686730674421464,
"learning_rate": 9.875526504362868e-06,
"loss": 0.1222,
"step": 80
},
{
"epoch": 0.14412811387900357,
"grad_norm": 0.8946944962778296,
"learning_rate": 9.872408603120187e-06,
"loss": 0.0933,
"step": 81
},
{
"epoch": 0.14590747330960854,
"grad_norm": 0.6636611567534423,
"learning_rate": 9.869252638217846e-06,
"loss": 0.0665,
"step": 82
},
{
"epoch": 0.14768683274021352,
"grad_norm": 0.6042787444443084,
"learning_rate": 9.866058634310503e-06,
"loss": 0.0662,
"step": 83
},
{
"epoch": 0.1494661921708185,
"grad_norm": 0.7550818041695985,
"learning_rate": 9.862826616349981e-06,
"loss": 0.0827,
"step": 84
},
{
"epoch": 0.1512455516014235,
"grad_norm": 0.7989700322537726,
"learning_rate": 9.859556609585075e-06,
"loss": 0.077,
"step": 85
},
{
"epoch": 0.15302491103202848,
"grad_norm": 0.5848917286005131,
"learning_rate": 9.856248639561346e-06,
"loss": 0.0641,
"step": 86
},
{
"epoch": 0.15480427046263345,
"grad_norm": 0.6577846738245182,
"learning_rate": 9.85290273212093e-06,
"loss": 0.0567,
"step": 87
},
{
"epoch": 0.15658362989323843,
"grad_norm": 0.6541171475273805,
"learning_rate": 9.849518913402334e-06,
"loss": 0.065,
"step": 88
},
{
"epoch": 0.1583629893238434,
"grad_norm": 0.7630511168905286,
"learning_rate": 9.84609720984023e-06,
"loss": 0.0827,
"step": 89
},
{
"epoch": 0.1601423487544484,
"grad_norm": 0.691520591015244,
"learning_rate": 9.84263764816525e-06,
"loss": 0.0678,
"step": 90
},
{
"epoch": 0.1619217081850534,
"grad_norm": 0.5651625716460608,
"learning_rate": 9.839140255403776e-06,
"loss": 0.0571,
"step": 91
},
{
"epoch": 0.16370106761565836,
"grad_norm": 0.7972692243013502,
"learning_rate": 9.83560505887773e-06,
"loss": 0.0818,
"step": 92
},
{
"epoch": 0.16548042704626334,
"grad_norm": 0.6342471485876013,
"learning_rate": 9.83203208620436e-06,
"loss": 0.0688,
"step": 93
},
{
"epoch": 0.16725978647686832,
"grad_norm": 0.7649862472508883,
"learning_rate": 9.828421365296023e-06,
"loss": 0.0746,
"step": 94
},
{
"epoch": 0.16903914590747332,
"grad_norm": 0.7511220083435702,
"learning_rate": 9.824772924359974e-06,
"loss": 0.0666,
"step": 95
},
{
"epoch": 0.1708185053380783,
"grad_norm": 0.6140864985810345,
"learning_rate": 9.821086791898133e-06,
"loss": 0.0612,
"step": 96
},
{
"epoch": 0.17259786476868327,
"grad_norm": 0.6106454836307731,
"learning_rate": 9.817362996706872e-06,
"loss": 0.0731,
"step": 97
},
{
"epoch": 0.17437722419928825,
"grad_norm": 0.5305202960928928,
"learning_rate": 9.81360156787679e-06,
"loss": 0.0527,
"step": 98
},
{
"epoch": 0.17615658362989323,
"grad_norm": 0.5702684824491315,
"learning_rate": 9.809802534792477e-06,
"loss": 0.0476,
"step": 99
},
{
"epoch": 0.17793594306049823,
"grad_norm": 0.8454050534685281,
"learning_rate": 9.805965927132294e-06,
"loss": 0.0892,
"step": 100
},
{
"epoch": 0.1797153024911032,
"grad_norm": 0.6548626597778442,
"learning_rate": 9.802091774868143e-06,
"loss": 0.0614,
"step": 101
},
{
"epoch": 0.18149466192170818,
"grad_norm": 0.5509279935307513,
"learning_rate": 9.798180108265218e-06,
"loss": 0.0585,
"step": 102
},
{
"epoch": 0.18327402135231316,
"grad_norm": 0.6970946382405296,
"learning_rate": 9.794230957881785e-06,
"loss": 0.0823,
"step": 103
},
{
"epoch": 0.18505338078291814,
"grad_norm": 0.7896773151945998,
"learning_rate": 9.79024435456893e-06,
"loss": 0.0807,
"step": 104
},
{
"epoch": 0.18683274021352314,
"grad_norm": 0.723360902835066,
"learning_rate": 9.786220329470334e-06,
"loss": 0.0667,
"step": 105
},
{
"epoch": 0.18861209964412812,
"grad_norm": 0.6540048268339445,
"learning_rate": 9.782158914022011e-06,
"loss": 0.067,
"step": 106
},
{
"epoch": 0.1903914590747331,
"grad_norm": 0.7050287689263087,
"learning_rate": 9.778060139952075e-06,
"loss": 0.063,
"step": 107
},
{
"epoch": 0.19217081850533807,
"grad_norm": 0.8968261818934136,
"learning_rate": 9.773924039280488e-06,
"loss": 0.0919,
"step": 108
},
{
"epoch": 0.19395017793594305,
"grad_norm": 0.541275460725368,
"learning_rate": 9.769750644318814e-06,
"loss": 0.0546,
"step": 109
},
{
"epoch": 0.19572953736654805,
"grad_norm": 0.7612403911408803,
"learning_rate": 9.765539987669956e-06,
"loss": 0.0706,
"step": 110
},
{
"epoch": 0.19750889679715303,
"grad_norm": 0.6915525049828356,
"learning_rate": 9.761292102227917e-06,
"loss": 0.0735,
"step": 111
},
{
"epoch": 0.199288256227758,
"grad_norm": 1.1930784631813578,
"learning_rate": 9.757007021177529e-06,
"loss": 0.0846,
"step": 112
},
{
"epoch": 0.20106761565836298,
"grad_norm": 0.4402389459874699,
"learning_rate": 9.752684777994197e-06,
"loss": 0.0441,
"step": 113
},
{
"epoch": 0.20284697508896798,
"grad_norm": 0.6973554951209938,
"learning_rate": 9.748325406443647e-06,
"loss": 0.0592,
"step": 114
},
{
"epoch": 0.20462633451957296,
"grad_norm": 0.7939621580401891,
"learning_rate": 9.743928940581646e-06,
"loss": 0.0751,
"step": 115
},
{
"epoch": 0.20640569395017794,
"grad_norm": 0.6872570076289733,
"learning_rate": 9.739495414753754e-06,
"loss": 0.0657,
"step": 116
},
{
"epoch": 0.20818505338078291,
"grad_norm": 0.69296021829135,
"learning_rate": 9.73502486359504e-06,
"loss": 0.0695,
"step": 117
},
{
"epoch": 0.2099644128113879,
"grad_norm": 0.6530260083762813,
"learning_rate": 9.73051732202982e-06,
"loss": 0.0543,
"step": 118
},
{
"epoch": 0.2117437722419929,
"grad_norm": 0.6369806425011684,
"learning_rate": 9.725972825271381e-06,
"loss": 0.0606,
"step": 119
},
{
"epoch": 0.21352313167259787,
"grad_norm": 0.6767056785033603,
"learning_rate": 9.721391408821713e-06,
"loss": 0.0758,
"step": 120
},
{
"epoch": 0.21530249110320285,
"grad_norm": 0.5720508981382746,
"learning_rate": 9.716773108471213e-06,
"loss": 0.0612,
"step": 121
},
{
"epoch": 0.21708185053380782,
"grad_norm": 0.765384143047754,
"learning_rate": 9.712117960298433e-06,
"loss": 0.0795,
"step": 122
},
{
"epoch": 0.2188612099644128,
"grad_norm": 0.6605036245337603,
"learning_rate": 9.707426000669773e-06,
"loss": 0.0696,
"step": 123
},
{
"epoch": 0.2206405693950178,
"grad_norm": 0.7069644842646444,
"learning_rate": 9.702697266239211e-06,
"loss": 0.0718,
"step": 124
},
{
"epoch": 0.22241992882562278,
"grad_norm": 0.6887891803042019,
"learning_rate": 9.697931793948012e-06,
"loss": 0.0739,
"step": 125
},
{
"epoch": 0.22419928825622776,
"grad_norm": 0.5802237523536334,
"learning_rate": 9.693129621024441e-06,
"loss": 0.0604,
"step": 126
},
{
"epoch": 0.22597864768683273,
"grad_norm": 0.6726231607910361,
"learning_rate": 9.68829078498347e-06,
"loss": 0.0748,
"step": 127
},
{
"epoch": 0.2277580071174377,
"grad_norm": 0.8366149002185953,
"learning_rate": 9.683415323626487e-06,
"loss": 0.0848,
"step": 128
},
{
"epoch": 0.22953736654804271,
"grad_norm": 0.9530248522788181,
"learning_rate": 9.678503275040997e-06,
"loss": 0.0956,
"step": 129
},
{
"epoch": 0.2313167259786477,
"grad_norm": 0.5973510355696137,
"learning_rate": 9.673554677600336e-06,
"loss": 0.0511,
"step": 130
},
{
"epoch": 0.23309608540925267,
"grad_norm": 0.602070086768945,
"learning_rate": 9.668569569963355e-06,
"loss": 0.0646,
"step": 131
},
{
"epoch": 0.23487544483985764,
"grad_norm": 0.5515919440621416,
"learning_rate": 9.663547991074129e-06,
"loss": 0.0507,
"step": 132
},
{
"epoch": 0.23665480427046262,
"grad_norm": 0.7661987719955262,
"learning_rate": 9.658489980161643e-06,
"loss": 0.0742,
"step": 133
},
{
"epoch": 0.23843416370106763,
"grad_norm": 0.7235194197327811,
"learning_rate": 9.653395576739504e-06,
"loss": 0.0757,
"step": 134
},
{
"epoch": 0.2402135231316726,
"grad_norm": 0.7361903944947585,
"learning_rate": 9.648264820605611e-06,
"loss": 0.0658,
"step": 135
},
{
"epoch": 0.24199288256227758,
"grad_norm": 0.5216011702292412,
"learning_rate": 9.643097751841854e-06,
"loss": 0.058,
"step": 136
},
{
"epoch": 0.24377224199288255,
"grad_norm": 0.658748324038026,
"learning_rate": 9.637894410813803e-06,
"loss": 0.0747,
"step": 137
},
{
"epoch": 0.24555160142348753,
"grad_norm": 0.7081146897847538,
"learning_rate": 9.632654838170393e-06,
"loss": 0.0655,
"step": 138
},
{
"epoch": 0.24733096085409254,
"grad_norm": 0.6500887432131869,
"learning_rate": 9.627379074843595e-06,
"loss": 0.0654,
"step": 139
},
{
"epoch": 0.2491103202846975,
"grad_norm": 0.7366684680268587,
"learning_rate": 9.622067162048111e-06,
"loss": 0.0781,
"step": 140
},
{
"epoch": 0.2508896797153025,
"grad_norm": 0.5937482387013016,
"learning_rate": 9.616719141281044e-06,
"loss": 0.061,
"step": 141
},
{
"epoch": 0.2526690391459075,
"grad_norm": 0.8730784749631377,
"learning_rate": 9.611335054321576e-06,
"loss": 0.0882,
"step": 142
},
{
"epoch": 0.25444839857651247,
"grad_norm": 0.8048594343697372,
"learning_rate": 9.605914943230637e-06,
"loss": 0.0773,
"step": 143
},
{
"epoch": 0.25622775800711745,
"grad_norm": 0.5746763559538758,
"learning_rate": 9.600458850350588e-06,
"loss": 0.0689,
"step": 144
},
{
"epoch": 0.2580071174377224,
"grad_norm": 0.8904218543959297,
"learning_rate": 9.594966818304875e-06,
"loss": 0.0802,
"step": 145
},
{
"epoch": 0.2597864768683274,
"grad_norm": 0.7064752155922343,
"learning_rate": 9.589438889997712e-06,
"loss": 0.0628,
"step": 146
},
{
"epoch": 0.2615658362989324,
"grad_norm": 0.6176761831142004,
"learning_rate": 9.583875108613727e-06,
"loss": 0.052,
"step": 147
},
{
"epoch": 0.26334519572953735,
"grad_norm": 0.7226831087836451,
"learning_rate": 9.578275517617646e-06,
"loss": 0.0686,
"step": 148
},
{
"epoch": 0.26512455516014233,
"grad_norm": 0.6862360098837336,
"learning_rate": 9.572640160753936e-06,
"loss": 0.0691,
"step": 149
},
{
"epoch": 0.2669039145907473,
"grad_norm": 0.6163595987795264,
"learning_rate": 9.566969082046471e-06,
"loss": 0.0617,
"step": 150
},
{
"epoch": 0.26868327402135234,
"grad_norm": 0.4878038359174218,
"learning_rate": 9.561262325798188e-06,
"loss": 0.0445,
"step": 151
},
{
"epoch": 0.2704626334519573,
"grad_norm": 0.600382503559289,
"learning_rate": 9.555519936590739e-06,
"loss": 0.0669,
"step": 152
},
{
"epoch": 0.2722419928825623,
"grad_norm": 0.6562128300557961,
"learning_rate": 9.549741959284147e-06,
"loss": 0.0649,
"step": 153
},
{
"epoch": 0.27402135231316727,
"grad_norm": 0.5730712826477441,
"learning_rate": 9.543928439016445e-06,
"loss": 0.053,
"step": 154
},
{
"epoch": 0.27580071174377224,
"grad_norm": 0.7007297094892871,
"learning_rate": 9.538079421203339e-06,
"loss": 0.0714,
"step": 155
},
{
"epoch": 0.2775800711743772,
"grad_norm": 0.8739947775273068,
"learning_rate": 9.532194951537838e-06,
"loss": 0.0849,
"step": 156
},
{
"epoch": 0.2793594306049822,
"grad_norm": 0.645371740082366,
"learning_rate": 9.52627507598991e-06,
"loss": 0.0597,
"step": 157
},
{
"epoch": 0.28113879003558717,
"grad_norm": 0.5828553901595871,
"learning_rate": 9.52031984080611e-06,
"loss": 0.0585,
"step": 158
},
{
"epoch": 0.28291814946619215,
"grad_norm": 0.6101159880882936,
"learning_rate": 9.514329292509227e-06,
"loss": 0.0659,
"step": 159
},
{
"epoch": 0.2846975088967972,
"grad_norm": 0.8851863661019759,
"learning_rate": 9.508303477897925e-06,
"loss": 0.0744,
"step": 160
},
{
"epoch": 0.28647686832740216,
"grad_norm": 0.7838978043928034,
"learning_rate": 9.502242444046365e-06,
"loss": 0.0776,
"step": 161
},
{
"epoch": 0.28825622775800713,
"grad_norm": 0.633239852147155,
"learning_rate": 9.496146238303846e-06,
"loss": 0.0634,
"step": 162
},
{
"epoch": 0.2900355871886121,
"grad_norm": 0.7171204886990744,
"learning_rate": 9.49001490829443e-06,
"loss": 0.0616,
"step": 163
},
{
"epoch": 0.2918149466192171,
"grad_norm": 0.8650956649130569,
"learning_rate": 9.483848501916578e-06,
"loss": 0.1028,
"step": 164
},
{
"epoch": 0.29359430604982206,
"grad_norm": 0.7884289849742133,
"learning_rate": 9.477647067342766e-06,
"loss": 0.0813,
"step": 165
},
{
"epoch": 0.29537366548042704,
"grad_norm": 0.6925967131732893,
"learning_rate": 9.471410653019115e-06,
"loss": 0.0707,
"step": 166
},
{
"epoch": 0.297153024911032,
"grad_norm": 0.6265538174180255,
"learning_rate": 9.46513930766501e-06,
"loss": 0.0642,
"step": 167
},
{
"epoch": 0.298932384341637,
"grad_norm": 0.6417001124412925,
"learning_rate": 9.458833080272723e-06,
"loss": 0.0614,
"step": 168
},
{
"epoch": 0.30071174377224197,
"grad_norm": 0.6702422520090452,
"learning_rate": 9.45249202010702e-06,
"loss": 0.0612,
"step": 169
},
{
"epoch": 0.302491103202847,
"grad_norm": 0.9284189746074317,
"learning_rate": 9.446116176704791e-06,
"loss": 0.089,
"step": 170
},
{
"epoch": 0.304270462633452,
"grad_norm": 0.6948377972356458,
"learning_rate": 9.439705599874653e-06,
"loss": 0.07,
"step": 171
},
{
"epoch": 0.30604982206405695,
"grad_norm": 0.7863036459212169,
"learning_rate": 9.433260339696564e-06,
"loss": 0.0786,
"step": 172
},
{
"epoch": 0.30782918149466193,
"grad_norm": 0.7648838638347256,
"learning_rate": 9.426780446521429e-06,
"loss": 0.0765,
"step": 173
},
{
"epoch": 0.3096085409252669,
"grad_norm": 0.6644265273258103,
"learning_rate": 9.42026597097071e-06,
"loss": 0.0662,
"step": 174
},
{
"epoch": 0.3113879003558719,
"grad_norm": 0.5690568275364777,
"learning_rate": 9.413716963936033e-06,
"loss": 0.0579,
"step": 175
},
{
"epoch": 0.31316725978647686,
"grad_norm": 0.6793974499399753,
"learning_rate": 9.407133476578778e-06,
"loss": 0.0597,
"step": 176
},
{
"epoch": 0.31494661921708184,
"grad_norm": 1.0091050574054283,
"learning_rate": 9.400515560329698e-06,
"loss": 0.1013,
"step": 177
},
{
"epoch": 0.3167259786476868,
"grad_norm": 0.6591769000915146,
"learning_rate": 9.393863266888501e-06,
"loss": 0.0489,
"step": 178
},
{
"epoch": 0.3185053380782918,
"grad_norm": 0.7710446107825684,
"learning_rate": 9.387176648223457e-06,
"loss": 0.0759,
"step": 179
},
{
"epoch": 0.3202846975088968,
"grad_norm": 1.0141065717973643,
"learning_rate": 9.38045575657098e-06,
"loss": 0.0749,
"step": 180
},
{
"epoch": 0.3220640569395018,
"grad_norm": 0.8716680231979295,
"learning_rate": 9.37370064443524e-06,
"loss": 0.0753,
"step": 181
},
{
"epoch": 0.3238434163701068,
"grad_norm": 0.5287598581031884,
"learning_rate": 9.366911364587726e-06,
"loss": 0.0485,
"step": 182
},
{
"epoch": 0.32562277580071175,
"grad_norm": 0.7080819550756755,
"learning_rate": 9.360087970066854e-06,
"loss": 0.0592,
"step": 183
},
{
"epoch": 0.3274021352313167,
"grad_norm": 0.6409930818741086,
"learning_rate": 9.353230514177553e-06,
"loss": 0.0538,
"step": 184
},
{
"epoch": 0.3291814946619217,
"grad_norm": 0.657904152935833,
"learning_rate": 9.346339050490832e-06,
"loss": 0.0651,
"step": 185
},
{
"epoch": 0.3309608540925267,
"grad_norm": 0.9159554730057862,
"learning_rate": 9.33941363284338e-06,
"loss": 0.0939,
"step": 186
},
{
"epoch": 0.33274021352313166,
"grad_norm": 0.7076967733007625,
"learning_rate": 9.332454315337129e-06,
"loss": 0.0727,
"step": 187
},
{
"epoch": 0.33451957295373663,
"grad_norm": 0.7915640372359511,
"learning_rate": 9.325461152338846e-06,
"loss": 0.0687,
"step": 188
},
{
"epoch": 0.33629893238434166,
"grad_norm": 0.8025801231759407,
"learning_rate": 9.3184341984797e-06,
"loss": 0.0817,
"step": 189
},
{
"epoch": 0.33807829181494664,
"grad_norm": 0.7063384420602035,
"learning_rate": 9.311373508654838e-06,
"loss": 0.0687,
"step": 190
},
{
"epoch": 0.3398576512455516,
"grad_norm": 0.8067957413993656,
"learning_rate": 9.30427913802295e-06,
"loss": 0.074,
"step": 191
},
{
"epoch": 0.3416370106761566,
"grad_norm": 1.0278741472051922,
"learning_rate": 9.297151142005852e-06,
"loss": 0.102,
"step": 192
},
{
"epoch": 0.34341637010676157,
"grad_norm": 0.6333587413022541,
"learning_rate": 9.289989576288035e-06,
"loss": 0.0669,
"step": 193
},
{
"epoch": 0.34519572953736655,
"grad_norm": 0.6882588238728015,
"learning_rate": 9.282794496816244e-06,
"loss": 0.0667,
"step": 194
},
{
"epoch": 0.3469750889679715,
"grad_norm": 0.6204558442733177,
"learning_rate": 9.27556595979904e-06,
"loss": 0.0553,
"step": 195
},
{
"epoch": 0.3487544483985765,
"grad_norm": 0.6910955644663985,
"learning_rate": 9.26830402170635e-06,
"loss": 0.0688,
"step": 196
},
{
"epoch": 0.3505338078291815,
"grad_norm": 0.7614659543258289,
"learning_rate": 9.261008739269035e-06,
"loss": 0.0765,
"step": 197
},
{
"epoch": 0.35231316725978645,
"grad_norm": 0.6258466645255691,
"learning_rate": 9.253680169478448e-06,
"loss": 0.0629,
"step": 198
},
{
"epoch": 0.3540925266903915,
"grad_norm": 0.7834521648360228,
"learning_rate": 9.246318369585983e-06,
"loss": 0.0739,
"step": 199
},
{
"epoch": 0.35587188612099646,
"grad_norm": 0.8694900104783158,
"learning_rate": 9.238923397102629e-06,
"loss": 0.0902,
"step": 200
},
{
"epoch": 0.35765124555160144,
"grad_norm": 0.6576925792644225,
"learning_rate": 9.231495309798525e-06,
"loss": 0.0601,
"step": 201
},
{
"epoch": 0.3594306049822064,
"grad_norm": 0.768530032295965,
"learning_rate": 9.224034165702506e-06,
"loss": 0.0716,
"step": 202
},
{
"epoch": 0.3612099644128114,
"grad_norm": 0.4886729385446261,
"learning_rate": 9.216540023101646e-06,
"loss": 0.048,
"step": 203
},
{
"epoch": 0.36298932384341637,
"grad_norm": 0.6257911888810285,
"learning_rate": 9.209012940540806e-06,
"loss": 0.0636,
"step": 204
},
{
"epoch": 0.36476868327402134,
"grad_norm": 0.6227633970830718,
"learning_rate": 9.20145297682218e-06,
"loss": 0.0621,
"step": 205
},
{
"epoch": 0.3665480427046263,
"grad_norm": 0.8627702458529797,
"learning_rate": 9.193860191004833e-06,
"loss": 0.0868,
"step": 206
},
{
"epoch": 0.3683274021352313,
"grad_norm": 0.6716058224954908,
"learning_rate": 9.186234642404234e-06,
"loss": 0.0662,
"step": 207
},
{
"epoch": 0.3701067615658363,
"grad_norm": 0.8321781074665348,
"learning_rate": 9.178576390591803e-06,
"loss": 0.0737,
"step": 208
},
{
"epoch": 0.3718861209964413,
"grad_norm": 0.631567956783431,
"learning_rate": 9.170885495394435e-06,
"loss": 0.0595,
"step": 209
},
{
"epoch": 0.3736654804270463,
"grad_norm": 0.9163664898914814,
"learning_rate": 9.16316201689404e-06,
"loss": 0.0855,
"step": 210
},
{
"epoch": 0.37544483985765126,
"grad_norm": 0.8015949321344831,
"learning_rate": 9.155406015427076e-06,
"loss": 0.077,
"step": 211
},
{
"epoch": 0.37722419928825623,
"grad_norm": 0.9519123331914713,
"learning_rate": 9.147617551584066e-06,
"loss": 0.0872,
"step": 212
},
{
"epoch": 0.3790035587188612,
"grad_norm": 0.7210512071930563,
"learning_rate": 9.139796686209135e-06,
"loss": 0.0992,
"step": 213
},
{
"epoch": 0.3807829181494662,
"grad_norm": 0.5997792848808455,
"learning_rate": 9.131943480399531e-06,
"loss": 0.0562,
"step": 214
},
{
"epoch": 0.38256227758007116,
"grad_norm": 0.6162558154690008,
"learning_rate": 9.124057995505148e-06,
"loss": 0.0697,
"step": 215
},
{
"epoch": 0.38434163701067614,
"grad_norm": 0.9557796113428656,
"learning_rate": 9.11614029312805e-06,
"loss": 0.1031,
"step": 216
},
{
"epoch": 0.3861209964412811,
"grad_norm": 0.7852542487610065,
"learning_rate": 9.108190435121982e-06,
"loss": 0.0809,
"step": 217
},
{
"epoch": 0.3879003558718861,
"grad_norm": 0.8722435551277069,
"learning_rate": 9.100208483591892e-06,
"loss": 0.0889,
"step": 218
},
{
"epoch": 0.3896797153024911,
"grad_norm": 0.8518499601063854,
"learning_rate": 9.092194500893448e-06,
"loss": 0.0879,
"step": 219
},
{
"epoch": 0.3914590747330961,
"grad_norm": 0.6462997777718903,
"learning_rate": 9.084148549632547e-06,
"loss": 0.0498,
"step": 220
},
{
"epoch": 0.3932384341637011,
"grad_norm": 0.7951379647091716,
"learning_rate": 9.076070692664827e-06,
"loss": 0.0812,
"step": 221
},
{
"epoch": 0.39501779359430605,
"grad_norm": 0.6515729780902559,
"learning_rate": 9.067960993095176e-06,
"loss": 0.0583,
"step": 222
},
{
"epoch": 0.39679715302491103,
"grad_norm": 0.8208416664111609,
"learning_rate": 9.059819514277238e-06,
"loss": 0.0788,
"step": 223
},
{
"epoch": 0.398576512455516,
"grad_norm": 0.7312842823397765,
"learning_rate": 9.05164631981292e-06,
"loss": 0.0813,
"step": 224
},
{
"epoch": 0.400355871886121,
"grad_norm": 0.6789403841894707,
"learning_rate": 9.043441473551893e-06,
"loss": 0.0733,
"step": 225
},
{
"epoch": 0.40213523131672596,
"grad_norm": 0.6522634736322597,
"learning_rate": 9.035205039591099e-06,
"loss": 0.0553,
"step": 226
},
{
"epoch": 0.40391459074733094,
"grad_norm": 1.1749269335176993,
"learning_rate": 9.02693708227424e-06,
"loss": 0.1124,
"step": 227
},
{
"epoch": 0.40569395017793597,
"grad_norm": 0.6522130676677695,
"learning_rate": 9.018637666191284e-06,
"loss": 0.0718,
"step": 228
},
{
"epoch": 0.40747330960854095,
"grad_norm": 0.8624795312858453,
"learning_rate": 9.010306856177958e-06,
"loss": 0.0851,
"step": 229
},
{
"epoch": 0.4092526690391459,
"grad_norm": 0.5904314457428553,
"learning_rate": 9.001944717315236e-06,
"loss": 0.0581,
"step": 230
},
{
"epoch": 0.4110320284697509,
"grad_norm": 0.7935670050987601,
"learning_rate": 8.993551314928846e-06,
"loss": 0.0825,
"step": 231
},
{
"epoch": 0.4128113879003559,
"grad_norm": 0.5914410055559369,
"learning_rate": 8.985126714588739e-06,
"loss": 0.0541,
"step": 232
},
{
"epoch": 0.41459074733096085,
"grad_norm": 0.6431224011224128,
"learning_rate": 8.976670982108591e-06,
"loss": 0.0712,
"step": 233
},
{
"epoch": 0.41637010676156583,
"grad_norm": 0.7826097260825766,
"learning_rate": 8.968184183545285e-06,
"loss": 0.0845,
"step": 234
},
{
"epoch": 0.4181494661921708,
"grad_norm": 0.7193453326638846,
"learning_rate": 8.959666385198396e-06,
"loss": 0.0648,
"step": 235
},
{
"epoch": 0.4199288256227758,
"grad_norm": 0.6252477186587466,
"learning_rate": 8.951117653609666e-06,
"loss": 0.0625,
"step": 236
},
{
"epoch": 0.42170818505338076,
"grad_norm": 0.9151745268751516,
"learning_rate": 8.9425380555625e-06,
"loss": 0.0912,
"step": 237
},
{
"epoch": 0.4234875444839858,
"grad_norm": 0.484996223256912,
"learning_rate": 8.933927658081423e-06,
"loss": 0.0461,
"step": 238
},
{
"epoch": 0.42526690391459077,
"grad_norm": 0.5697523309909434,
"learning_rate": 8.925286528431578e-06,
"loss": 0.0649,
"step": 239
},
{
"epoch": 0.42704626334519574,
"grad_norm": 0.6908536719964652,
"learning_rate": 8.916614734118184e-06,
"loss": 0.0675,
"step": 240
},
{
"epoch": 0.4288256227758007,
"grad_norm": 0.5198893984950097,
"learning_rate": 8.907912342886016e-06,
"loss": 0.0479,
"step": 241
},
{
"epoch": 0.4306049822064057,
"grad_norm": 0.7498759611208019,
"learning_rate": 8.899179422718877e-06,
"loss": 0.0706,
"step": 242
},
{
"epoch": 0.43238434163701067,
"grad_norm": 0.7314628910814697,
"learning_rate": 8.890416041839061e-06,
"loss": 0.0569,
"step": 243
},
{
"epoch": 0.43416370106761565,
"grad_norm": 0.678298957281942,
"learning_rate": 8.881622268706825e-06,
"loss": 0.0683,
"step": 244
},
{
"epoch": 0.4359430604982206,
"grad_norm": 0.6415922539886235,
"learning_rate": 8.872798172019856e-06,
"loss": 0.0649,
"step": 245
},
{
"epoch": 0.4377224199288256,
"grad_norm": 0.628152820130725,
"learning_rate": 8.863943820712726e-06,
"loss": 0.0592,
"step": 246
},
{
"epoch": 0.4395017793594306,
"grad_norm": 0.759819499389858,
"learning_rate": 8.855059283956363e-06,
"loss": 0.0904,
"step": 247
},
{
"epoch": 0.4412811387900356,
"grad_norm": 0.8239144847966118,
"learning_rate": 8.8461446311575e-06,
"loss": 0.0831,
"step": 248
},
{
"epoch": 0.4430604982206406,
"grad_norm": 0.6454891285667056,
"learning_rate": 8.837199931958147e-06,
"loss": 0.0609,
"step": 249
},
{
"epoch": 0.44483985765124556,
"grad_norm": 0.6801231067148992,
"learning_rate": 8.828225256235035e-06,
"loss": 0.0629,
"step": 250
},
{
"epoch": 0.44661921708185054,
"grad_norm": 0.7861130085783277,
"learning_rate": 8.819220674099074e-06,
"loss": 0.0742,
"step": 251
},
{
"epoch": 0.4483985765124555,
"grad_norm": 0.6228146514914071,
"learning_rate": 8.810186255894804e-06,
"loss": 0.0602,
"step": 252
},
{
"epoch": 0.4501779359430605,
"grad_norm": 0.7572339079493575,
"learning_rate": 8.801122072199848e-06,
"loss": 0.0697,
"step": 253
},
{
"epoch": 0.45195729537366547,
"grad_norm": 0.6639221182220044,
"learning_rate": 8.792028193824364e-06,
"loss": 0.0669,
"step": 254
},
{
"epoch": 0.45373665480427045,
"grad_norm": 0.5560650795964978,
"learning_rate": 8.782904691810478e-06,
"loss": 0.0469,
"step": 255
},
{
"epoch": 0.4555160142348754,
"grad_norm": 0.7113533515389538,
"learning_rate": 8.77375163743175e-06,
"loss": 0.0615,
"step": 256
},
{
"epoch": 0.45729537366548045,
"grad_norm": 0.6996953152939676,
"learning_rate": 8.764569102192593e-06,
"loss": 0.085,
"step": 257
},
{
"epoch": 0.45907473309608543,
"grad_norm": 0.7278133546007519,
"learning_rate": 8.755357157827735e-06,
"loss": 0.0742,
"step": 258
},
{
"epoch": 0.4608540925266904,
"grad_norm": 0.9180098288781817,
"learning_rate": 8.746115876301651e-06,
"loss": 0.0942,
"step": 259
},
{
"epoch": 0.4626334519572954,
"grad_norm": 0.651769852014177,
"learning_rate": 8.736845329807994e-06,
"loss": 0.0634,
"step": 260
},
{
"epoch": 0.46441281138790036,
"grad_norm": 0.7900515253103204,
"learning_rate": 8.727545590769044e-06,
"loss": 0.0767,
"step": 261
},
{
"epoch": 0.46619217081850534,
"grad_norm": 0.5754117708142552,
"learning_rate": 8.718216731835131e-06,
"loss": 0.0596,
"step": 262
},
{
"epoch": 0.4679715302491103,
"grad_norm": 0.7292261292392734,
"learning_rate": 8.708858825884075e-06,
"loss": 0.0832,
"step": 263
},
{
"epoch": 0.4697508896797153,
"grad_norm": 0.7644135851249976,
"learning_rate": 8.699471946020612e-06,
"loss": 0.0761,
"step": 264
},
{
"epoch": 0.47153024911032027,
"grad_norm": 0.7043033648059586,
"learning_rate": 8.690056165575825e-06,
"loss": 0.0654,
"step": 265
},
{
"epoch": 0.47330960854092524,
"grad_norm": 0.6755981278707108,
"learning_rate": 8.680611558106571e-06,
"loss": 0.0684,
"step": 266
},
{
"epoch": 0.4750889679715303,
"grad_norm": 0.6235005875138784,
"learning_rate": 8.671138197394907e-06,
"loss": 0.0534,
"step": 267
},
{
"epoch": 0.47686832740213525,
"grad_norm": 0.5100564089208902,
"learning_rate": 8.661636157447511e-06,
"loss": 0.0581,
"step": 268
},
{
"epoch": 0.4786476868327402,
"grad_norm": 0.9576974445258026,
"learning_rate": 8.652105512495106e-06,
"loss": 0.0785,
"step": 269
},
{
"epoch": 0.4804270462633452,
"grad_norm": 0.6616255422283299,
"learning_rate": 8.64254633699188e-06,
"loss": 0.0601,
"step": 270
},
{
"epoch": 0.4822064056939502,
"grad_norm": 0.813645011518409,
"learning_rate": 8.632958705614905e-06,
"loss": 0.0791,
"step": 271
},
{
"epoch": 0.48398576512455516,
"grad_norm": 0.6511721408615697,
"learning_rate": 8.623342693263549e-06,
"loss": 0.0633,
"step": 272
},
{
"epoch": 0.48576512455516013,
"grad_norm": 0.7335056749805812,
"learning_rate": 8.6136983750589e-06,
"loss": 0.0814,
"step": 273
},
{
"epoch": 0.4875444839857651,
"grad_norm": 0.603529192463855,
"learning_rate": 8.604025826343167e-06,
"loss": 0.0652,
"step": 274
},
{
"epoch": 0.4893238434163701,
"grad_norm": 0.7293321601307705,
"learning_rate": 8.594325122679107e-06,
"loss": 0.0581,
"step": 275
},
{
"epoch": 0.49110320284697506,
"grad_norm": 0.7703175404812748,
"learning_rate": 8.584596339849419e-06,
"loss": 0.1029,
"step": 276
},
{
"epoch": 0.4928825622775801,
"grad_norm": 0.4848214361265286,
"learning_rate": 8.574839553856157e-06,
"loss": 0.0463,
"step": 277
},
{
"epoch": 0.49466192170818507,
"grad_norm": 0.7924699346710438,
"learning_rate": 8.565054840920145e-06,
"loss": 0.082,
"step": 278
},
{
"epoch": 0.49644128113879005,
"grad_norm": 0.4824698625985039,
"learning_rate": 8.55524227748037e-06,
"loss": 0.0451,
"step": 279
},
{
"epoch": 0.498220640569395,
"grad_norm": 0.6832585031935443,
"learning_rate": 8.545401940193392e-06,
"loss": 0.0563,
"step": 280
},
{
"epoch": 0.5,
"grad_norm": 0.6546997520090722,
"learning_rate": 8.535533905932739e-06,
"loss": 0.0625,
"step": 281
},
{
"epoch": 0.501779359430605,
"grad_norm": 1.1292933888909238,
"learning_rate": 8.525638251788312e-06,
"loss": 0.1142,
"step": 282
},
{
"epoch": 0.50355871886121,
"grad_norm": 0.738025227365862,
"learning_rate": 8.515715055065783e-06,
"loss": 0.0922,
"step": 283
},
{
"epoch": 0.505338078291815,
"grad_norm": 0.7181605808065559,
"learning_rate": 8.505764393285985e-06,
"loss": 0.077,
"step": 284
},
{
"epoch": 0.5071174377224199,
"grad_norm": 0.996929885291125,
"learning_rate": 8.495786344184314e-06,
"loss": 0.091,
"step": 285
},
{
"epoch": 0.5088967971530249,
"grad_norm": 0.6051375859094301,
"learning_rate": 8.485780985710113e-06,
"loss": 0.0562,
"step": 286
},
{
"epoch": 0.5106761565836299,
"grad_norm": 0.6262489677291431,
"learning_rate": 8.475748396026074e-06,
"loss": 0.0709,
"step": 287
},
{
"epoch": 0.5124555160142349,
"grad_norm": 0.935462458064743,
"learning_rate": 8.46568865350762e-06,
"loss": 0.0865,
"step": 288
},
{
"epoch": 0.5142348754448398,
"grad_norm": 0.8903384378751757,
"learning_rate": 8.45560183674229e-06,
"loss": 0.0883,
"step": 289
},
{
"epoch": 0.5160142348754448,
"grad_norm": 0.7286855902692326,
"learning_rate": 8.445488024529133e-06,
"loss": 0.0745,
"step": 290
},
{
"epoch": 0.5177935943060499,
"grad_norm": 0.6445658589254484,
"learning_rate": 8.435347295878087e-06,
"loss": 0.0648,
"step": 291
},
{
"epoch": 0.5195729537366548,
"grad_norm": 0.6060124105394451,
"learning_rate": 8.425179730009368e-06,
"loss": 0.0579,
"step": 292
},
{
"epoch": 0.5213523131672598,
"grad_norm": 0.535503366784967,
"learning_rate": 8.41498540635284e-06,
"loss": 0.0552,
"step": 293
},
{
"epoch": 0.5231316725978647,
"grad_norm": 0.7679411378383981,
"learning_rate": 8.404764404547404e-06,
"loss": 0.0713,
"step": 294
},
{
"epoch": 0.5249110320284698,
"grad_norm": 0.8078617445189842,
"learning_rate": 8.394516804440374e-06,
"loss": 0.091,
"step": 295
},
{
"epoch": 0.5266903914590747,
"grad_norm": 0.7800420953366541,
"learning_rate": 8.384242686086848e-06,
"loss": 0.0816,
"step": 296
},
{
"epoch": 0.5284697508896797,
"grad_norm": 0.6450657948237661,
"learning_rate": 8.373942129749094e-06,
"loss": 0.0698,
"step": 297
},
{
"epoch": 0.5302491103202847,
"grad_norm": 0.5126701692253715,
"learning_rate": 8.363615215895908e-06,
"loss": 0.0552,
"step": 298
},
{
"epoch": 0.5320284697508897,
"grad_norm": 0.7474821469479275,
"learning_rate": 8.353262025202e-06,
"loss": 0.0752,
"step": 299
},
{
"epoch": 0.5338078291814946,
"grad_norm": 0.6502592320251632,
"learning_rate": 8.342882638547351e-06,
"loss": 0.0601,
"step": 300
},
{
"epoch": 0.5355871886120996,
"grad_norm": 0.5953828045874886,
"learning_rate": 8.332477137016587e-06,
"loss": 0.0591,
"step": 301
},
{
"epoch": 0.5373665480427047,
"grad_norm": 0.48588698677737296,
"learning_rate": 8.322045601898354e-06,
"loss": 0.047,
"step": 302
},
{
"epoch": 0.5391459074733096,
"grad_norm": 0.5304771958046166,
"learning_rate": 8.311588114684665e-06,
"loss": 0.0571,
"step": 303
},
{
"epoch": 0.5409252669039146,
"grad_norm": 0.6973405912688413,
"learning_rate": 8.301104757070276e-06,
"loss": 0.0938,
"step": 304
},
{
"epoch": 0.5427046263345195,
"grad_norm": 0.6201619994315529,
"learning_rate": 8.290595610952045e-06,
"loss": 0.0626,
"step": 305
},
{
"epoch": 0.5444839857651246,
"grad_norm": 0.8778029393308154,
"learning_rate": 8.280060758428294e-06,
"loss": 0.0909,
"step": 306
},
{
"epoch": 0.5462633451957295,
"grad_norm": 0.4824899964282266,
"learning_rate": 8.269500281798164e-06,
"loss": 0.047,
"step": 307
},
{
"epoch": 0.5480427046263345,
"grad_norm": 0.7365896083701101,
"learning_rate": 8.258914263560971e-06,
"loss": 0.071,
"step": 308
},
{
"epoch": 0.5498220640569395,
"grad_norm": 0.6448411570636917,
"learning_rate": 8.248302786415567e-06,
"loss": 0.0608,
"step": 309
},
{
"epoch": 0.5516014234875445,
"grad_norm": 0.7928088237628461,
"learning_rate": 8.237665933259693e-06,
"loss": 0.0758,
"step": 310
},
{
"epoch": 0.5533807829181495,
"grad_norm": 0.631180955087727,
"learning_rate": 8.227003787189323e-06,
"loss": 0.0618,
"step": 311
},
{
"epoch": 0.5551601423487544,
"grad_norm": 0.48085104028148584,
"learning_rate": 8.216316431498028e-06,
"loss": 0.047,
"step": 312
},
{
"epoch": 0.5569395017793595,
"grad_norm": 0.6747073624423635,
"learning_rate": 8.205603949676317e-06,
"loss": 0.0704,
"step": 313
},
{
"epoch": 0.5587188612099644,
"grad_norm": 0.7097344849630727,
"learning_rate": 8.194866425410984e-06,
"loss": 0.0803,
"step": 314
},
{
"epoch": 0.5604982206405694,
"grad_norm": 0.722672955831541,
"learning_rate": 8.184103942584456e-06,
"loss": 0.0653,
"step": 315
},
{
"epoch": 0.5622775800711743,
"grad_norm": 0.7483051082145639,
"learning_rate": 8.173316585274144e-06,
"loss": 0.067,
"step": 316
},
{
"epoch": 0.5640569395017794,
"grad_norm": 0.796806495490532,
"learning_rate": 8.162504437751775e-06,
"loss": 0.0755,
"step": 317
},
{
"epoch": 0.5658362989323843,
"grad_norm": 0.619163216690872,
"learning_rate": 8.151667584482742e-06,
"loss": 0.0608,
"step": 318
},
{
"epoch": 0.5676156583629893,
"grad_norm": 0.8108165494399907,
"learning_rate": 8.140806110125442e-06,
"loss": 0.0862,
"step": 319
},
{
"epoch": 0.5693950177935944,
"grad_norm": 0.6580580131746375,
"learning_rate": 8.129920099530608e-06,
"loss": 0.0696,
"step": 320
},
{
"epoch": 0.5711743772241993,
"grad_norm": 0.5149311412674462,
"learning_rate": 8.119009637740663e-06,
"loss": 0.0426,
"step": 321
},
{
"epoch": 0.5729537366548043,
"grad_norm": 0.5686528087561702,
"learning_rate": 8.108074809989032e-06,
"loss": 0.0635,
"step": 322
},
{
"epoch": 0.5747330960854092,
"grad_norm": 0.9348519704845141,
"learning_rate": 8.097115701699498e-06,
"loss": 0.0803,
"step": 323
},
{
"epoch": 0.5765124555160143,
"grad_norm": 0.7552240539476781,
"learning_rate": 8.086132398485525e-06,
"loss": 0.0884,
"step": 324
},
{
"epoch": 0.5782918149466192,
"grad_norm": 0.4038915109018739,
"learning_rate": 8.075124986149583e-06,
"loss": 0.041,
"step": 325
},
{
"epoch": 0.5800711743772242,
"grad_norm": 0.7264877761111537,
"learning_rate": 8.064093550682494e-06,
"loss": 0.0647,
"step": 326
},
{
"epoch": 0.5818505338078291,
"grad_norm": 0.6806711363388825,
"learning_rate": 8.053038178262742e-06,
"loss": 0.0526,
"step": 327
},
{
"epoch": 0.5836298932384342,
"grad_norm": 0.9164999406714408,
"learning_rate": 8.041958955255815e-06,
"loss": 0.0751,
"step": 328
},
{
"epoch": 0.5854092526690391,
"grad_norm": 0.7325885031738016,
"learning_rate": 8.030855968213518e-06,
"loss": 0.0745,
"step": 329
},
{
"epoch": 0.5871886120996441,
"grad_norm": 0.7625301323338678,
"learning_rate": 8.019729303873307e-06,
"loss": 0.0741,
"step": 330
},
{
"epoch": 0.5889679715302492,
"grad_norm": 0.5818217406072602,
"learning_rate": 8.008579049157607e-06,
"loss": 0.0532,
"step": 331
},
{
"epoch": 0.5907473309608541,
"grad_norm": 0.5684271184673029,
"learning_rate": 7.99740529117313e-06,
"loss": 0.06,
"step": 332
},
{
"epoch": 0.5925266903914591,
"grad_norm": 0.6522638389553282,
"learning_rate": 7.986208117210198e-06,
"loss": 0.0634,
"step": 333
},
{
"epoch": 0.594306049822064,
"grad_norm": 0.8411790737856999,
"learning_rate": 7.974987614742066e-06,
"loss": 0.0868,
"step": 334
},
{
"epoch": 0.5960854092526691,
"grad_norm": 0.7304080022709364,
"learning_rate": 7.963743871424224e-06,
"loss": 0.066,
"step": 335
},
{
"epoch": 0.597864768683274,
"grad_norm": 0.6593715937402292,
"learning_rate": 7.952476975093729e-06,
"loss": 0.0609,
"step": 336
},
{
"epoch": 0.599644128113879,
"grad_norm": 0.7697482528347014,
"learning_rate": 7.941187013768508e-06,
"loss": 0.0881,
"step": 337
},
{
"epoch": 0.6014234875444839,
"grad_norm": 0.6835140384099027,
"learning_rate": 7.929874075646673e-06,
"loss": 0.0657,
"step": 338
},
{
"epoch": 0.603202846975089,
"grad_norm": 0.7772488090497259,
"learning_rate": 7.918538249105835e-06,
"loss": 0.069,
"step": 339
},
{
"epoch": 0.604982206405694,
"grad_norm": 0.8924826661632429,
"learning_rate": 7.907179622702409e-06,
"loss": 0.0947,
"step": 340
},
{
"epoch": 0.6067615658362989,
"grad_norm": 0.5340536156792078,
"learning_rate": 7.895798285170927e-06,
"loss": 0.0553,
"step": 341
},
{
"epoch": 0.608540925266904,
"grad_norm": 0.62671526073955,
"learning_rate": 7.88439432542334e-06,
"loss": 0.059,
"step": 342
},
{
"epoch": 0.6103202846975089,
"grad_norm": 0.5523010203044488,
"learning_rate": 7.872967832548327e-06,
"loss": 0.0632,
"step": 343
},
{
"epoch": 0.6120996441281139,
"grad_norm": 0.8693266051291886,
"learning_rate": 7.861518895810597e-06,
"loss": 0.1028,
"step": 344
},
{
"epoch": 0.6138790035587188,
"grad_norm": 0.5935520434112935,
"learning_rate": 7.850047604650188e-06,
"loss": 0.0616,
"step": 345
},
{
"epoch": 0.6156583629893239,
"grad_norm": 0.6260807498790539,
"learning_rate": 7.838554048681783e-06,
"loss": 0.0609,
"step": 346
},
{
"epoch": 0.6174377224199288,
"grad_norm": 0.8002780413407536,
"learning_rate": 7.827038317693988e-06,
"loss": 0.0937,
"step": 347
},
{
"epoch": 0.6192170818505338,
"grad_norm": 0.6866358983290911,
"learning_rate": 7.815500501648654e-06,
"loss": 0.065,
"step": 348
},
{
"epoch": 0.6209964412811388,
"grad_norm": 0.9047133625551661,
"learning_rate": 7.80394069068015e-06,
"loss": 0.0812,
"step": 349
},
{
"epoch": 0.6227758007117438,
"grad_norm": 0.8214649505167316,
"learning_rate": 7.79235897509468e-06,
"loss": 0.0718,
"step": 350
},
{
"epoch": 0.6245551601423488,
"grad_norm": 0.7421416611333658,
"learning_rate": 7.780755445369563e-06,
"loss": 0.0811,
"step": 351
},
{
"epoch": 0.6263345195729537,
"grad_norm": 0.9645661062646296,
"learning_rate": 7.769130192152538e-06,
"loss": 0.1019,
"step": 352
},
{
"epoch": 0.6281138790035588,
"grad_norm": 0.8980775557585837,
"learning_rate": 7.757483306261042e-06,
"loss": 0.0903,
"step": 353
},
{
"epoch": 0.6298932384341637,
"grad_norm": 0.9397455199795565,
"learning_rate": 7.745814878681516e-06,
"loss": 0.0994,
"step": 354
},
{
"epoch": 0.6316725978647687,
"grad_norm": 0.5734335584076025,
"learning_rate": 7.734125000568684e-06,
"loss": 0.0535,
"step": 355
},
{
"epoch": 0.6334519572953736,
"grad_norm": 0.6199988130829074,
"learning_rate": 7.722413763244837e-06,
"loss": 0.0511,
"step": 356
},
{
"epoch": 0.6352313167259787,
"grad_norm": 0.5153211378768511,
"learning_rate": 7.710681258199136e-06,
"loss": 0.0544,
"step": 357
},
{
"epoch": 0.6370106761565836,
"grad_norm": 0.75224816579673,
"learning_rate": 7.69892757708688e-06,
"loss": 0.0717,
"step": 358
},
{
"epoch": 0.6387900355871886,
"grad_norm": 0.6427436266956046,
"learning_rate": 7.687152811728799e-06,
"loss": 0.0651,
"step": 359
},
{
"epoch": 0.6405693950177936,
"grad_norm": 0.676134355412856,
"learning_rate": 7.675357054110337e-06,
"loss": 0.0671,
"step": 360
},
{
"epoch": 0.6423487544483986,
"grad_norm": 0.6223826380380137,
"learning_rate": 7.663540396380931e-06,
"loss": 0.0589,
"step": 361
},
{
"epoch": 0.6441281138790036,
"grad_norm": 0.634791193063349,
"learning_rate": 7.651702930853287e-06,
"loss": 0.0623,
"step": 362
},
{
"epoch": 0.6459074733096085,
"grad_norm": 0.8335402163276783,
"learning_rate": 7.639844750002668e-06,
"loss": 0.082,
"step": 363
},
{
"epoch": 0.6476868327402135,
"grad_norm": 0.6716093900265061,
"learning_rate": 7.627965946466167e-06,
"loss": 0.0653,
"step": 364
},
{
"epoch": 0.6494661921708185,
"grad_norm": 0.7523516176225902,
"learning_rate": 7.616066613041977e-06,
"loss": 0.0774,
"step": 365
},
{
"epoch": 0.6512455516014235,
"grad_norm": 0.884868397207999,
"learning_rate": 7.6041468426886785e-06,
"loss": 0.0844,
"step": 366
},
{
"epoch": 0.6530249110320284,
"grad_norm": 0.4312842256040485,
"learning_rate": 7.592206728524507e-06,
"loss": 0.0373,
"step": 367
},
{
"epoch": 0.6548042704626335,
"grad_norm": 0.9070914972583358,
"learning_rate": 7.580246363826621e-06,
"loss": 0.0925,
"step": 368
},
{
"epoch": 0.6565836298932385,
"grad_norm": 0.895489601891725,
"learning_rate": 7.568265842030381e-06,
"loss": 0.0998,
"step": 369
},
{
"epoch": 0.6583629893238434,
"grad_norm": 0.5654260225814501,
"learning_rate": 7.556265256728618e-06,
"loss": 0.0586,
"step": 370
},
{
"epoch": 0.6601423487544484,
"grad_norm": 0.7660885593790417,
"learning_rate": 7.544244701670894e-06,
"loss": 0.0806,
"step": 371
},
{
"epoch": 0.6619217081850534,
"grad_norm": 0.7178779459149556,
"learning_rate": 7.532204270762786e-06,
"loss": 0.0688,
"step": 372
},
{
"epoch": 0.6637010676156584,
"grad_norm": 0.989777550107603,
"learning_rate": 7.520144058065133e-06,
"loss": 0.0713,
"step": 373
},
{
"epoch": 0.6654804270462633,
"grad_norm": 0.8082382988960944,
"learning_rate": 7.50806415779332e-06,
"loss": 0.0713,
"step": 374
},
{
"epoch": 0.6672597864768683,
"grad_norm": 0.7675290395877853,
"learning_rate": 7.495964664316525e-06,
"loss": 0.0777,
"step": 375
},
{
"epoch": 0.6690391459074733,
"grad_norm": 0.8041997552501599,
"learning_rate": 7.4838456721569975e-06,
"loss": 0.084,
"step": 376
},
{
"epoch": 0.6708185053380783,
"grad_norm": 0.5576466441700395,
"learning_rate": 7.471707275989304e-06,
"loss": 0.0652,
"step": 377
},
{
"epoch": 0.6725978647686833,
"grad_norm": 0.5590482211802095,
"learning_rate": 7.459549570639602e-06,
"loss": 0.0501,
"step": 378
},
{
"epoch": 0.6743772241992882,
"grad_norm": 0.7603814119173841,
"learning_rate": 7.447372651084896e-06,
"loss": 0.071,
"step": 379
},
{
"epoch": 0.6761565836298933,
"grad_norm": 0.6025128230421307,
"learning_rate": 7.435176612452286e-06,
"loss": 0.0564,
"step": 380
},
{
"epoch": 0.6779359430604982,
"grad_norm": 0.643771855748427,
"learning_rate": 7.4229615500182396e-06,
"loss": 0.052,
"step": 381
},
{
"epoch": 0.6797153024911032,
"grad_norm": 0.552092345273491,
"learning_rate": 7.4107275592078345e-06,
"loss": 0.0575,
"step": 382
},
{
"epoch": 0.6814946619217082,
"grad_norm": 0.6425712357425862,
"learning_rate": 7.398474735594022e-06,
"loss": 0.0605,
"step": 383
},
{
"epoch": 0.6832740213523132,
"grad_norm": 0.8224092488681517,
"learning_rate": 7.386203174896872e-06,
"loss": 0.0757,
"step": 384
},
{
"epoch": 0.6850533807829181,
"grad_norm": 0.7183533069622634,
"learning_rate": 7.373912972982838e-06,
"loss": 0.0617,
"step": 385
},
{
"epoch": 0.6868327402135231,
"grad_norm": 0.6409236590427451,
"learning_rate": 7.361604225863992e-06,
"loss": 0.0635,
"step": 386
},
{
"epoch": 0.6886120996441281,
"grad_norm": 0.714881163770755,
"learning_rate": 7.349277029697287e-06,
"loss": 0.0652,
"step": 387
},
{
"epoch": 0.6903914590747331,
"grad_norm": 0.7018082964532534,
"learning_rate": 7.336931480783801e-06,
"loss": 0.0794,
"step": 388
},
{
"epoch": 0.6921708185053381,
"grad_norm": 0.6929221000164245,
"learning_rate": 7.3245676755679854e-06,
"loss": 0.0724,
"step": 389
},
{
"epoch": 0.693950177935943,
"grad_norm": 0.792814848910068,
"learning_rate": 7.312185710636911e-06,
"loss": 0.0776,
"step": 390
},
{
"epoch": 0.6957295373665481,
"grad_norm": 0.6985521046077963,
"learning_rate": 7.299785682719512e-06,
"loss": 0.0807,
"step": 391
},
{
"epoch": 0.697508896797153,
"grad_norm": 0.5520024217077182,
"learning_rate": 7.287367688685835e-06,
"loss": 0.0573,
"step": 392
},
{
"epoch": 0.699288256227758,
"grad_norm": 1.0316414347756817,
"learning_rate": 7.274931825546279e-06,
"loss": 0.1039,
"step": 393
},
{
"epoch": 0.701067615658363,
"grad_norm": 0.7526772348210492,
"learning_rate": 7.262478190450834e-06,
"loss": 0.0685,
"step": 394
},
{
"epoch": 0.702846975088968,
"grad_norm": 0.8610424296821027,
"learning_rate": 7.250006880688332e-06,
"loss": 0.0779,
"step": 395
},
{
"epoch": 0.7046263345195729,
"grad_norm": 0.8339678175075566,
"learning_rate": 7.2375179936856775e-06,
"loss": 0.0841,
"step": 396
},
{
"epoch": 0.7064056939501779,
"grad_norm": 0.6453500908547836,
"learning_rate": 7.22501162700709e-06,
"loss": 0.0619,
"step": 397
},
{
"epoch": 0.708185053380783,
"grad_norm": 0.6472889872553371,
"learning_rate": 7.21248787835334e-06,
"loss": 0.0538,
"step": 398
},
{
"epoch": 0.7099644128113879,
"grad_norm": 0.5419348678396662,
"learning_rate": 7.199946845560994e-06,
"loss": 0.0429,
"step": 399
},
{
"epoch": 0.7117437722419929,
"grad_norm": 0.7041041085118886,
"learning_rate": 7.1873886266016365e-06,
"loss": 0.0671,
"step": 400
},
{
"epoch": 0.7135231316725978,
"grad_norm": 0.6010926389793982,
"learning_rate": 7.174813319581115e-06,
"loss": 0.0624,
"step": 401
},
{
"epoch": 0.7153024911032029,
"grad_norm": 0.8212800041588307,
"learning_rate": 7.162221022738768e-06,
"loss": 0.0678,
"step": 402
},
{
"epoch": 0.7170818505338078,
"grad_norm": 0.6094853842816113,
"learning_rate": 7.149611834446664e-06,
"loss": 0.066,
"step": 403
},
{
"epoch": 0.7188612099644128,
"grad_norm": 0.8105624719908342,
"learning_rate": 7.136985853208824e-06,
"loss": 0.0754,
"step": 404
},
{
"epoch": 0.7206405693950177,
"grad_norm": 0.7117356859108068,
"learning_rate": 7.124343177660462e-06,
"loss": 0.0718,
"step": 405
},
{
"epoch": 0.7224199288256228,
"grad_norm": 1.0888870675282183,
"learning_rate": 7.111683906567206e-06,
"loss": 0.0826,
"step": 406
},
{
"epoch": 0.7241992882562278,
"grad_norm": 0.7616565668745587,
"learning_rate": 7.099008138824329e-06,
"loss": 0.0676,
"step": 407
},
{
"epoch": 0.7259786476868327,
"grad_norm": 0.6893942775351468,
"learning_rate": 7.086315973455982e-06,
"loss": 0.0776,
"step": 408
},
{
"epoch": 0.7277580071174378,
"grad_norm": 0.6653010726452421,
"learning_rate": 7.0736075096144084e-06,
"loss": 0.0701,
"step": 409
},
{
"epoch": 0.7295373665480427,
"grad_norm": 0.5602768806482805,
"learning_rate": 7.060882846579182e-06,
"loss": 0.0459,
"step": 410
},
{
"epoch": 0.7313167259786477,
"grad_norm": 0.5751271698200824,
"learning_rate": 7.048142083756427e-06,
"loss": 0.0555,
"step": 411
},
{
"epoch": 0.7330960854092526,
"grad_norm": 0.7593317853468826,
"learning_rate": 7.035385320678035e-06,
"loss": 0.0689,
"step": 412
},
{
"epoch": 0.7348754448398577,
"grad_norm": 0.9775690867946747,
"learning_rate": 7.022612657000898e-06,
"loss": 0.0985,
"step": 413
},
{
"epoch": 0.7366548042704626,
"grad_norm": 0.7499055520750201,
"learning_rate": 7.0098241925061215e-06,
"loss": 0.0689,
"step": 414
},
{
"epoch": 0.7384341637010676,
"grad_norm": 0.7350503110640004,
"learning_rate": 6.997020027098249e-06,
"loss": 0.0684,
"step": 415
},
{
"epoch": 0.7402135231316725,
"grad_norm": 0.5905328820897843,
"learning_rate": 6.9842002608044844e-06,
"loss": 0.0588,
"step": 416
},
{
"epoch": 0.7419928825622776,
"grad_norm": 0.605332340886857,
"learning_rate": 6.971364993773901e-06,
"loss": 0.0563,
"step": 417
},
{
"epoch": 0.7437722419928826,
"grad_norm": 0.6661744807467044,
"learning_rate": 6.958514326276669e-06,
"loss": 0.0549,
"step": 418
},
{
"epoch": 0.7455516014234875,
"grad_norm": 0.5660721841359815,
"learning_rate": 6.945648358703269e-06,
"loss": 0.0542,
"step": 419
},
{
"epoch": 0.7473309608540926,
"grad_norm": 0.6569406664697057,
"learning_rate": 6.932767191563703e-06,
"loss": 0.0676,
"step": 420
},
{
"epoch": 0.7491103202846975,
"grad_norm": 0.5606341609682046,
"learning_rate": 6.919870925486718e-06,
"loss": 0.0518,
"step": 421
},
{
"epoch": 0.7508896797153025,
"grad_norm": 0.5856711411202924,
"learning_rate": 6.906959661219011e-06,
"loss": 0.0602,
"step": 422
},
{
"epoch": 0.7526690391459074,
"grad_norm": 0.6217308820746088,
"learning_rate": 6.8940334996244505e-06,
"loss": 0.0636,
"step": 423
},
{
"epoch": 0.7544483985765125,
"grad_norm": 0.7949975408737848,
"learning_rate": 6.881092541683279e-06,
"loss": 0.075,
"step": 424
},
{
"epoch": 0.7562277580071174,
"grad_norm": 0.612524967707652,
"learning_rate": 6.8681368884913345e-06,
"loss": 0.0541,
"step": 425
},
{
"epoch": 0.7580071174377224,
"grad_norm": 0.6858037675273395,
"learning_rate": 6.855166641259252e-06,
"loss": 0.0759,
"step": 426
},
{
"epoch": 0.7597864768683275,
"grad_norm": 0.6440337463580342,
"learning_rate": 6.8421819013116766e-06,
"loss": 0.0759,
"step": 427
},
{
"epoch": 0.7615658362989324,
"grad_norm": 0.6005268303581746,
"learning_rate": 6.829182770086474e-06,
"loss": 0.0643,
"step": 428
},
{
"epoch": 0.7633451957295374,
"grad_norm": 0.7063462904065313,
"learning_rate": 6.816169349133934e-06,
"loss": 0.0707,
"step": 429
},
{
"epoch": 0.7651245551601423,
"grad_norm": 0.6257020513804384,
"learning_rate": 6.803141740115979e-06,
"loss": 0.0523,
"step": 430
},
{
"epoch": 0.7669039145907474,
"grad_norm": 1.0616065300369593,
"learning_rate": 6.7901000448053676e-06,
"loss": 0.0938,
"step": 431
},
{
"epoch": 0.7686832740213523,
"grad_norm": 0.6027818313958888,
"learning_rate": 6.777044365084907e-06,
"loss": 0.0592,
"step": 432
},
{
"epoch": 0.7704626334519573,
"grad_norm": 0.5854053258817512,
"learning_rate": 6.763974802946649e-06,
"loss": 0.0529,
"step": 433
},
{
"epoch": 0.7722419928825622,
"grad_norm": 0.6718409724935704,
"learning_rate": 6.750891460491093e-06,
"loss": 0.0702,
"step": 434
},
{
"epoch": 0.7740213523131673,
"grad_norm": 0.8396495939725284,
"learning_rate": 6.737794439926395e-06,
"loss": 0.0858,
"step": 435
},
{
"epoch": 0.7758007117437722,
"grad_norm": 0.6883104969649357,
"learning_rate": 6.724683843567567e-06,
"loss": 0.0689,
"step": 436
},
{
"epoch": 0.7775800711743772,
"grad_norm": 0.7137481647049393,
"learning_rate": 6.711559773835672e-06,
"loss": 0.0682,
"step": 437
},
{
"epoch": 0.7793594306049823,
"grad_norm": 0.7563030032011214,
"learning_rate": 6.69842233325703e-06,
"loss": 0.0736,
"step": 438
},
{
"epoch": 0.7811387900355872,
"grad_norm": 0.7329095458397233,
"learning_rate": 6.685271624462416e-06,
"loss": 0.0654,
"step": 439
},
{
"epoch": 0.7829181494661922,
"grad_norm": 0.6124474807640846,
"learning_rate": 6.672107750186255e-06,
"loss": 0.0534,
"step": 440
},
{
"epoch": 0.7846975088967971,
"grad_norm": 0.41032838224060697,
"learning_rate": 6.658930813265825e-06,
"loss": 0.0364,
"step": 441
},
{
"epoch": 0.7864768683274022,
"grad_norm": 0.6764828681905263,
"learning_rate": 6.645740916640449e-06,
"loss": 0.074,
"step": 442
},
{
"epoch": 0.7882562277580071,
"grad_norm": 0.6101725823710006,
"learning_rate": 6.63253816335069e-06,
"loss": 0.0501,
"step": 443
},
{
"epoch": 0.7900355871886121,
"grad_norm": 0.5716280356508288,
"learning_rate": 6.619322656537552e-06,
"loss": 0.0534,
"step": 444
},
{
"epoch": 0.791814946619217,
"grad_norm": 0.7384422816241225,
"learning_rate": 6.606094499441671e-06,
"loss": 0.0678,
"step": 445
},
{
"epoch": 0.7935943060498221,
"grad_norm": 0.6811167106946131,
"learning_rate": 6.592853795402502e-06,
"loss": 0.0639,
"step": 446
},
{
"epoch": 0.7953736654804271,
"grad_norm": 0.6219935446553387,
"learning_rate": 6.579600647857525e-06,
"loss": 0.0515,
"step": 447
},
{
"epoch": 0.797153024911032,
"grad_norm": 1.1217026096642748,
"learning_rate": 6.566335160341425e-06,
"loss": 0.0873,
"step": 448
},
{
"epoch": 0.798932384341637,
"grad_norm": 0.5462727754705452,
"learning_rate": 6.553057436485289e-06,
"loss": 0.0572,
"step": 449
},
{
"epoch": 0.800711743772242,
"grad_norm": 0.577188457911302,
"learning_rate": 6.539767580015799e-06,
"loss": 0.0716,
"step": 450
},
{
"epoch": 0.802491103202847,
"grad_norm": 0.9746504693361744,
"learning_rate": 6.52646569475441e-06,
"loss": 0.0875,
"step": 451
},
{
"epoch": 0.8042704626334519,
"grad_norm": 0.6603531635103416,
"learning_rate": 6.513151884616556e-06,
"loss": 0.0658,
"step": 452
},
{
"epoch": 0.806049822064057,
"grad_norm": 0.8463029376414105,
"learning_rate": 6.499826253610823e-06,
"loss": 0.0759,
"step": 453
},
{
"epoch": 0.8078291814946619,
"grad_norm": 0.637518256808881,
"learning_rate": 6.486488905838143e-06,
"loss": 0.0574,
"step": 454
},
{
"epoch": 0.8096085409252669,
"grad_norm": 0.6203939718167747,
"learning_rate": 6.473139945490984e-06,
"loss": 0.0552,
"step": 455
},
{
"epoch": 0.8113879003558719,
"grad_norm": 0.8950329055840025,
"learning_rate": 6.459779476852528e-06,
"loss": 0.0755,
"step": 456
},
{
"epoch": 0.8131672597864769,
"grad_norm": 0.6295717439962958,
"learning_rate": 6.446407604295863e-06,
"loss": 0.0638,
"step": 457
},
{
"epoch": 0.8149466192170819,
"grad_norm": 0.8229892213016687,
"learning_rate": 6.433024432283169e-06,
"loss": 0.0699,
"step": 458
},
{
"epoch": 0.8167259786476868,
"grad_norm": 0.5747555859150263,
"learning_rate": 6.41963006536489e-06,
"loss": 0.0493,
"step": 459
},
{
"epoch": 0.8185053380782918,
"grad_norm": 0.6765423750455049,
"learning_rate": 6.4062246081789316e-06,
"loss": 0.0679,
"step": 460
},
{
"epoch": 0.8202846975088968,
"grad_norm": 0.5250470156424751,
"learning_rate": 6.392808165449836e-06,
"loss": 0.045,
"step": 461
},
{
"epoch": 0.8220640569395018,
"grad_norm": 0.5765821532501635,
"learning_rate": 6.379380841987965e-06,
"loss": 0.0603,
"step": 462
},
{
"epoch": 0.8238434163701067,
"grad_norm": 0.7959793586158014,
"learning_rate": 6.365942742688684e-06,
"loss": 0.0682,
"step": 463
},
{
"epoch": 0.8256227758007118,
"grad_norm": 0.7275158514609877,
"learning_rate": 6.352493972531535e-06,
"loss": 0.0618,
"step": 464
},
{
"epoch": 0.8274021352313167,
"grad_norm": 0.8615877414742431,
"learning_rate": 6.339034636579425e-06,
"loss": 0.074,
"step": 465
},
{
"epoch": 0.8291814946619217,
"grad_norm": 0.613751798058676,
"learning_rate": 6.325564839977802e-06,
"loss": 0.0566,
"step": 466
},
{
"epoch": 0.8309608540925267,
"grad_norm": 0.5968688309291743,
"learning_rate": 6.312084687953835e-06,
"loss": 0.0583,
"step": 467
},
{
"epoch": 0.8327402135231317,
"grad_norm": 0.9714507202074504,
"learning_rate": 6.298594285815585e-06,
"loss": 0.0897,
"step": 468
},
{
"epoch": 0.8345195729537367,
"grad_norm": 0.7511034043929439,
"learning_rate": 6.2850937389511936e-06,
"loss": 0.0768,
"step": 469
},
{
"epoch": 0.8362989323843416,
"grad_norm": 0.5856720030345731,
"learning_rate": 6.271583152828049e-06,
"loss": 0.0606,
"step": 470
},
{
"epoch": 0.8380782918149466,
"grad_norm": 0.6314776501285371,
"learning_rate": 6.258062632991972e-06,
"loss": 0.0516,
"step": 471
},
{
"epoch": 0.8398576512455516,
"grad_norm": 0.5711004700337038,
"learning_rate": 6.244532285066382e-06,
"loss": 0.0538,
"step": 472
},
{
"epoch": 0.8416370106761566,
"grad_norm": 0.5454199245177097,
"learning_rate": 6.2309922147514775e-06,
"loss": 0.0473,
"step": 473
},
{
"epoch": 0.8434163701067615,
"grad_norm": 0.7292400182344442,
"learning_rate": 6.2174425278234115e-06,
"loss": 0.0959,
"step": 474
},
{
"epoch": 0.8451957295373665,
"grad_norm": 0.7106161254796619,
"learning_rate": 6.20388333013346e-06,
"loss": 0.0714,
"step": 475
},
{
"epoch": 0.8469750889679716,
"grad_norm": 0.622547092618638,
"learning_rate": 6.190314727607196e-06,
"loss": 0.0628,
"step": 476
},
{
"epoch": 0.8487544483985765,
"grad_norm": 0.7101405088354685,
"learning_rate": 6.176736826243671e-06,
"loss": 0.071,
"step": 477
},
{
"epoch": 0.8505338078291815,
"grad_norm": 0.690277573505413,
"learning_rate": 6.163149732114571e-06,
"loss": 0.0773,
"step": 478
},
{
"epoch": 0.8523131672597865,
"grad_norm": 0.7539795087418848,
"learning_rate": 6.149553551363404e-06,
"loss": 0.0664,
"step": 479
},
{
"epoch": 0.8540925266903915,
"grad_norm": 0.7064668670179876,
"learning_rate": 6.1359483902046605e-06,
"loss": 0.0597,
"step": 480
},
{
"epoch": 0.8558718861209964,
"grad_norm": 0.6877440183143942,
"learning_rate": 6.122334354922984e-06,
"loss": 0.0649,
"step": 481
},
{
"epoch": 0.8576512455516014,
"grad_norm": 0.5628881042086711,
"learning_rate": 6.108711551872347e-06,
"loss": 0.0643,
"step": 482
},
{
"epoch": 0.8594306049822064,
"grad_norm": 0.8292765627391548,
"learning_rate": 6.095080087475218e-06,
"loss": 0.0741,
"step": 483
},
{
"epoch": 0.8612099644128114,
"grad_norm": 0.7008251314610137,
"learning_rate": 6.0814400682217236e-06,
"loss": 0.076,
"step": 484
},
{
"epoch": 0.8629893238434164,
"grad_norm": 0.5509293059848297,
"learning_rate": 6.067791600668823e-06,
"loss": 0.0408,
"step": 485
},
{
"epoch": 0.8647686832740213,
"grad_norm": 0.6283251083397737,
"learning_rate": 6.054134791439479e-06,
"loss": 0.0538,
"step": 486
},
{
"epoch": 0.8665480427046264,
"grad_norm": 0.7280882160407783,
"learning_rate": 6.040469747221815e-06,
"loss": 0.0674,
"step": 487
},
{
"epoch": 0.8683274021352313,
"grad_norm": 0.6264414289722235,
"learning_rate": 6.026796574768288e-06,
"loss": 0.0549,
"step": 488
},
{
"epoch": 0.8701067615658363,
"grad_norm": 0.6071972826698729,
"learning_rate": 6.013115380894854e-06,
"loss": 0.0538,
"step": 489
},
{
"epoch": 0.8718861209964412,
"grad_norm": 0.4868092261357424,
"learning_rate": 5.999426272480133e-06,
"loss": 0.041,
"step": 490
},
{
"epoch": 0.8736654804270463,
"grad_norm": 0.6183917526395925,
"learning_rate": 5.985729356464575e-06,
"loss": 0.0663,
"step": 491
},
{
"epoch": 0.8754448398576512,
"grad_norm": 0.6201415258159154,
"learning_rate": 5.972024739849622e-06,
"loss": 0.0643,
"step": 492
},
{
"epoch": 0.8772241992882562,
"grad_norm": 0.45032975185909824,
"learning_rate": 5.958312529696874e-06,
"loss": 0.0363,
"step": 493
},
{
"epoch": 0.8790035587188612,
"grad_norm": 0.6570052049475356,
"learning_rate": 5.944592833127253e-06,
"loss": 0.0712,
"step": 494
},
{
"epoch": 0.8807829181494662,
"grad_norm": 0.6992426127421228,
"learning_rate": 5.9308657573201645e-06,
"loss": 0.0758,
"step": 495
},
{
"epoch": 0.8825622775800712,
"grad_norm": 0.6016813631557385,
"learning_rate": 5.917131409512663e-06,
"loss": 0.056,
"step": 496
},
{
"epoch": 0.8843416370106761,
"grad_norm": 0.6988562844795229,
"learning_rate": 5.903389896998611e-06,
"loss": 0.0576,
"step": 497
},
{
"epoch": 0.8861209964412812,
"grad_norm": 0.6994123371717575,
"learning_rate": 5.889641327127843e-06,
"loss": 0.0634,
"step": 498
},
{
"epoch": 0.8879003558718861,
"grad_norm": 0.5803170933678522,
"learning_rate": 5.875885807305326e-06,
"loss": 0.0544,
"step": 499
},
{
"epoch": 0.8896797153024911,
"grad_norm": 0.6293090970385533,
"learning_rate": 5.862123444990319e-06,
"loss": 0.0612,
"step": 500
},
{
"epoch": 0.8896797153024911,
"eval_loss": 0.07298383861780167,
"eval_runtime": 1.9274,
"eval_samples_per_second": 23.867,
"eval_steps_per_second": 6.226,
"step": 500
},
{
"epoch": 0.891459074733096,
"grad_norm": 0.5682636402667952,
"learning_rate": 5.848354347695537e-06,
"loss": 0.0662,
"step": 501
},
{
"epoch": 0.8932384341637011,
"grad_norm": 0.6639439735702557,
"learning_rate": 5.83457862298631e-06,
"loss": 0.067,
"step": 502
},
{
"epoch": 0.895017793594306,
"grad_norm": 0.7268688453082539,
"learning_rate": 5.8207963784797396e-06,
"loss": 0.057,
"step": 503
},
{
"epoch": 0.896797153024911,
"grad_norm": 0.6193018816563773,
"learning_rate": 5.807007721843862e-06,
"loss": 0.076,
"step": 504
},
{
"epoch": 0.8985765124555161,
"grad_norm": 0.6316054501319838,
"learning_rate": 5.793212760796804e-06,
"loss": 0.0548,
"step": 505
},
{
"epoch": 0.900355871886121,
"grad_norm": 0.783022021001074,
"learning_rate": 5.779411603105947e-06,
"loss": 0.0732,
"step": 506
},
{
"epoch": 0.902135231316726,
"grad_norm": 0.6273431009429979,
"learning_rate": 5.765604356587076e-06,
"loss": 0.0555,
"step": 507
},
{
"epoch": 0.9039145907473309,
"grad_norm": 0.6425377325026403,
"learning_rate": 5.751791129103545e-06,
"loss": 0.0661,
"step": 508
},
{
"epoch": 0.905693950177936,
"grad_norm": 0.7936713212867098,
"learning_rate": 5.737972028565431e-06,
"loss": 0.0877,
"step": 509
},
{
"epoch": 0.9074733096085409,
"grad_norm": 1.0263688863363118,
"learning_rate": 5.7241471629286934e-06,
"loss": 0.0997,
"step": 510
},
{
"epoch": 0.9092526690391459,
"grad_norm": 0.8423806886759521,
"learning_rate": 5.7103166401943276e-06,
"loss": 0.0825,
"step": 511
},
{
"epoch": 0.9110320284697508,
"grad_norm": 0.841455786927593,
"learning_rate": 5.696480568407523e-06,
"loss": 0.0746,
"step": 512
},
{
"epoch": 0.9128113879003559,
"grad_norm": 0.7565808062417881,
"learning_rate": 5.682639055656817e-06,
"loss": 0.0754,
"step": 513
},
{
"epoch": 0.9145907473309609,
"grad_norm": 0.9988966398860025,
"learning_rate": 5.668792210073255e-06,
"loss": 0.1031,
"step": 514
},
{
"epoch": 0.9163701067615658,
"grad_norm": 0.7563474883938855,
"learning_rate": 5.654940139829544e-06,
"loss": 0.0774,
"step": 515
},
{
"epoch": 0.9181494661921709,
"grad_norm": 0.6104890847005197,
"learning_rate": 5.641082953139201e-06,
"loss": 0.0585,
"step": 516
},
{
"epoch": 0.9199288256227758,
"grad_norm": 0.5227387576958703,
"learning_rate": 5.6272207582557195e-06,
"loss": 0.0561,
"step": 517
},
{
"epoch": 0.9217081850533808,
"grad_norm": 0.7499438627928076,
"learning_rate": 5.61335366347171e-06,
"loss": 0.0652,
"step": 518
},
{
"epoch": 0.9234875444839857,
"grad_norm": 0.6210683250090191,
"learning_rate": 5.599481777118071e-06,
"loss": 0.0606,
"step": 519
},
{
"epoch": 0.9252669039145908,
"grad_norm": 0.6476625620578094,
"learning_rate": 5.585605207563124e-06,
"loss": 0.0579,
"step": 520
},
{
"epoch": 0.9270462633451957,
"grad_norm": 0.5822469566879743,
"learning_rate": 5.571724063211782e-06,
"loss": 0.0529,
"step": 521
},
{
"epoch": 0.9288256227758007,
"grad_norm": 0.5040068333073178,
"learning_rate": 5.557838452504692e-06,
"loss": 0.0348,
"step": 522
},
{
"epoch": 0.9306049822064056,
"grad_norm": 0.9023390779850642,
"learning_rate": 5.5439484839173996e-06,
"loss": 0.0769,
"step": 523
},
{
"epoch": 0.9323843416370107,
"grad_norm": 0.5934034163042189,
"learning_rate": 5.530054265959486e-06,
"loss": 0.055,
"step": 524
},
{
"epoch": 0.9341637010676157,
"grad_norm": 0.6539384745244389,
"learning_rate": 5.516155907173735e-06,
"loss": 0.0581,
"step": 525
},
{
"epoch": 0.9359430604982206,
"grad_norm": 0.5246470564915263,
"learning_rate": 5.5022535161352764e-06,
"loss": 0.057,
"step": 526
},
{
"epoch": 0.9377224199288257,
"grad_norm": 0.725400446979043,
"learning_rate": 5.488347201450741e-06,
"loss": 0.0608,
"step": 527
},
{
"epoch": 0.9395017793594306,
"grad_norm": 0.6129830685953794,
"learning_rate": 5.47443707175741e-06,
"loss": 0.0575,
"step": 528
},
{
"epoch": 0.9412811387900356,
"grad_norm": 0.5948231168188022,
"learning_rate": 5.46052323572237e-06,
"loss": 0.0482,
"step": 529
},
{
"epoch": 0.9430604982206405,
"grad_norm": 0.6348494711831387,
"learning_rate": 5.446605802041662e-06,
"loss": 0.0615,
"step": 530
},
{
"epoch": 0.9448398576512456,
"grad_norm": 0.5683438037419098,
"learning_rate": 5.432684879439428e-06,
"loss": 0.0601,
"step": 531
},
{
"epoch": 0.9466192170818505,
"grad_norm": 0.7205477677786964,
"learning_rate": 5.418760576667071e-06,
"loss": 0.0517,
"step": 532
},
{
"epoch": 0.9483985765124555,
"grad_norm": 0.6874573080027957,
"learning_rate": 5.404833002502398e-06,
"loss": 0.0622,
"step": 533
},
{
"epoch": 0.9501779359430605,
"grad_norm": 0.7724526906742538,
"learning_rate": 5.39090226574877e-06,
"loss": 0.0612,
"step": 534
},
{
"epoch": 0.9519572953736655,
"grad_norm": 0.9675551601876531,
"learning_rate": 5.376968475234258e-06,
"loss": 0.0915,
"step": 535
},
{
"epoch": 0.9537366548042705,
"grad_norm": 0.5642798061868978,
"learning_rate": 5.363031739810787e-06,
"loss": 0.0594,
"step": 536
},
{
"epoch": 0.9555160142348754,
"grad_norm": 0.5368616450989243,
"learning_rate": 5.349092168353291e-06,
"loss": 0.0552,
"step": 537
},
{
"epoch": 0.9572953736654805,
"grad_norm": 0.6196114734830861,
"learning_rate": 5.335149869758855e-06,
"loss": 0.0513,
"step": 538
},
{
"epoch": 0.9590747330960854,
"grad_norm": 1.2435668714789774,
"learning_rate": 5.32120495294587e-06,
"loss": 0.068,
"step": 539
},
{
"epoch": 0.9608540925266904,
"grad_norm": 0.5837344755854572,
"learning_rate": 5.3072575268531835e-06,
"loss": 0.0569,
"step": 540
},
{
"epoch": 0.9626334519572953,
"grad_norm": 0.7688110179771342,
"learning_rate": 5.293307700439242e-06,
"loss": 0.0727,
"step": 541
},
{
"epoch": 0.9644128113879004,
"grad_norm": 0.7835251885336217,
"learning_rate": 5.2793555826812456e-06,
"loss": 0.0662,
"step": 542
},
{
"epoch": 0.9661921708185054,
"grad_norm": 0.4411254530256784,
"learning_rate": 5.265401282574294e-06,
"loss": 0.0303,
"step": 543
},
{
"epoch": 0.9679715302491103,
"grad_norm": 0.8434321135033864,
"learning_rate": 5.2514449091305375e-06,
"loss": 0.0857,
"step": 544
},
{
"epoch": 0.9697508896797153,
"grad_norm": 0.6358677882929963,
"learning_rate": 5.237486571378317e-06,
"loss": 0.0634,
"step": 545
},
{
"epoch": 0.9715302491103203,
"grad_norm": 0.5991205746694954,
"learning_rate": 5.22352637836133e-06,
"loss": 0.0513,
"step": 546
},
{
"epoch": 0.9733096085409253,
"grad_norm": 0.5179303859860476,
"learning_rate": 5.209564439137755e-06,
"loss": 0.052,
"step": 547
},
{
"epoch": 0.9750889679715302,
"grad_norm": 0.787237555471461,
"learning_rate": 5.195600862779421e-06,
"loss": 0.0828,
"step": 548
},
{
"epoch": 0.9768683274021353,
"grad_norm": 0.5452495632140479,
"learning_rate": 5.181635758370942e-06,
"loss": 0.0426,
"step": 549
},
{
"epoch": 0.9786476868327402,
"grad_norm": 0.6904119057070438,
"learning_rate": 5.167669235008871e-06,
"loss": 0.0715,
"step": 550
},
{
"epoch": 0.9804270462633452,
"grad_norm": 0.8295539284524873,
"learning_rate": 5.153701401800845e-06,
"loss": 0.0853,
"step": 551
},
{
"epoch": 0.9822064056939501,
"grad_norm": 0.7184223766368749,
"learning_rate": 5.139732367864736e-06,
"loss": 0.0678,
"step": 552
},
{
"epoch": 0.9839857651245552,
"grad_norm": 0.6520513489594142,
"learning_rate": 5.1257622423277934e-06,
"loss": 0.0601,
"step": 553
},
{
"epoch": 0.9857651245551602,
"grad_norm": 0.6577830754609172,
"learning_rate": 5.111791134325793e-06,
"loss": 0.0706,
"step": 554
},
{
"epoch": 0.9875444839857651,
"grad_norm": 0.7703315996831352,
"learning_rate": 5.097819153002192e-06,
"loss": 0.0784,
"step": 555
},
{
"epoch": 0.9893238434163701,
"grad_norm": 0.5380766735646697,
"learning_rate": 5.083846407507263e-06,
"loss": 0.0499,
"step": 556
},
{
"epoch": 0.9911032028469751,
"grad_norm": 0.8501478847380034,
"learning_rate": 5.0698730069972535e-06,
"loss": 0.0856,
"step": 557
},
{
"epoch": 0.9928825622775801,
"grad_norm": 0.6041315186131391,
"learning_rate": 5.055899060633524e-06,
"loss": 0.0521,
"step": 558
},
{
"epoch": 0.994661921708185,
"grad_norm": 0.6644356530220107,
"learning_rate": 5.041924677581702e-06,
"loss": 0.0763,
"step": 559
},
{
"epoch": 0.99644128113879,
"grad_norm": 0.5759052618815151,
"learning_rate": 5.0279499670108245e-06,
"loss": 0.0565,
"step": 560
},
{
"epoch": 0.998220640569395,
"grad_norm": 0.575120678946544,
"learning_rate": 5.013975038092491e-06,
"loss": 0.0629,
"step": 561
},
{
"epoch": 1.0,
"grad_norm": 0.5372084668812244,
"learning_rate": 5e-06,
"loss": 0.0457,
"step": 562
},
{
"epoch": 1.001779359430605,
"grad_norm": 0.5001655991754118,
"learning_rate": 4.98602496190751e-06,
"loss": 0.0385,
"step": 563
},
{
"epoch": 1.00355871886121,
"grad_norm": 0.38154470666969476,
"learning_rate": 4.9720500329891755e-06,
"loss": 0.0279,
"step": 564
},
{
"epoch": 1.0053380782918149,
"grad_norm": 0.44498006275293056,
"learning_rate": 4.9580753224183005e-06,
"loss": 0.0316,
"step": 565
},
{
"epoch": 1.00711743772242,
"grad_norm": 0.39043268237867323,
"learning_rate": 4.944100939366478e-06,
"loss": 0.0279,
"step": 566
},
{
"epoch": 1.008896797153025,
"grad_norm": 0.3588505807838101,
"learning_rate": 4.930126993002748e-06,
"loss": 0.0265,
"step": 567
},
{
"epoch": 1.01067615658363,
"grad_norm": 0.3308402823845662,
"learning_rate": 4.9161535924927375e-06,
"loss": 0.0259,
"step": 568
},
{
"epoch": 1.0124555160142348,
"grad_norm": 0.38264774439205845,
"learning_rate": 4.90218084699781e-06,
"loss": 0.0273,
"step": 569
},
{
"epoch": 1.0142348754448398,
"grad_norm": 0.6256983559984098,
"learning_rate": 4.888208865674208e-06,
"loss": 0.0455,
"step": 570
},
{
"epoch": 1.0160142348754448,
"grad_norm": 0.3900630779080784,
"learning_rate": 4.874237757672209e-06,
"loss": 0.0317,
"step": 571
},
{
"epoch": 1.0177935943060499,
"grad_norm": 0.4419008155726269,
"learning_rate": 4.8602676321352646e-06,
"loss": 0.0263,
"step": 572
},
{
"epoch": 1.019572953736655,
"grad_norm": 0.42323674285693136,
"learning_rate": 4.846298598199155e-06,
"loss": 0.0274,
"step": 573
},
{
"epoch": 1.0213523131672597,
"grad_norm": 0.3961781358164811,
"learning_rate": 4.832330764991131e-06,
"loss": 0.0277,
"step": 574
},
{
"epoch": 1.0231316725978647,
"grad_norm": 0.429918311908298,
"learning_rate": 4.81836424162906e-06,
"loss": 0.0324,
"step": 575
},
{
"epoch": 1.0249110320284698,
"grad_norm": 0.2921716807575885,
"learning_rate": 4.80439913722058e-06,
"loss": 0.0178,
"step": 576
},
{
"epoch": 1.0266903914590748,
"grad_norm": 0.3245201283086529,
"learning_rate": 4.790435560862247e-06,
"loss": 0.0212,
"step": 577
},
{
"epoch": 1.0284697508896796,
"grad_norm": 0.3516281830354792,
"learning_rate": 4.776473621638673e-06,
"loss": 0.0209,
"step": 578
},
{
"epoch": 1.0302491103202847,
"grad_norm": 0.3258455833043046,
"learning_rate": 4.762513428621684e-06,
"loss": 0.0203,
"step": 579
},
{
"epoch": 1.0320284697508897,
"grad_norm": 0.41638471774476854,
"learning_rate": 4.748555090869464e-06,
"loss": 0.0263,
"step": 580
},
{
"epoch": 1.0338078291814947,
"grad_norm": 0.30260692792372684,
"learning_rate": 4.734598717425706e-06,
"loss": 0.02,
"step": 581
},
{
"epoch": 1.0355871886120998,
"grad_norm": 0.5292256307809499,
"learning_rate": 4.720644417318755e-06,
"loss": 0.0324,
"step": 582
},
{
"epoch": 1.0373665480427046,
"grad_norm": 0.5326720236337824,
"learning_rate": 4.70669229956076e-06,
"loss": 0.0309,
"step": 583
},
{
"epoch": 1.0391459074733096,
"grad_norm": 0.4124433349249828,
"learning_rate": 4.692742473146818e-06,
"loss": 0.0268,
"step": 584
},
{
"epoch": 1.0409252669039146,
"grad_norm": 0.46930655713283065,
"learning_rate": 4.678795047054131e-06,
"loss": 0.0338,
"step": 585
},
{
"epoch": 1.0427046263345197,
"grad_norm": 0.3570770990255731,
"learning_rate": 4.664850130241146e-06,
"loss": 0.0231,
"step": 586
},
{
"epoch": 1.0444839857651245,
"grad_norm": 0.29332867322560496,
"learning_rate": 4.650907831646711e-06,
"loss": 0.0163,
"step": 587
},
{
"epoch": 1.0462633451957295,
"grad_norm": 0.3951351943900231,
"learning_rate": 4.636968260189214e-06,
"loss": 0.0281,
"step": 588
},
{
"epoch": 1.0480427046263345,
"grad_norm": 0.36653963620805824,
"learning_rate": 4.623031524765744e-06,
"loss": 0.0242,
"step": 589
},
{
"epoch": 1.0498220640569396,
"grad_norm": 0.36870077391007183,
"learning_rate": 4.609097734251231e-06,
"loss": 0.0198,
"step": 590
},
{
"epoch": 1.0516014234875444,
"grad_norm": 0.4562061277027047,
"learning_rate": 4.595166997497605e-06,
"loss": 0.0253,
"step": 591
},
{
"epoch": 1.0533807829181494,
"grad_norm": 0.34372525692342787,
"learning_rate": 4.58123942333293e-06,
"loss": 0.0198,
"step": 592
},
{
"epoch": 1.0551601423487544,
"grad_norm": 0.4276794328468661,
"learning_rate": 4.567315120560573e-06,
"loss": 0.0278,
"step": 593
},
{
"epoch": 1.0569395017793595,
"grad_norm": 0.2936205050515641,
"learning_rate": 4.553394197958339e-06,
"loss": 0.0182,
"step": 594
},
{
"epoch": 1.0587188612099645,
"grad_norm": 0.3621443501794496,
"learning_rate": 4.539476764277631e-06,
"loss": 0.0205,
"step": 595
},
{
"epoch": 1.0604982206405693,
"grad_norm": 0.38708866905954004,
"learning_rate": 4.525562928242592e-06,
"loss": 0.0249,
"step": 596
},
{
"epoch": 1.0622775800711743,
"grad_norm": 0.45558098402927705,
"learning_rate": 4.511652798549261e-06,
"loss": 0.028,
"step": 597
},
{
"epoch": 1.0640569395017794,
"grad_norm": 0.38925015376240757,
"learning_rate": 4.497746483864725e-06,
"loss": 0.0208,
"step": 598
},
{
"epoch": 1.0658362989323844,
"grad_norm": 0.4341424098370843,
"learning_rate": 4.483844092826267e-06,
"loss": 0.0211,
"step": 599
},
{
"epoch": 1.0676156583629894,
"grad_norm": 0.31518195325427384,
"learning_rate": 4.469945734040516e-06,
"loss": 0.0188,
"step": 600
},
{
"epoch": 1.0693950177935942,
"grad_norm": 0.3924957226977899,
"learning_rate": 4.456051516082603e-06,
"loss": 0.0208,
"step": 601
},
{
"epoch": 1.0711743772241993,
"grad_norm": 0.49171024767163984,
"learning_rate": 4.442161547495309e-06,
"loss": 0.0297,
"step": 602
},
{
"epoch": 1.0729537366548043,
"grad_norm": 0.4761066750612135,
"learning_rate": 4.42827593678822e-06,
"loss": 0.0254,
"step": 603
},
{
"epoch": 1.0747330960854093,
"grad_norm": 0.451371727502243,
"learning_rate": 4.414394792436877e-06,
"loss": 0.0247,
"step": 604
},
{
"epoch": 1.0765124555160142,
"grad_norm": 0.48619354119553027,
"learning_rate": 4.400518222881931e-06,
"loss": 0.0309,
"step": 605
},
{
"epoch": 1.0782918149466192,
"grad_norm": 0.3898938299499558,
"learning_rate": 4.386646336528291e-06,
"loss": 0.0216,
"step": 606
},
{
"epoch": 1.0800711743772242,
"grad_norm": 0.6063780268544902,
"learning_rate": 4.372779241744282e-06,
"loss": 0.0294,
"step": 607
},
{
"epoch": 1.0818505338078293,
"grad_norm": 0.3790682261049408,
"learning_rate": 4.358917046860799e-06,
"loss": 0.0216,
"step": 608
},
{
"epoch": 1.083629893238434,
"grad_norm": 0.5897013185843848,
"learning_rate": 4.345059860170458e-06,
"loss": 0.0298,
"step": 609
},
{
"epoch": 1.085409252669039,
"grad_norm": 0.33642581481340955,
"learning_rate": 4.331207789926746e-06,
"loss": 0.0155,
"step": 610
},
{
"epoch": 1.0871886120996441,
"grad_norm": 0.3248294041702718,
"learning_rate": 4.317360944343184e-06,
"loss": 0.0174,
"step": 611
},
{
"epoch": 1.0889679715302492,
"grad_norm": 0.4196542237480155,
"learning_rate": 4.303519431592479e-06,
"loss": 0.0212,
"step": 612
},
{
"epoch": 1.0907473309608542,
"grad_norm": 0.5756669743063417,
"learning_rate": 4.289683359805673e-06,
"loss": 0.0286,
"step": 613
},
{
"epoch": 1.092526690391459,
"grad_norm": 0.4579399569178108,
"learning_rate": 4.275852837071309e-06,
"loss": 0.0209,
"step": 614
},
{
"epoch": 1.094306049822064,
"grad_norm": 0.3041588225973077,
"learning_rate": 4.26202797143457e-06,
"loss": 0.0152,
"step": 615
},
{
"epoch": 1.096085409252669,
"grad_norm": 0.511530677111693,
"learning_rate": 4.248208870896456e-06,
"loss": 0.0273,
"step": 616
},
{
"epoch": 1.097864768683274,
"grad_norm": 0.4312958240881722,
"learning_rate": 4.234395643412925e-06,
"loss": 0.0199,
"step": 617
},
{
"epoch": 1.099644128113879,
"grad_norm": 0.49439659343712655,
"learning_rate": 4.220588396894055e-06,
"loss": 0.0228,
"step": 618
},
{
"epoch": 1.101423487544484,
"grad_norm": 0.481749138733836,
"learning_rate": 4.2067872392031965e-06,
"loss": 0.0355,
"step": 619
},
{
"epoch": 1.103202846975089,
"grad_norm": 0.5208639949986622,
"learning_rate": 4.192992278156141e-06,
"loss": 0.025,
"step": 620
},
{
"epoch": 1.104982206405694,
"grad_norm": 0.5281381976566879,
"learning_rate": 4.179203621520262e-06,
"loss": 0.0294,
"step": 621
},
{
"epoch": 1.106761565836299,
"grad_norm": 0.39964326817540397,
"learning_rate": 4.165421377013691e-06,
"loss": 0.0139,
"step": 622
},
{
"epoch": 1.1085409252669038,
"grad_norm": 0.4993837857187084,
"learning_rate": 4.151645652304465e-06,
"loss": 0.0261,
"step": 623
},
{
"epoch": 1.1103202846975089,
"grad_norm": 0.5030711070218273,
"learning_rate": 4.137876555009684e-06,
"loss": 0.0254,
"step": 624
},
{
"epoch": 1.112099644128114,
"grad_norm": 0.6281844073263249,
"learning_rate": 4.124114192694676e-06,
"loss": 0.0393,
"step": 625
},
{
"epoch": 1.113879003558719,
"grad_norm": 0.47279352326949575,
"learning_rate": 4.110358672872158e-06,
"loss": 0.0257,
"step": 626
},
{
"epoch": 1.1156583629893237,
"grad_norm": 0.4390309982206838,
"learning_rate": 4.0966101030013915e-06,
"loss": 0.0222,
"step": 627
},
{
"epoch": 1.1174377224199288,
"grad_norm": 0.4145013107009972,
"learning_rate": 4.082868590487339e-06,
"loss": 0.0201,
"step": 628
},
{
"epoch": 1.1192170818505338,
"grad_norm": 0.2798217126199585,
"learning_rate": 4.069134242679837e-06,
"loss": 0.0139,
"step": 629
},
{
"epoch": 1.1209964412811388,
"grad_norm": 0.4945011824822153,
"learning_rate": 4.055407166872748e-06,
"loss": 0.0218,
"step": 630
},
{
"epoch": 1.1227758007117439,
"grad_norm": 0.4574736645891099,
"learning_rate": 4.041687470303127e-06,
"loss": 0.0273,
"step": 631
},
{
"epoch": 1.1245551601423487,
"grad_norm": 0.4487313799941942,
"learning_rate": 4.02797526015038e-06,
"loss": 0.028,
"step": 632
},
{
"epoch": 1.1263345195729537,
"grad_norm": 0.4443117699515725,
"learning_rate": 4.014270643535427e-06,
"loss": 0.0191,
"step": 633
},
{
"epoch": 1.1281138790035588,
"grad_norm": 0.4366831647127634,
"learning_rate": 4.000573727519868e-06,
"loss": 0.0216,
"step": 634
},
{
"epoch": 1.1298932384341638,
"grad_norm": 0.4055176179580648,
"learning_rate": 3.9868846191051465e-06,
"loss": 0.019,
"step": 635
},
{
"epoch": 1.1316725978647686,
"grad_norm": 0.49748088363447035,
"learning_rate": 3.973203425231715e-06,
"loss": 0.0256,
"step": 636
},
{
"epoch": 1.1334519572953736,
"grad_norm": 0.4136208408236497,
"learning_rate": 3.959530252778187e-06,
"loss": 0.0268,
"step": 637
},
{
"epoch": 1.1352313167259787,
"grad_norm": 0.5230492645217799,
"learning_rate": 3.945865208560522e-06,
"loss": 0.0254,
"step": 638
},
{
"epoch": 1.1370106761565837,
"grad_norm": 0.5501055507128955,
"learning_rate": 3.932208399331177e-06,
"loss": 0.0292,
"step": 639
},
{
"epoch": 1.1387900355871885,
"grad_norm": 0.5499646183015098,
"learning_rate": 3.918559931778277e-06,
"loss": 0.028,
"step": 640
},
{
"epoch": 1.1405693950177935,
"grad_norm": 0.3775244944838647,
"learning_rate": 3.904919912524784e-06,
"loss": 0.0169,
"step": 641
},
{
"epoch": 1.1423487544483986,
"grad_norm": 0.3485170834529513,
"learning_rate": 3.891288448127654e-06,
"loss": 0.0179,
"step": 642
},
{
"epoch": 1.1441281138790036,
"grad_norm": 0.45316553478728105,
"learning_rate": 3.877665645077017e-06,
"loss": 0.0216,
"step": 643
},
{
"epoch": 1.1459074733096086,
"grad_norm": 0.433647410334812,
"learning_rate": 3.86405160979534e-06,
"loss": 0.0196,
"step": 644
},
{
"epoch": 1.1476868327402134,
"grad_norm": 0.3855993219839802,
"learning_rate": 3.850446448636597e-06,
"loss": 0.0158,
"step": 645
},
{
"epoch": 1.1494661921708185,
"grad_norm": 0.4138255589972193,
"learning_rate": 3.8368502678854296e-06,
"loss": 0.0167,
"step": 646
},
{
"epoch": 1.1512455516014235,
"grad_norm": 0.3228609370829825,
"learning_rate": 3.8232631737563306e-06,
"loss": 0.0182,
"step": 647
},
{
"epoch": 1.1530249110320285,
"grad_norm": 0.5041353006154433,
"learning_rate": 3.809685272392804e-06,
"loss": 0.0242,
"step": 648
},
{
"epoch": 1.1548042704626336,
"grad_norm": 0.38730171138123676,
"learning_rate": 3.796116669866543e-06,
"loss": 0.0179,
"step": 649
},
{
"epoch": 1.1565836298932384,
"grad_norm": 0.48333658616256264,
"learning_rate": 3.78255747217659e-06,
"loss": 0.023,
"step": 650
},
{
"epoch": 1.1583629893238434,
"grad_norm": 0.4992512508088218,
"learning_rate": 3.769007785248523e-06,
"loss": 0.0249,
"step": 651
},
{
"epoch": 1.1601423487544484,
"grad_norm": 0.45886273301057917,
"learning_rate": 3.7554677149336186e-06,
"loss": 0.0203,
"step": 652
},
{
"epoch": 1.1619217081850535,
"grad_norm": 0.4675725698180337,
"learning_rate": 3.7419373670080284e-06,
"loss": 0.0202,
"step": 653
},
{
"epoch": 1.1637010676156583,
"grad_norm": 0.5028084844783673,
"learning_rate": 3.7284168471719527e-06,
"loss": 0.0232,
"step": 654
},
{
"epoch": 1.1654804270462633,
"grad_norm": 0.48095533972009663,
"learning_rate": 3.7149062610488085e-06,
"loss": 0.0273,
"step": 655
},
{
"epoch": 1.1672597864768683,
"grad_norm": 0.5019673310396005,
"learning_rate": 3.701405714184416e-06,
"loss": 0.0225,
"step": 656
},
{
"epoch": 1.1690391459074734,
"grad_norm": 0.32822061468660285,
"learning_rate": 3.687915312046166e-06,
"loss": 0.0153,
"step": 657
},
{
"epoch": 1.1708185053380782,
"grad_norm": 0.5901048612017039,
"learning_rate": 3.6744351600221994e-06,
"loss": 0.0202,
"step": 658
},
{
"epoch": 1.1725978647686832,
"grad_norm": 0.46632793735709954,
"learning_rate": 3.6609653634205773e-06,
"loss": 0.0313,
"step": 659
},
{
"epoch": 1.1743772241992882,
"grad_norm": 0.6317727989427925,
"learning_rate": 3.647506027468467e-06,
"loss": 0.022,
"step": 660
},
{
"epoch": 1.1761565836298933,
"grad_norm": 0.3830800859911667,
"learning_rate": 3.6340572573113176e-06,
"loss": 0.0183,
"step": 661
},
{
"epoch": 1.1779359430604983,
"grad_norm": 0.3822326187171165,
"learning_rate": 3.6206191580120346e-06,
"loss": 0.02,
"step": 662
},
{
"epoch": 1.1797153024911031,
"grad_norm": 0.5618269541774977,
"learning_rate": 3.6071918345501655e-06,
"loss": 0.027,
"step": 663
},
{
"epoch": 1.1814946619217082,
"grad_norm": 0.3964165268999999,
"learning_rate": 3.5937753918210705e-06,
"loss": 0.0167,
"step": 664
},
{
"epoch": 1.1832740213523132,
"grad_norm": 0.541698409269263,
"learning_rate": 3.5803699346351117e-06,
"loss": 0.017,
"step": 665
},
{
"epoch": 1.1850533807829182,
"grad_norm": 0.3373750970069219,
"learning_rate": 3.566975567716833e-06,
"loss": 0.015,
"step": 666
},
{
"epoch": 1.1868327402135233,
"grad_norm": 0.4675125638084268,
"learning_rate": 3.5535923957041374e-06,
"loss": 0.0199,
"step": 667
},
{
"epoch": 1.188612099644128,
"grad_norm": 0.5254026423059593,
"learning_rate": 3.540220523147474e-06,
"loss": 0.022,
"step": 668
},
{
"epoch": 1.190391459074733,
"grad_norm": 0.5629285573334403,
"learning_rate": 3.5268600545090183e-06,
"loss": 0.0198,
"step": 669
},
{
"epoch": 1.1921708185053381,
"grad_norm": 0.49234493401428336,
"learning_rate": 3.513511094161858e-06,
"loss": 0.0217,
"step": 670
},
{
"epoch": 1.193950177935943,
"grad_norm": 1.2909573215866423,
"learning_rate": 3.5001737463891793e-06,
"loss": 0.0399,
"step": 671
},
{
"epoch": 1.195729537366548,
"grad_norm": 0.4072249304190132,
"learning_rate": 3.4868481153834454e-06,
"loss": 0.0169,
"step": 672
},
{
"epoch": 1.197508896797153,
"grad_norm": 0.3813572823509891,
"learning_rate": 3.4735343052455905e-06,
"loss": 0.0132,
"step": 673
},
{
"epoch": 1.199288256227758,
"grad_norm": 0.618705633146954,
"learning_rate": 3.4602324199842026e-06,
"loss": 0.0306,
"step": 674
},
{
"epoch": 1.201067615658363,
"grad_norm": 0.4611625154639999,
"learning_rate": 3.446942563514711e-06,
"loss": 0.0215,
"step": 675
},
{
"epoch": 1.2028469750889679,
"grad_norm": 0.7963266121535011,
"learning_rate": 3.4336648396585777e-06,
"loss": 0.0213,
"step": 676
},
{
"epoch": 1.204626334519573,
"grad_norm": 0.39214564933481616,
"learning_rate": 3.4203993521424774e-06,
"loss": 0.0194,
"step": 677
},
{
"epoch": 1.206405693950178,
"grad_norm": 0.3333666384359917,
"learning_rate": 3.407146204597499e-06,
"loss": 0.0124,
"step": 678
},
{
"epoch": 1.208185053380783,
"grad_norm": 0.5486401689507133,
"learning_rate": 3.3939055005583305e-06,
"loss": 0.0276,
"step": 679
},
{
"epoch": 1.209964412811388,
"grad_norm": 0.7095609535370889,
"learning_rate": 3.3806773434624475e-06,
"loss": 0.0333,
"step": 680
},
{
"epoch": 1.2117437722419928,
"grad_norm": 0.5652092121760228,
"learning_rate": 3.3674618366493117e-06,
"loss": 0.0327,
"step": 681
},
{
"epoch": 1.2135231316725978,
"grad_norm": 0.6018134146917381,
"learning_rate": 3.3542590833595533e-06,
"loss": 0.0285,
"step": 682
},
{
"epoch": 1.2153024911032029,
"grad_norm": 0.6376321022452747,
"learning_rate": 3.341069186734176e-06,
"loss": 0.0241,
"step": 683
},
{
"epoch": 1.217081850533808,
"grad_norm": 0.38318249219940465,
"learning_rate": 3.3278922498137455e-06,
"loss": 0.0147,
"step": 684
},
{
"epoch": 1.2188612099644127,
"grad_norm": 0.6007758231003585,
"learning_rate": 3.314728375537587e-06,
"loss": 0.0281,
"step": 685
},
{
"epoch": 1.2206405693950177,
"grad_norm": 0.45001073979264594,
"learning_rate": 3.3015776667429724e-06,
"loss": 0.0237,
"step": 686
},
{
"epoch": 1.2224199288256228,
"grad_norm": 0.47919888077240347,
"learning_rate": 3.2884402261643296e-06,
"loss": 0.019,
"step": 687
},
{
"epoch": 1.2241992882562278,
"grad_norm": 0.5464571962635737,
"learning_rate": 3.2753161564324344e-06,
"loss": 0.0217,
"step": 688
},
{
"epoch": 1.2259786476868326,
"grad_norm": 0.7136843816838695,
"learning_rate": 3.262205560073605e-06,
"loss": 0.0237,
"step": 689
},
{
"epoch": 1.2277580071174377,
"grad_norm": 0.364314923687308,
"learning_rate": 3.249108539508909e-06,
"loss": 0.016,
"step": 690
},
{
"epoch": 1.2295373665480427,
"grad_norm": 0.5531559672367855,
"learning_rate": 3.2360251970533527e-06,
"loss": 0.0199,
"step": 691
},
{
"epoch": 1.2313167259786477,
"grad_norm": 0.4792095016958291,
"learning_rate": 3.2229556349150947e-06,
"loss": 0.0234,
"step": 692
},
{
"epoch": 1.2330960854092528,
"grad_norm": 0.45256265485240615,
"learning_rate": 3.2098999551946337e-06,
"loss": 0.0216,
"step": 693
},
{
"epoch": 1.2348754448398576,
"grad_norm": 0.5308112298963644,
"learning_rate": 3.1968582598840234e-06,
"loss": 0.0199,
"step": 694
},
{
"epoch": 1.2366548042704626,
"grad_norm": 0.4833030223070796,
"learning_rate": 3.183830650866068e-06,
"loss": 0.0186,
"step": 695
},
{
"epoch": 1.2384341637010676,
"grad_norm": 0.45297406731634793,
"learning_rate": 3.1708172299135266e-06,
"loss": 0.0217,
"step": 696
},
{
"epoch": 1.2402135231316727,
"grad_norm": 0.563069487172996,
"learning_rate": 3.1578180986883234e-06,
"loss": 0.0241,
"step": 697
},
{
"epoch": 1.2419928825622777,
"grad_norm": 0.4173506315881807,
"learning_rate": 3.1448333587407486e-06,
"loss": 0.0169,
"step": 698
},
{
"epoch": 1.2437722419928825,
"grad_norm": 0.46408869061113045,
"learning_rate": 3.131863111508667e-06,
"loss": 0.0207,
"step": 699
},
{
"epoch": 1.2455516014234875,
"grad_norm": 0.5193829271326897,
"learning_rate": 3.118907458316722e-06,
"loss": 0.0242,
"step": 700
},
{
"epoch": 1.2473309608540926,
"grad_norm": 0.4273272058290953,
"learning_rate": 3.105966500375551e-06,
"loss": 0.0174,
"step": 701
},
{
"epoch": 1.2491103202846976,
"grad_norm": 0.5433645115870799,
"learning_rate": 3.0930403387809892e-06,
"loss": 0.0263,
"step": 702
},
{
"epoch": 1.2508896797153026,
"grad_norm": 0.47009295029666054,
"learning_rate": 3.080129074513285e-06,
"loss": 0.0201,
"step": 703
},
{
"epoch": 1.2526690391459074,
"grad_norm": 0.5736617510520102,
"learning_rate": 3.067232808436299e-06,
"loss": 0.0278,
"step": 704
},
{
"epoch": 1.2544483985765125,
"grad_norm": 0.5538587887091225,
"learning_rate": 3.0543516412967327e-06,
"loss": 0.0278,
"step": 705
},
{
"epoch": 1.2562277580071175,
"grad_norm": 0.40592483676625524,
"learning_rate": 3.041485673723331e-06,
"loss": 0.0191,
"step": 706
},
{
"epoch": 1.2580071174377223,
"grad_norm": 0.3809116041625743,
"learning_rate": 3.0286350062261017e-06,
"loss": 0.0143,
"step": 707
},
{
"epoch": 1.2597864768683273,
"grad_norm": 0.46053210606084194,
"learning_rate": 3.0157997391955172e-06,
"loss": 0.0174,
"step": 708
},
{
"epoch": 1.2615658362989324,
"grad_norm": 0.5762997175536367,
"learning_rate": 3.0029799729017518e-06,
"loss": 0.0222,
"step": 709
},
{
"epoch": 1.2633451957295374,
"grad_norm": 0.5410977142435384,
"learning_rate": 2.9901758074938797e-06,
"loss": 0.0249,
"step": 710
},
{
"epoch": 1.2651245551601424,
"grad_norm": 0.7631112137371882,
"learning_rate": 2.977387342999103e-06,
"loss": 0.0345,
"step": 711
},
{
"epoch": 1.2669039145907472,
"grad_norm": 0.4671695768811203,
"learning_rate": 2.964614679321966e-06,
"loss": 0.0198,
"step": 712
},
{
"epoch": 1.2686832740213523,
"grad_norm": 0.5553679662308749,
"learning_rate": 2.951857916243574e-06,
"loss": 0.0203,
"step": 713
},
{
"epoch": 1.2704626334519573,
"grad_norm": 0.5917808581319681,
"learning_rate": 2.9391171534208185e-06,
"loss": 0.0254,
"step": 714
},
{
"epoch": 1.2722419928825623,
"grad_norm": 0.632214049636021,
"learning_rate": 2.9263924903855932e-06,
"loss": 0.0297,
"step": 715
},
{
"epoch": 1.2740213523131674,
"grad_norm": 0.641368521960054,
"learning_rate": 2.9136840265440213e-06,
"loss": 0.0236,
"step": 716
},
{
"epoch": 1.2758007117437722,
"grad_norm": 0.4754682221096714,
"learning_rate": 2.9009918611756732e-06,
"loss": 0.0192,
"step": 717
},
{
"epoch": 1.2775800711743772,
"grad_norm": 0.43337042623912403,
"learning_rate": 2.8883160934327968e-06,
"loss": 0.0161,
"step": 718
},
{
"epoch": 1.2793594306049823,
"grad_norm": 0.6583246135907656,
"learning_rate": 2.8756568223395396e-06,
"loss": 0.0286,
"step": 719
},
{
"epoch": 1.281138790035587,
"grad_norm": 0.5134051640075226,
"learning_rate": 2.8630141467911777e-06,
"loss": 0.022,
"step": 720
},
{
"epoch": 1.282918149466192,
"grad_norm": 0.539329925736207,
"learning_rate": 2.8503881655533395e-06,
"loss": 0.0167,
"step": 721
},
{
"epoch": 1.2846975088967971,
"grad_norm": 0.44715466679413357,
"learning_rate": 2.837778977261235e-06,
"loss": 0.0223,
"step": 722
},
{
"epoch": 1.2864768683274022,
"grad_norm": 0.5414773027672198,
"learning_rate": 2.8251866804188875e-06,
"loss": 0.0232,
"step": 723
},
{
"epoch": 1.2882562277580072,
"grad_norm": 0.5120286847309797,
"learning_rate": 2.812611373398365e-06,
"loss": 0.0259,
"step": 724
},
{
"epoch": 1.290035587188612,
"grad_norm": 0.6171158902707203,
"learning_rate": 2.8000531544390064e-06,
"loss": 0.0247,
"step": 725
},
{
"epoch": 1.291814946619217,
"grad_norm": 0.5134282112495162,
"learning_rate": 2.7875121216466595e-06,
"loss": 0.0204,
"step": 726
},
{
"epoch": 1.293594306049822,
"grad_norm": 0.5385203126185941,
"learning_rate": 2.7749883729929105e-06,
"loss": 0.017,
"step": 727
},
{
"epoch": 1.295373665480427,
"grad_norm": 0.4416968209605626,
"learning_rate": 2.762482006314324e-06,
"loss": 0.0166,
"step": 728
},
{
"epoch": 1.2971530249110321,
"grad_norm": 0.592396794403802,
"learning_rate": 2.7499931193116692e-06,
"loss": 0.0199,
"step": 729
},
{
"epoch": 1.298932384341637,
"grad_norm": 0.4324017771681348,
"learning_rate": 2.737521809549167e-06,
"loss": 0.0169,
"step": 730
},
{
"epoch": 1.300711743772242,
"grad_norm": 0.42177140376949085,
"learning_rate": 2.725068174453722e-06,
"loss": 0.017,
"step": 731
},
{
"epoch": 1.302491103202847,
"grad_norm": 0.5495481275730962,
"learning_rate": 2.712632311314165e-06,
"loss": 0.0205,
"step": 732
},
{
"epoch": 1.304270462633452,
"grad_norm": 0.7460401447084133,
"learning_rate": 2.7002143172804875e-06,
"loss": 0.0379,
"step": 733
},
{
"epoch": 1.306049822064057,
"grad_norm": 0.43318763568465063,
"learning_rate": 2.6878142893630904e-06,
"loss": 0.0175,
"step": 734
},
{
"epoch": 1.3078291814946619,
"grad_norm": 0.47921107784031003,
"learning_rate": 2.6754323244320154e-06,
"loss": 0.0198,
"step": 735
},
{
"epoch": 1.309608540925267,
"grad_norm": 0.7049182265261095,
"learning_rate": 2.6630685192161995e-06,
"loss": 0.0322,
"step": 736
},
{
"epoch": 1.311387900355872,
"grad_norm": 0.5954822421699036,
"learning_rate": 2.650722970302714e-06,
"loss": 0.0184,
"step": 737
},
{
"epoch": 1.3131672597864767,
"grad_norm": 0.624289677249861,
"learning_rate": 2.638395774136009e-06,
"loss": 0.0231,
"step": 738
},
{
"epoch": 1.3149466192170818,
"grad_norm": 0.858401952966649,
"learning_rate": 2.6260870270171645e-06,
"loss": 0.0272,
"step": 739
},
{
"epoch": 1.3167259786476868,
"grad_norm": 0.5079667300186227,
"learning_rate": 2.613796825103129e-06,
"loss": 0.0248,
"step": 740
},
{
"epoch": 1.3185053380782918,
"grad_norm": 0.6612579187095262,
"learning_rate": 2.60152526440598e-06,
"loss": 0.0255,
"step": 741
},
{
"epoch": 1.3202846975088969,
"grad_norm": 0.4215894099628918,
"learning_rate": 2.5892724407921667e-06,
"loss": 0.0164,
"step": 742
},
{
"epoch": 1.3220640569395017,
"grad_norm": 0.7998616288069056,
"learning_rate": 2.577038449981763e-06,
"loss": 0.0308,
"step": 743
},
{
"epoch": 1.3238434163701067,
"grad_norm": 0.6408435866143708,
"learning_rate": 2.564823387547716e-06,
"loss": 0.0273,
"step": 744
},
{
"epoch": 1.3256227758007118,
"grad_norm": 0.7334289656460562,
"learning_rate": 2.552627348915106e-06,
"loss": 0.037,
"step": 745
},
{
"epoch": 1.3274021352313168,
"grad_norm": 0.5737178029849802,
"learning_rate": 2.5404504293603983e-06,
"loss": 0.0245,
"step": 746
},
{
"epoch": 1.3291814946619218,
"grad_norm": 0.770944867558618,
"learning_rate": 2.528292724010697e-06,
"loss": 0.0354,
"step": 747
},
{
"epoch": 1.3309608540925266,
"grad_norm": 0.5111104274968983,
"learning_rate": 2.5161543278430055e-06,
"loss": 0.0224,
"step": 748
},
{
"epoch": 1.3327402135231317,
"grad_norm": 0.5005226536619571,
"learning_rate": 2.5040353356834756e-06,
"loss": 0.0165,
"step": 749
},
{
"epoch": 1.3345195729537367,
"grad_norm": 0.4353164103176695,
"learning_rate": 2.4919358422066816e-06,
"loss": 0.0193,
"step": 750
},
{
"epoch": 1.3362989323843417,
"grad_norm": 0.44406297273330664,
"learning_rate": 2.4798559419348672e-06,
"loss": 0.016,
"step": 751
},
{
"epoch": 1.3380782918149468,
"grad_norm": 0.3832738243405496,
"learning_rate": 2.4677957292372166e-06,
"loss": 0.0182,
"step": 752
},
{
"epoch": 1.3398576512455516,
"grad_norm": 0.3997483794874491,
"learning_rate": 2.455755298329107e-06,
"loss": 0.0147,
"step": 753
},
{
"epoch": 1.3416370106761566,
"grad_norm": 0.4209271167576176,
"learning_rate": 2.4437347432713838e-06,
"loss": 0.0196,
"step": 754
},
{
"epoch": 1.3434163701067616,
"grad_norm": 0.5441291606043887,
"learning_rate": 2.431734157969619e-06,
"loss": 0.0217,
"step": 755
},
{
"epoch": 1.3451957295373664,
"grad_norm": 0.48632163646916565,
"learning_rate": 2.4197536361733792e-06,
"loss": 0.0254,
"step": 756
},
{
"epoch": 1.3469750889679715,
"grad_norm": 0.4659983039673738,
"learning_rate": 2.407793271475495e-06,
"loss": 0.0211,
"step": 757
},
{
"epoch": 1.3487544483985765,
"grad_norm": 0.45178211282553266,
"learning_rate": 2.3958531573113223e-06,
"loss": 0.0284,
"step": 758
},
{
"epoch": 1.3505338078291815,
"grad_norm": 0.5249529700914765,
"learning_rate": 2.3839333869580243e-06,
"loss": 0.0234,
"step": 759
},
{
"epoch": 1.3523131672597866,
"grad_norm": 0.33804635314130616,
"learning_rate": 2.372034053533835e-06,
"loss": 0.0125,
"step": 760
},
{
"epoch": 1.3540925266903914,
"grad_norm": 0.45510346268677493,
"learning_rate": 2.360155249997334e-06,
"loss": 0.0219,
"step": 761
},
{
"epoch": 1.3558718861209964,
"grad_norm": 0.4290008263395087,
"learning_rate": 2.348297069146715e-06,
"loss": 0.0174,
"step": 762
},
{
"epoch": 1.3576512455516014,
"grad_norm": 0.39483425749383966,
"learning_rate": 2.3364596036190706e-06,
"loss": 0.0155,
"step": 763
},
{
"epoch": 1.3594306049822065,
"grad_norm": 0.6166553988132547,
"learning_rate": 2.3246429458896637e-06,
"loss": 0.0162,
"step": 764
},
{
"epoch": 1.3612099644128115,
"grad_norm": 0.5465245626521924,
"learning_rate": 2.312847188271203e-06,
"loss": 0.0327,
"step": 765
},
{
"epoch": 1.3629893238434163,
"grad_norm": 0.6006452797351305,
"learning_rate": 2.301072422913123e-06,
"loss": 0.0204,
"step": 766
},
{
"epoch": 1.3647686832740213,
"grad_norm": 0.5065696017993653,
"learning_rate": 2.2893187418008666e-06,
"loss": 0.0247,
"step": 767
},
{
"epoch": 1.3665480427046264,
"grad_norm": 0.5260397665616747,
"learning_rate": 2.2775862367551642e-06,
"loss": 0.0193,
"step": 768
},
{
"epoch": 1.3683274021352312,
"grad_norm": 0.5170754928488476,
"learning_rate": 2.265874999431318e-06,
"loss": 0.0196,
"step": 769
},
{
"epoch": 1.3701067615658362,
"grad_norm": 0.4948722067119337,
"learning_rate": 2.254185121318484e-06,
"loss": 0.0177,
"step": 770
},
{
"epoch": 1.3718861209964412,
"grad_norm": 0.7938192669486235,
"learning_rate": 2.2425166937389596e-06,
"loss": 0.0271,
"step": 771
},
{
"epoch": 1.3736654804270463,
"grad_norm": 0.46591544684027353,
"learning_rate": 2.2308698078474645e-06,
"loss": 0.0226,
"step": 772
},
{
"epoch": 1.3754448398576513,
"grad_norm": 0.6195771759864835,
"learning_rate": 2.219244554630438e-06,
"loss": 0.022,
"step": 773
},
{
"epoch": 1.3772241992882561,
"grad_norm": 0.6545114739851292,
"learning_rate": 2.207641024905322e-06,
"loss": 0.0375,
"step": 774
},
{
"epoch": 1.3790035587188612,
"grad_norm": 0.4294196859720872,
"learning_rate": 2.1960593093198508e-06,
"loss": 0.0131,
"step": 775
},
{
"epoch": 1.3807829181494662,
"grad_norm": 0.46518500738403457,
"learning_rate": 2.184499498351347e-06,
"loss": 0.0175,
"step": 776
},
{
"epoch": 1.3825622775800712,
"grad_norm": 0.4380215625025954,
"learning_rate": 2.172961682306011e-06,
"loss": 0.0196,
"step": 777
},
{
"epoch": 1.3843416370106763,
"grad_norm": 0.4816919299339168,
"learning_rate": 2.1614459513182173e-06,
"loss": 0.0226,
"step": 778
},
{
"epoch": 1.386120996441281,
"grad_norm": 0.5003285694844123,
"learning_rate": 2.149952395349813e-06,
"loss": 0.0193,
"step": 779
},
{
"epoch": 1.387900355871886,
"grad_norm": 0.6171675802940203,
"learning_rate": 2.1384811041894055e-06,
"loss": 0.0295,
"step": 780
},
{
"epoch": 1.3896797153024911,
"grad_norm": 0.4407408822277648,
"learning_rate": 2.1270321674516736e-06,
"loss": 0.0199,
"step": 781
},
{
"epoch": 1.3914590747330962,
"grad_norm": 0.5190637961840711,
"learning_rate": 2.1156056745766593e-06,
"loss": 0.0196,
"step": 782
},
{
"epoch": 1.3932384341637012,
"grad_norm": 0.5360574310722005,
"learning_rate": 2.104201714829074e-06,
"loss": 0.0229,
"step": 783
},
{
"epoch": 1.395017793594306,
"grad_norm": 0.5322955603619314,
"learning_rate": 2.0928203772975917e-06,
"loss": 0.0287,
"step": 784
},
{
"epoch": 1.396797153024911,
"grad_norm": 0.5456075966352554,
"learning_rate": 2.081461750894166e-06,
"loss": 0.0226,
"step": 785
},
{
"epoch": 1.398576512455516,
"grad_norm": 0.43454195972109483,
"learning_rate": 2.070125924353328e-06,
"loss": 0.0147,
"step": 786
},
{
"epoch": 1.4003558718861209,
"grad_norm": 0.40041120595663793,
"learning_rate": 2.058812986231493e-06,
"loss": 0.0144,
"step": 787
},
{
"epoch": 1.402135231316726,
"grad_norm": 0.6030472238548137,
"learning_rate": 2.0475230249062727e-06,
"loss": 0.0292,
"step": 788
},
{
"epoch": 1.403914590747331,
"grad_norm": 0.706143644579018,
"learning_rate": 2.0362561285757766e-06,
"loss": 0.0341,
"step": 789
},
{
"epoch": 1.405693950177936,
"grad_norm": 0.6538957024858132,
"learning_rate": 2.0250123852579347e-06,
"loss": 0.0222,
"step": 790
},
{
"epoch": 1.407473309608541,
"grad_norm": 0.36789802627358015,
"learning_rate": 2.013791882789801e-06,
"loss": 0.0159,
"step": 791
},
{
"epoch": 1.4092526690391458,
"grad_norm": 0.520107342350963,
"learning_rate": 2.0025947088268714e-06,
"loss": 0.0182,
"step": 792
},
{
"epoch": 1.4110320284697508,
"grad_norm": 0.6180866178897662,
"learning_rate": 1.9914209508423943e-06,
"loss": 0.0258,
"step": 793
},
{
"epoch": 1.4128113879003559,
"grad_norm": 0.41197617001982345,
"learning_rate": 1.9802706961266936e-06,
"loss": 0.017,
"step": 794
},
{
"epoch": 1.414590747330961,
"grad_norm": 0.6601949592765903,
"learning_rate": 1.969144031786483e-06,
"loss": 0.037,
"step": 795
},
{
"epoch": 1.416370106761566,
"grad_norm": 0.5303804645212132,
"learning_rate": 1.958041044744186e-06,
"loss": 0.0243,
"step": 796
},
{
"epoch": 1.4181494661921707,
"grad_norm": 0.4221202198921774,
"learning_rate": 1.94696182173726e-06,
"loss": 0.0159,
"step": 797
},
{
"epoch": 1.4199288256227758,
"grad_norm": 0.48425994204487605,
"learning_rate": 1.9359064493175077e-06,
"loss": 0.0193,
"step": 798
},
{
"epoch": 1.4217081850533808,
"grad_norm": 0.4379234430938787,
"learning_rate": 1.9248750138504176e-06,
"loss": 0.0186,
"step": 799
},
{
"epoch": 1.4234875444839858,
"grad_norm": 0.490522324383623,
"learning_rate": 1.9138676015144765e-06,
"loss": 0.0176,
"step": 800
},
{
"epoch": 1.4252669039145909,
"grad_norm": 0.5195896886907801,
"learning_rate": 1.9028842983005036e-06,
"loss": 0.0268,
"step": 801
},
{
"epoch": 1.4270462633451957,
"grad_norm": 0.6134552522986715,
"learning_rate": 1.8919251900109697e-06,
"loss": 0.0314,
"step": 802
},
{
"epoch": 1.4288256227758007,
"grad_norm": 0.35783173710430316,
"learning_rate": 1.8809903622593395e-06,
"loss": 0.0152,
"step": 803
},
{
"epoch": 1.4306049822064058,
"grad_norm": 0.47187731265748467,
"learning_rate": 1.870079900469392e-06,
"loss": 0.0226,
"step": 804
},
{
"epoch": 1.4323843416370106,
"grad_norm": 0.49008349747565916,
"learning_rate": 1.8591938898745593e-06,
"loss": 0.0232,
"step": 805
},
{
"epoch": 1.4341637010676156,
"grad_norm": 0.3792494939169566,
"learning_rate": 1.8483324155172594e-06,
"loss": 0.0114,
"step": 806
},
{
"epoch": 1.4359430604982206,
"grad_norm": 0.47217446266737334,
"learning_rate": 1.837495562248226e-06,
"loss": 0.0198,
"step": 807
},
{
"epoch": 1.4377224199288257,
"grad_norm": 0.43238438735396173,
"learning_rate": 1.8266834147258577e-06,
"loss": 0.0157,
"step": 808
},
{
"epoch": 1.4395017793594307,
"grad_norm": 0.4910245024478601,
"learning_rate": 1.8158960574155455e-06,
"loss": 0.0172,
"step": 809
},
{
"epoch": 1.4412811387900355,
"grad_norm": 0.61893105111085,
"learning_rate": 1.8051335745890196e-06,
"loss": 0.0226,
"step": 810
},
{
"epoch": 1.4430604982206405,
"grad_norm": 0.3710869412596552,
"learning_rate": 1.7943960503236856e-06,
"loss": 0.0139,
"step": 811
},
{
"epoch": 1.4448398576512456,
"grad_norm": 0.43314226770459807,
"learning_rate": 1.7836835685019732e-06,
"loss": 0.017,
"step": 812
},
{
"epoch": 1.4466192170818506,
"grad_norm": 0.42678354465733004,
"learning_rate": 1.7729962128106787e-06,
"loss": 0.0183,
"step": 813
},
{
"epoch": 1.4483985765124556,
"grad_norm": 0.45048607608096025,
"learning_rate": 1.7623340667403089e-06,
"loss": 0.0185,
"step": 814
},
{
"epoch": 1.4501779359430604,
"grad_norm": 0.6674423054720211,
"learning_rate": 1.7516972135844352e-06,
"loss": 0.0266,
"step": 815
},
{
"epoch": 1.4519572953736655,
"grad_norm": 0.604385421573245,
"learning_rate": 1.741085736439031e-06,
"loss": 0.0235,
"step": 816
},
{
"epoch": 1.4537366548042705,
"grad_norm": 0.5327606343352189,
"learning_rate": 1.730499718201838e-06,
"loss": 0.0209,
"step": 817
},
{
"epoch": 1.4555160142348753,
"grad_norm": 0.6801389549736783,
"learning_rate": 1.7199392415717064e-06,
"loss": 0.0271,
"step": 818
},
{
"epoch": 1.4572953736654806,
"grad_norm": 0.47906779259644433,
"learning_rate": 1.7094043890479557e-06,
"loss": 0.0238,
"step": 819
},
{
"epoch": 1.4590747330960854,
"grad_norm": 0.4255317836386466,
"learning_rate": 1.698895242929725e-06,
"loss": 0.0174,
"step": 820
},
{
"epoch": 1.4608540925266904,
"grad_norm": 0.5097865407665956,
"learning_rate": 1.6884118853153358e-06,
"loss": 0.0255,
"step": 821
},
{
"epoch": 1.4626334519572954,
"grad_norm": 1.1345695826275857,
"learning_rate": 1.6779543981016478e-06,
"loss": 0.0402,
"step": 822
},
{
"epoch": 1.4644128113879002,
"grad_norm": 0.48242007122289937,
"learning_rate": 1.6675228629834133e-06,
"loss": 0.0243,
"step": 823
},
{
"epoch": 1.4661921708185053,
"grad_norm": 0.3666833449944579,
"learning_rate": 1.657117361452651e-06,
"loss": 0.016,
"step": 824
},
{
"epoch": 1.4679715302491103,
"grad_norm": 0.40523596627166125,
"learning_rate": 1.6467379747980011e-06,
"loss": 0.0205,
"step": 825
},
{
"epoch": 1.4697508896797153,
"grad_norm": 0.34467970545350696,
"learning_rate": 1.6363847841040914e-06,
"loss": 0.0118,
"step": 826
},
{
"epoch": 1.4715302491103204,
"grad_norm": 0.5007611855778333,
"learning_rate": 1.626057870250906e-06,
"loss": 0.0246,
"step": 827
},
{
"epoch": 1.4733096085409252,
"grad_norm": 0.3656396606475968,
"learning_rate": 1.6157573139131527e-06,
"loss": 0.013,
"step": 828
},
{
"epoch": 1.4750889679715302,
"grad_norm": 0.4117031903860015,
"learning_rate": 1.605483195559628e-06,
"loss": 0.0166,
"step": 829
},
{
"epoch": 1.4768683274021353,
"grad_norm": 0.5476893042638611,
"learning_rate": 1.5952355954525966e-06,
"loss": 0.0202,
"step": 830
},
{
"epoch": 1.4786476868327403,
"grad_norm": 0.5246521970923796,
"learning_rate": 1.5850145936471607e-06,
"loss": 0.0204,
"step": 831
},
{
"epoch": 1.4804270462633453,
"grad_norm": 0.7093224262739016,
"learning_rate": 1.5748202699906335e-06,
"loss": 0.0305,
"step": 832
},
{
"epoch": 1.4822064056939501,
"grad_norm": 0.6402753849526811,
"learning_rate": 1.5646527041219128e-06,
"loss": 0.0243,
"step": 833
},
{
"epoch": 1.4839857651245552,
"grad_norm": 0.45839667870918444,
"learning_rate": 1.5545119754708682e-06,
"loss": 0.0174,
"step": 834
},
{
"epoch": 1.4857651245551602,
"grad_norm": 0.718632210924249,
"learning_rate": 1.544398163257711e-06,
"loss": 0.0265,
"step": 835
},
{
"epoch": 1.487544483985765,
"grad_norm": 0.5790513223856332,
"learning_rate": 1.5343113464923808e-06,
"loss": 0.0244,
"step": 836
},
{
"epoch": 1.48932384341637,
"grad_norm": 0.38032087738754,
"learning_rate": 1.524251603973927e-06,
"loss": 0.0172,
"step": 837
},
{
"epoch": 1.491103202846975,
"grad_norm": 0.5239443287560103,
"learning_rate": 1.5142190142898883e-06,
"loss": 0.02,
"step": 838
},
{
"epoch": 1.49288256227758,
"grad_norm": 0.46495922418555236,
"learning_rate": 1.5042136558156883e-06,
"loss": 0.019,
"step": 839
},
{
"epoch": 1.4946619217081851,
"grad_norm": 0.5542948868713746,
"learning_rate": 1.4942356067140162e-06,
"loss": 0.024,
"step": 840
},
{
"epoch": 1.49644128113879,
"grad_norm": 0.41037196421819966,
"learning_rate": 1.4842849449342195e-06,
"loss": 0.0158,
"step": 841
},
{
"epoch": 1.498220640569395,
"grad_norm": 0.4820571678279581,
"learning_rate": 1.4743617482116896e-06,
"loss": 0.0205,
"step": 842
},
{
"epoch": 1.5,
"grad_norm": 0.5034322268240278,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.0182,
"step": 843
},
{
"epoch": 1.501779359430605,
"grad_norm": 0.34839108779550243,
"learning_rate": 1.454598059806609e-06,
"loss": 0.0132,
"step": 844
},
{
"epoch": 1.50355871886121,
"grad_norm": 0.5540170675375056,
"learning_rate": 1.4447577225196296e-06,
"loss": 0.0229,
"step": 845
},
{
"epoch": 1.5053380782918149,
"grad_norm": 0.5069549958553468,
"learning_rate": 1.4349451590798564e-06,
"loss": 0.0204,
"step": 846
},
{
"epoch": 1.50711743772242,
"grad_norm": 0.5579393160230254,
"learning_rate": 1.4251604461438444e-06,
"loss": 0.019,
"step": 847
},
{
"epoch": 1.508896797153025,
"grad_norm": 0.4249633498077031,
"learning_rate": 1.4154036601505834e-06,
"loss": 0.015,
"step": 848
},
{
"epoch": 1.5106761565836297,
"grad_norm": 0.5720345706006639,
"learning_rate": 1.4056748773208933e-06,
"loss": 0.0216,
"step": 849
},
{
"epoch": 1.512455516014235,
"grad_norm": 0.5150797528703756,
"learning_rate": 1.3959741736568339e-06,
"loss": 0.0198,
"step": 850
},
{
"epoch": 1.5142348754448398,
"grad_norm": 0.45789045528714983,
"learning_rate": 1.3863016249411027e-06,
"loss": 0.0198,
"step": 851
},
{
"epoch": 1.5160142348754448,
"grad_norm": 0.4842344594520872,
"learning_rate": 1.376657306736453e-06,
"loss": 0.02,
"step": 852
},
{
"epoch": 1.5177935943060499,
"grad_norm": 0.5953213467761111,
"learning_rate": 1.3670412943850975e-06,
"loss": 0.0253,
"step": 853
},
{
"epoch": 1.5195729537366547,
"grad_norm": 0.566319367811969,
"learning_rate": 1.3574536630081208e-06,
"loss": 0.0241,
"step": 854
},
{
"epoch": 1.52135231316726,
"grad_norm": 0.5031043521098744,
"learning_rate": 1.347894487504896e-06,
"loss": 0.0249,
"step": 855
},
{
"epoch": 1.5231316725978647,
"grad_norm": 0.4845713391408967,
"learning_rate": 1.3383638425524909e-06,
"loss": 0.0181,
"step": 856
},
{
"epoch": 1.5249110320284698,
"grad_norm": 0.46143421257280676,
"learning_rate": 1.3288618026050943e-06,
"loss": 0.02,
"step": 857
},
{
"epoch": 1.5266903914590748,
"grad_norm": 0.6152816196336514,
"learning_rate": 1.31938844189343e-06,
"loss": 0.0271,
"step": 858
},
{
"epoch": 1.5284697508896796,
"grad_norm": 0.5608714018375411,
"learning_rate": 1.3099438344241777e-06,
"loss": 0.0238,
"step": 859
},
{
"epoch": 1.5302491103202847,
"grad_norm": 0.4430582873937365,
"learning_rate": 1.3005280539793908e-06,
"loss": 0.019,
"step": 860
},
{
"epoch": 1.5320284697508897,
"grad_norm": 0.3978196707900395,
"learning_rate": 1.2911411741159273e-06,
"loss": 0.0159,
"step": 861
},
{
"epoch": 1.5338078291814945,
"grad_norm": 0.7307459220000226,
"learning_rate": 1.2817832681648712e-06,
"loss": 0.0292,
"step": 862
},
{
"epoch": 1.5355871886120998,
"grad_norm": 0.49714095399011266,
"learning_rate": 1.2724544092309581e-06,
"loss": 0.022,
"step": 863
},
{
"epoch": 1.5373665480427046,
"grad_norm": 0.5005354442341268,
"learning_rate": 1.2631546701920073e-06,
"loss": 0.0197,
"step": 864
},
{
"epoch": 1.5391459074733096,
"grad_norm": 0.485320184039084,
"learning_rate": 1.2538841236983519e-06,
"loss": 0.0212,
"step": 865
},
{
"epoch": 1.5409252669039146,
"grad_norm": 0.46472179443198164,
"learning_rate": 1.244642842172266e-06,
"loss": 0.0195,
"step": 866
},
{
"epoch": 1.5427046263345194,
"grad_norm": 0.42574641462722274,
"learning_rate": 1.2354308978074088e-06,
"loss": 0.0152,
"step": 867
},
{
"epoch": 1.5444839857651247,
"grad_norm": 0.5249876680222282,
"learning_rate": 1.2262483625682514e-06,
"loss": 0.0223,
"step": 868
},
{
"epoch": 1.5462633451957295,
"grad_norm": 0.5861380638267182,
"learning_rate": 1.2170953081895214e-06,
"loss": 0.0259,
"step": 869
},
{
"epoch": 1.5480427046263345,
"grad_norm": 0.4183001251568824,
"learning_rate": 1.2079718061756369e-06,
"loss": 0.0147,
"step": 870
},
{
"epoch": 1.5498220640569396,
"grad_norm": 0.5588402237903976,
"learning_rate": 1.1988779278001517e-06,
"loss": 0.0242,
"step": 871
},
{
"epoch": 1.5516014234875444,
"grad_norm": 0.3156076203053354,
"learning_rate": 1.1898137441051982e-06,
"loss": 0.0138,
"step": 872
},
{
"epoch": 1.5533807829181496,
"grad_norm": 0.5871485450716929,
"learning_rate": 1.1807793259009282e-06,
"loss": 0.023,
"step": 873
},
{
"epoch": 1.5551601423487544,
"grad_norm": 0.49871820203905354,
"learning_rate": 1.1717747437649657e-06,
"loss": 0.0252,
"step": 874
},
{
"epoch": 1.5569395017793595,
"grad_norm": 0.5465484638964518,
"learning_rate": 1.1628000680418533e-06,
"loss": 0.0206,
"step": 875
},
{
"epoch": 1.5587188612099645,
"grad_norm": 0.6314577052337949,
"learning_rate": 1.1538553688425002e-06,
"loss": 0.0219,
"step": 876
},
{
"epoch": 1.5604982206405693,
"grad_norm": 0.5184594786443107,
"learning_rate": 1.14494071604364e-06,
"loss": 0.0203,
"step": 877
},
{
"epoch": 1.5622775800711743,
"grad_norm": 0.45595838125405164,
"learning_rate": 1.1360561792872754e-06,
"loss": 0.0146,
"step": 878
},
{
"epoch": 1.5640569395017794,
"grad_norm": 0.7914339814897812,
"learning_rate": 1.127201827980145e-06,
"loss": 0.0293,
"step": 879
},
{
"epoch": 1.5658362989323842,
"grad_norm": 0.48653020395034047,
"learning_rate": 1.1183777312931748e-06,
"loss": 0.0147,
"step": 880
},
{
"epoch": 1.5676156583629894,
"grad_norm": 0.3685952505863919,
"learning_rate": 1.1095839581609407e-06,
"loss": 0.0147,
"step": 881
},
{
"epoch": 1.5693950177935942,
"grad_norm": 0.5101453884911401,
"learning_rate": 1.1008205772811248e-06,
"loss": 0.022,
"step": 882
},
{
"epoch": 1.5711743772241993,
"grad_norm": 0.5224907530054623,
"learning_rate": 1.0920876571139843e-06,
"loss": 0.0217,
"step": 883
},
{
"epoch": 1.5729537366548043,
"grad_norm": 0.5485376717927303,
"learning_rate": 1.0833852658818167e-06,
"loss": 0.0225,
"step": 884
},
{
"epoch": 1.5747330960854091,
"grad_norm": 0.4065459157495074,
"learning_rate": 1.0747134715684221e-06,
"loss": 0.0131,
"step": 885
},
{
"epoch": 1.5765124555160144,
"grad_norm": 0.4696807146925384,
"learning_rate": 1.0660723419185776e-06,
"loss": 0.0233,
"step": 886
},
{
"epoch": 1.5782918149466192,
"grad_norm": 0.37301033588596155,
"learning_rate": 1.0574619444375017e-06,
"loss": 0.0149,
"step": 887
},
{
"epoch": 1.5800711743772242,
"grad_norm": 0.48855760716085866,
"learning_rate": 1.0488823463903341e-06,
"loss": 0.0213,
"step": 888
},
{
"epoch": 1.5818505338078293,
"grad_norm": 0.558531919853563,
"learning_rate": 1.0403336148016053e-06,
"loss": 0.025,
"step": 889
},
{
"epoch": 1.583629893238434,
"grad_norm": 0.390199148226862,
"learning_rate": 1.0318158164547159e-06,
"loss": 0.0193,
"step": 890
},
{
"epoch": 1.585409252669039,
"grad_norm": 0.6138310626366668,
"learning_rate": 1.0233290178914096e-06,
"loss": 0.0199,
"step": 891
},
{
"epoch": 1.5871886120996441,
"grad_norm": 0.4900670192428447,
"learning_rate": 1.014873285411262e-06,
"loss": 0.0202,
"step": 892
},
{
"epoch": 1.5889679715302492,
"grad_norm": 0.4512432406893941,
"learning_rate": 1.006448685071154e-06,
"loss": 0.0184,
"step": 893
},
{
"epoch": 1.5907473309608542,
"grad_norm": 0.4907173411829678,
"learning_rate": 9.980552826847635e-07,
"loss": 0.0169,
"step": 894
},
{
"epoch": 1.592526690391459,
"grad_norm": 0.5497716389584371,
"learning_rate": 9.896931438220453e-07,
"loss": 0.0265,
"step": 895
},
{
"epoch": 1.594306049822064,
"grad_norm": 0.9343966233510561,
"learning_rate": 9.813623338087181e-07,
"loss": 0.0364,
"step": 896
},
{
"epoch": 1.596085409252669,
"grad_norm": 0.5630128133257485,
"learning_rate": 9.730629177257623e-07,
"loss": 0.02,
"step": 897
},
{
"epoch": 1.5978647686832739,
"grad_norm": 0.541524493706779,
"learning_rate": 9.64794960408903e-07,
"loss": 0.0199,
"step": 898
},
{
"epoch": 1.5996441281138791,
"grad_norm": 0.5387904785306966,
"learning_rate": 9.565585264481092e-07,
"loss": 0.0237,
"step": 899
},
{
"epoch": 1.601423487544484,
"grad_norm": 0.4012849694259753,
"learning_rate": 9.483536801870835e-07,
"loss": 0.0151,
"step": 900
},
{
"epoch": 1.603202846975089,
"grad_norm": 0.33208929124046044,
"learning_rate": 9.401804857227648e-07,
"loss": 0.0121,
"step": 901
},
{
"epoch": 1.604982206405694,
"grad_norm": 0.5416522168231634,
"learning_rate": 9.320390069048258e-07,
"loss": 0.0195,
"step": 902
},
{
"epoch": 1.6067615658362988,
"grad_norm": 0.46941905183937926,
"learning_rate": 9.239293073351735e-07,
"loss": 0.0174,
"step": 903
},
{
"epoch": 1.608540925266904,
"grad_norm": 0.5498715105415954,
"learning_rate": 9.158514503674543e-07,
"loss": 0.0199,
"step": 904
},
{
"epoch": 1.6103202846975089,
"grad_norm": 0.41245377902918556,
"learning_rate": 9.078054991065532e-07,
"loss": 0.0147,
"step": 905
},
{
"epoch": 1.612099644128114,
"grad_norm": 0.5106935222796881,
"learning_rate": 8.997915164081095e-07,
"loss": 0.0243,
"step": 906
},
{
"epoch": 1.613879003558719,
"grad_norm": 0.507744894118689,
"learning_rate": 8.918095648780195e-07,
"loss": 0.0198,
"step": 907
},
{
"epoch": 1.6156583629893237,
"grad_norm": 0.5922625864397197,
"learning_rate": 8.838597068719518e-07,
"loss": 0.0187,
"step": 908
},
{
"epoch": 1.6174377224199288,
"grad_norm": 0.7425252683616577,
"learning_rate": 8.75942004494853e-07,
"loss": 0.0314,
"step": 909
},
{
"epoch": 1.6192170818505338,
"grad_norm": 0.4627758082723585,
"learning_rate": 8.680565196004704e-07,
"loss": 0.0192,
"step": 910
},
{
"epoch": 1.6209964412811388,
"grad_norm": 0.45904927743320867,
"learning_rate": 8.602033137908666e-07,
"loss": 0.0147,
"step": 911
},
{
"epoch": 1.6227758007117439,
"grad_norm": 0.3976355822672067,
"learning_rate": 8.523824484159348e-07,
"loss": 0.0133,
"step": 912
},
{
"epoch": 1.6245551601423487,
"grad_norm": 0.4364646790819707,
"learning_rate": 8.445939845729245e-07,
"loss": 0.0167,
"step": 913
},
{
"epoch": 1.6263345195729537,
"grad_norm": 0.3851567059738438,
"learning_rate": 8.368379831059592e-07,
"loss": 0.0155,
"step": 914
},
{
"epoch": 1.6281138790035588,
"grad_norm": 0.46173992428063265,
"learning_rate": 8.29114504605566e-07,
"loss": 0.0213,
"step": 915
},
{
"epoch": 1.6298932384341636,
"grad_norm": 0.6038390325634172,
"learning_rate": 8.21423609408199e-07,
"loss": 0.0262,
"step": 916
},
{
"epoch": 1.6316725978647688,
"grad_norm": 0.42568971923737864,
"learning_rate": 8.137653575957666e-07,
"loss": 0.0165,
"step": 917
},
{
"epoch": 1.6334519572953736,
"grad_norm": 0.4732919276079248,
"learning_rate": 8.061398089951678e-07,
"loss": 0.0208,
"step": 918
},
{
"epoch": 1.6352313167259787,
"grad_norm": 0.4151869904030583,
"learning_rate": 7.985470231778203e-07,
"loss": 0.0175,
"step": 919
},
{
"epoch": 1.6370106761565837,
"grad_norm": 0.5604618536374466,
"learning_rate": 7.909870594591951e-07,
"loss": 0.0245,
"step": 920
},
{
"epoch": 1.6387900355871885,
"grad_norm": 0.6306937736476492,
"learning_rate": 7.834599768983553e-07,
"loss": 0.0259,
"step": 921
},
{
"epoch": 1.6405693950177938,
"grad_norm": 0.37874855194140244,
"learning_rate": 7.759658342974951e-07,
"loss": 0.014,
"step": 922
},
{
"epoch": 1.6423487544483986,
"grad_norm": 0.42384195290753635,
"learning_rate": 7.685046902014747e-07,
"loss": 0.0156,
"step": 923
},
{
"epoch": 1.6441281138790036,
"grad_norm": 0.8120248657686265,
"learning_rate": 7.61076602897371e-07,
"loss": 0.031,
"step": 924
},
{
"epoch": 1.6459074733096086,
"grad_norm": 0.4741418504607924,
"learning_rate": 7.536816304140177e-07,
"loss": 0.0209,
"step": 925
},
{
"epoch": 1.6476868327402134,
"grad_norm": 0.7349143836484207,
"learning_rate": 7.46319830521553e-07,
"loss": 0.0303,
"step": 926
},
{
"epoch": 1.6494661921708185,
"grad_norm": 0.5013871311662859,
"learning_rate": 7.389912607309662e-07,
"loss": 0.021,
"step": 927
},
{
"epoch": 1.6512455516014235,
"grad_norm": 0.5182791758800255,
"learning_rate": 7.316959782936516e-07,
"loss": 0.0192,
"step": 928
},
{
"epoch": 1.6530249110320283,
"grad_norm": 0.48466771361426775,
"learning_rate": 7.244340402009608e-07,
"loss": 0.0234,
"step": 929
},
{
"epoch": 1.6548042704626336,
"grad_norm": 0.657460811484988,
"learning_rate": 7.172055031837572e-07,
"loss": 0.0212,
"step": 930
},
{
"epoch": 1.6565836298932384,
"grad_norm": 0.5366839181326499,
"learning_rate": 7.100104237119676e-07,
"loss": 0.0192,
"step": 931
},
{
"epoch": 1.6583629893238434,
"grad_norm": 0.3956315916608553,
"learning_rate": 7.028488579941506e-07,
"loss": 0.0157,
"step": 932
},
{
"epoch": 1.6601423487544484,
"grad_norm": 0.4921224533236747,
"learning_rate": 6.957208619770505e-07,
"loss": 0.0197,
"step": 933
},
{
"epoch": 1.6619217081850532,
"grad_norm": 0.5523371143209881,
"learning_rate": 6.886264913451635e-07,
"loss": 0.026,
"step": 934
},
{
"epoch": 1.6637010676156585,
"grad_norm": 0.6022985960233896,
"learning_rate": 6.815658015203014e-07,
"loss": 0.0279,
"step": 935
},
{
"epoch": 1.6654804270462633,
"grad_norm": 0.5647602810734895,
"learning_rate": 6.745388476611553e-07,
"loss": 0.0239,
"step": 936
},
{
"epoch": 1.6672597864768683,
"grad_norm": 0.4990725235269463,
"learning_rate": 6.67545684662873e-07,
"loss": 0.02,
"step": 937
},
{
"epoch": 1.6690391459074734,
"grad_norm": 0.4668541870848106,
"learning_rate": 6.605863671566221e-07,
"loss": 0.0181,
"step": 938
},
{
"epoch": 1.6708185053380782,
"grad_norm": 0.4348644673867675,
"learning_rate": 6.536609495091695e-07,
"loss": 0.0177,
"step": 939
},
{
"epoch": 1.6725978647686834,
"grad_norm": 0.49572122445529704,
"learning_rate": 6.467694858224488e-07,
"loss": 0.0237,
"step": 940
},
{
"epoch": 1.6743772241992882,
"grad_norm": 0.4479776222700876,
"learning_rate": 6.399120299331468e-07,
"loss": 0.017,
"step": 941
},
{
"epoch": 1.6761565836298933,
"grad_norm": 0.4657411696983214,
"learning_rate": 6.330886354122768e-07,
"loss": 0.0171,
"step": 942
},
{
"epoch": 1.6779359430604983,
"grad_norm": 0.48501316325759597,
"learning_rate": 6.262993555647617e-07,
"loss": 0.0198,
"step": 943
},
{
"epoch": 1.6797153024911031,
"grad_norm": 0.4141372186820852,
"learning_rate": 6.1954424342902e-07,
"loss": 0.0177,
"step": 944
},
{
"epoch": 1.6814946619217082,
"grad_norm": 0.5768520981298544,
"learning_rate": 6.128233517765448e-07,
"loss": 0.0234,
"step": 945
},
{
"epoch": 1.6832740213523132,
"grad_norm": 0.5709441522653244,
"learning_rate": 6.061367331114992e-07,
"loss": 0.0223,
"step": 946
},
{
"epoch": 1.685053380782918,
"grad_norm": 0.5448556785300659,
"learning_rate": 5.994844396703025e-07,
"loss": 0.0237,
"step": 947
},
{
"epoch": 1.6868327402135233,
"grad_norm": 0.4548794218574884,
"learning_rate": 5.928665234212233e-07,
"loss": 0.0168,
"step": 948
},
{
"epoch": 1.688612099644128,
"grad_norm": 0.45458648783753686,
"learning_rate": 5.862830360639698e-07,
"loss": 0.0178,
"step": 949
},
{
"epoch": 1.690391459074733,
"grad_norm": 0.5236485887055707,
"learning_rate": 5.797340290292907e-07,
"loss": 0.0204,
"step": 950
},
{
"epoch": 1.6921708185053381,
"grad_norm": 0.5747000403102759,
"learning_rate": 5.732195534785723e-07,
"loss": 0.0257,
"step": 951
},
{
"epoch": 1.693950177935943,
"grad_norm": 0.5421953997915431,
"learning_rate": 5.667396603034369e-07,
"loss": 0.0185,
"step": 952
},
{
"epoch": 1.6957295373665482,
"grad_norm": 0.5383715950714395,
"learning_rate": 5.602944001253486e-07,
"loss": 0.0171,
"step": 953
},
{
"epoch": 1.697508896797153,
"grad_norm": 0.45150403360884067,
"learning_rate": 5.538838232952104e-07,
"loss": 0.0184,
"step": 954
},
{
"epoch": 1.699288256227758,
"grad_norm": 0.5343979256670289,
"learning_rate": 5.475079798929816e-07,
"loss": 0.0218,
"step": 955
},
{
"epoch": 1.701067615658363,
"grad_norm": 0.3936870172269373,
"learning_rate": 5.411669197272795e-07,
"loss": 0.009,
"step": 956
},
{
"epoch": 1.7028469750889679,
"grad_norm": 0.4238680319994359,
"learning_rate": 5.348606923349903e-07,
"loss": 0.016,
"step": 957
},
{
"epoch": 1.704626334519573,
"grad_norm": 0.4221198628189162,
"learning_rate": 5.285893469808855e-07,
"loss": 0.0151,
"step": 958
},
{
"epoch": 1.706405693950178,
"grad_norm": 0.7548563743177794,
"learning_rate": 5.223529326572352e-07,
"loss": 0.028,
"step": 959
},
{
"epoch": 1.708185053380783,
"grad_norm": 0.5326084459614833,
"learning_rate": 5.161514980834232e-07,
"loss": 0.0197,
"step": 960
},
{
"epoch": 1.709964412811388,
"grad_norm": 0.5046668457379697,
"learning_rate": 5.099850917055709e-07,
"loss": 0.0216,
"step": 961
},
{
"epoch": 1.7117437722419928,
"grad_norm": 0.5687090288338201,
"learning_rate": 5.038537616961559e-07,
"loss": 0.021,
"step": 962
},
{
"epoch": 1.7135231316725978,
"grad_norm": 0.5128052073581784,
"learning_rate": 4.977575559536358e-07,
"loss": 0.0211,
"step": 963
},
{
"epoch": 1.7153024911032029,
"grad_norm": 0.531504027911695,
"learning_rate": 4.916965221020753e-07,
"loss": 0.0268,
"step": 964
},
{
"epoch": 1.7170818505338077,
"grad_norm": 0.6373540569342278,
"learning_rate": 4.856707074907729e-07,
"loss": 0.0255,
"step": 965
},
{
"epoch": 1.718861209964413,
"grad_norm": 0.552537633367182,
"learning_rate": 4.796801591938922e-07,
"loss": 0.0215,
"step": 966
},
{
"epoch": 1.7206405693950177,
"grad_norm": 0.4913651329664384,
"learning_rate": 4.737249240100911e-07,
"loss": 0.016,
"step": 967
},
{
"epoch": 1.7224199288256228,
"grad_norm": 0.45037635689990874,
"learning_rate": 4.6780504846216155e-07,
"loss": 0.015,
"step": 968
},
{
"epoch": 1.7241992882562278,
"grad_norm": 0.5140704297108523,
"learning_rate": 4.619205787966613e-07,
"loss": 0.0177,
"step": 969
},
{
"epoch": 1.7259786476868326,
"grad_norm": 0.5324119451911834,
"learning_rate": 4.560715609835548e-07,
"loss": 0.0254,
"step": 970
},
{
"epoch": 1.7277580071174379,
"grad_norm": 0.4483983993731682,
"learning_rate": 4.5025804071585464e-07,
"loss": 0.0162,
"step": 971
},
{
"epoch": 1.7295373665480427,
"grad_norm": 0.8024751449382543,
"learning_rate": 4.4448006340926163e-07,
"loss": 0.029,
"step": 972
},
{
"epoch": 1.7313167259786477,
"grad_norm": 0.5049557657908982,
"learning_rate": 4.3873767420181344e-07,
"loss": 0.0208,
"step": 973
},
{
"epoch": 1.7330960854092528,
"grad_norm": 0.5407283247556349,
"learning_rate": 4.3303091795353024e-07,
"loss": 0.0217,
"step": 974
},
{
"epoch": 1.7348754448398576,
"grad_norm": 0.4393663540869399,
"learning_rate": 4.2735983924606596e-07,
"loss": 0.0188,
"step": 975
},
{
"epoch": 1.7366548042704626,
"grad_norm": 0.5952378336729619,
"learning_rate": 4.2172448238235464e-07,
"loss": 0.0207,
"step": 976
},
{
"epoch": 1.7384341637010676,
"grad_norm": 0.6106130763678146,
"learning_rate": 4.161248913862731e-07,
"loss": 0.0216,
"step": 977
},
{
"epoch": 1.7402135231316724,
"grad_norm": 0.4713722466317228,
"learning_rate": 4.1056111000228937e-07,
"loss": 0.0193,
"step": 978
},
{
"epoch": 1.7419928825622777,
"grad_norm": 0.5650874130283632,
"learning_rate": 4.0503318169512417e-07,
"loss": 0.026,
"step": 979
},
{
"epoch": 1.7437722419928825,
"grad_norm": 0.4575770723935713,
"learning_rate": 3.9954114964941336e-07,
"loss": 0.0196,
"step": 980
},
{
"epoch": 1.7455516014234875,
"grad_norm": 0.5281738806288874,
"learning_rate": 3.9408505676936327e-07,
"loss": 0.0226,
"step": 981
},
{
"epoch": 1.7473309608540926,
"grad_norm": 0.5793123527844682,
"learning_rate": 3.886649456784253e-07,
"loss": 0.0261,
"step": 982
},
{
"epoch": 1.7491103202846974,
"grad_norm": 0.5345885281075365,
"learning_rate": 3.8328085871895624e-07,
"loss": 0.0239,
"step": 983
},
{
"epoch": 1.7508896797153026,
"grad_norm": 0.45584403836263787,
"learning_rate": 3.779328379518898e-07,
"loss": 0.0184,
"step": 984
},
{
"epoch": 1.7526690391459074,
"grad_norm": 0.41807423622983675,
"learning_rate": 3.7262092515640556e-07,
"loss": 0.015,
"step": 985
},
{
"epoch": 1.7544483985765125,
"grad_norm": 0.3892954041570445,
"learning_rate": 3.673451618296081e-07,
"loss": 0.0151,
"step": 986
},
{
"epoch": 1.7562277580071175,
"grad_norm": 0.4949427787782939,
"learning_rate": 3.621055891861963e-07,
"loss": 0.0227,
"step": 987
},
{
"epoch": 1.7580071174377223,
"grad_norm": 0.5977847561281114,
"learning_rate": 3.56902248158148e-07,
"loss": 0.0224,
"step": 988
},
{
"epoch": 1.7597864768683276,
"grad_norm": 0.5319976887777412,
"learning_rate": 3.517351793943913e-07,
"loss": 0.0208,
"step": 989
},
{
"epoch": 1.7615658362989324,
"grad_norm": 0.6985739136609802,
"learning_rate": 3.4660442326049704e-07,
"loss": 0.0272,
"step": 990
},
{
"epoch": 1.7633451957295374,
"grad_norm": 0.4430883655183018,
"learning_rate": 3.4151001983835696e-07,
"loss": 0.0185,
"step": 991
},
{
"epoch": 1.7651245551601424,
"grad_norm": 0.5277584223480933,
"learning_rate": 3.364520089258727e-07,
"loss": 0.0173,
"step": 992
},
{
"epoch": 1.7669039145907472,
"grad_norm": 0.44357989361782046,
"learning_rate": 3.314304300366461e-07,
"loss": 0.0179,
"step": 993
},
{
"epoch": 1.7686832740213523,
"grad_norm": 0.42940394684191535,
"learning_rate": 3.2644532239966444e-07,
"loss": 0.0202,
"step": 994
},
{
"epoch": 1.7704626334519573,
"grad_norm": 0.30114143858467224,
"learning_rate": 3.2149672495900286e-07,
"loss": 0.0145,
"step": 995
},
{
"epoch": 1.7722419928825621,
"grad_norm": 0.4637181866776221,
"learning_rate": 3.165846763735153e-07,
"loss": 0.0222,
"step": 996
},
{
"epoch": 1.7740213523131674,
"grad_norm": 0.5488287432874124,
"learning_rate": 3.117092150165324e-07,
"loss": 0.0236,
"step": 997
},
{
"epoch": 1.7758007117437722,
"grad_norm": 0.66958815864923,
"learning_rate": 3.068703789755606e-07,
"loss": 0.0251,
"step": 998
},
{
"epoch": 1.7775800711743772,
"grad_norm": 0.5099537314387769,
"learning_rate": 3.020682060519886e-07,
"loss": 0.0185,
"step": 999
},
{
"epoch": 1.7793594306049823,
"grad_norm": 0.6309455605154024,
"learning_rate": 2.9730273376078923e-07,
"loss": 0.026,
"step": 1000
},
{
"epoch": 1.7793594306049823,
"eval_loss": 0.08554951101541519,
"eval_runtime": 1.8862,
"eval_samples_per_second": 24.387,
"eval_steps_per_second": 6.362,
"step": 1000
},
{
"epoch": 1.781138790035587,
"grad_norm": 0.4918888146262323,
"learning_rate": 2.9257399933022737e-07,
"loss": 0.0179,
"step": 1001
},
{
"epoch": 1.7829181494661923,
"grad_norm": 0.664928386728144,
"learning_rate": 2.8788203970156805e-07,
"loss": 0.0304,
"step": 1002
},
{
"epoch": 1.7846975088967971,
"grad_norm": 0.4655111152736626,
"learning_rate": 2.832268915287878e-07,
"loss": 0.021,
"step": 1003
},
{
"epoch": 1.7864768683274022,
"grad_norm": 0.6013046497693293,
"learning_rate": 2.7860859117828985e-07,
"loss": 0.0286,
"step": 1004
},
{
"epoch": 1.7882562277580072,
"grad_norm": 0.5733580866681125,
"learning_rate": 2.740271747286194e-07,
"loss": 0.0225,
"step": 1005
},
{
"epoch": 1.790035587188612,
"grad_norm": 0.5019568144104094,
"learning_rate": 2.6948267797018145e-07,
"loss": 0.0176,
"step": 1006
},
{
"epoch": 1.791814946619217,
"grad_norm": 0.46353760082586437,
"learning_rate": 2.649751364049613e-07,
"loss": 0.0117,
"step": 1007
},
{
"epoch": 1.793594306049822,
"grad_norm": 0.4168201293440238,
"learning_rate": 2.6050458524624735e-07,
"loss": 0.0139,
"step": 1008
},
{
"epoch": 1.795373665480427,
"grad_norm": 0.5232632789481222,
"learning_rate": 2.560710594183552e-07,
"loss": 0.0188,
"step": 1009
},
{
"epoch": 1.7971530249110321,
"grad_norm": 0.4628487756692675,
"learning_rate": 2.5167459355635524e-07,
"loss": 0.0168,
"step": 1010
},
{
"epoch": 1.798932384341637,
"grad_norm": 0.5587033872805784,
"learning_rate": 2.473152220058039e-07,
"loss": 0.0207,
"step": 1011
},
{
"epoch": 1.800711743772242,
"grad_norm": 0.5172004207975792,
"learning_rate": 2.429929788224722e-07,
"loss": 0.0207,
"step": 1012
},
{
"epoch": 1.802491103202847,
"grad_norm": 0.4782451602443775,
"learning_rate": 2.38707897772083e-07,
"loss": 0.0211,
"step": 1013
},
{
"epoch": 1.8042704626334518,
"grad_norm": 0.6034158584556891,
"learning_rate": 2.3446001233004333e-07,
"loss": 0.0231,
"step": 1014
},
{
"epoch": 1.806049822064057,
"grad_norm": 0.43350310557356275,
"learning_rate": 2.3024935568118745e-07,
"loss": 0.018,
"step": 1015
},
{
"epoch": 1.8078291814946619,
"grad_norm": 0.4994397027838582,
"learning_rate": 2.2607596071951288e-07,
"loss": 0.0177,
"step": 1016
},
{
"epoch": 1.809608540925267,
"grad_norm": 0.5024489708289125,
"learning_rate": 2.2193986004792667e-07,
"loss": 0.0222,
"step": 1017
},
{
"epoch": 1.811387900355872,
"grad_norm": 0.37368688650276854,
"learning_rate": 2.1784108597799058e-07,
"loss": 0.0126,
"step": 1018
},
{
"epoch": 1.8131672597864767,
"grad_norm": 0.3779868984963832,
"learning_rate": 2.1377967052966685e-07,
"loss": 0.0112,
"step": 1019
},
{
"epoch": 1.814946619217082,
"grad_norm": 0.38724525805942295,
"learning_rate": 2.0975564543107007e-07,
"loss": 0.0166,
"step": 1020
},
{
"epoch": 1.8167259786476868,
"grad_norm": 0.4306096057870547,
"learning_rate": 2.057690421182168e-07,
"loss": 0.0125,
"step": 1021
},
{
"epoch": 1.8185053380782918,
"grad_norm": 0.41782507127108337,
"learning_rate": 2.01819891734783e-07,
"loss": 0.0155,
"step": 1022
},
{
"epoch": 1.8202846975088969,
"grad_norm": 0.5646811465921464,
"learning_rate": 1.979082251318576e-07,
"loss": 0.0212,
"step": 1023
},
{
"epoch": 1.8220640569395017,
"grad_norm": 0.664036356858342,
"learning_rate": 1.9403407286770592e-07,
"loss": 0.0218,
"step": 1024
},
{
"epoch": 1.8238434163701067,
"grad_norm": 0.6013287465855081,
"learning_rate": 1.9019746520752502e-07,
"loss": 0.0237,
"step": 1025
},
{
"epoch": 1.8256227758007118,
"grad_norm": 0.4465350692321817,
"learning_rate": 1.8639843212321206e-07,
"loss": 0.0171,
"step": 1026
},
{
"epoch": 1.8274021352313166,
"grad_norm": 0.45772384028065144,
"learning_rate": 1.826370032931285e-07,
"loss": 0.0205,
"step": 1027
},
{
"epoch": 1.8291814946619218,
"grad_norm": 0.6246908380468256,
"learning_rate": 1.789132081018674e-07,
"loss": 0.0171,
"step": 1028
},
{
"epoch": 1.8309608540925266,
"grad_norm": 0.4641513486679748,
"learning_rate": 1.7522707564002706e-07,
"loss": 0.016,
"step": 1029
},
{
"epoch": 1.8327402135231317,
"grad_norm": 0.6290092576270153,
"learning_rate": 1.7157863470397718e-07,
"loss": 0.0243,
"step": 1030
},
{
"epoch": 1.8345195729537367,
"grad_norm": 0.44966198038557703,
"learning_rate": 1.6796791379564138e-07,
"loss": 0.017,
"step": 1031
},
{
"epoch": 1.8362989323843415,
"grad_norm": 0.3317322408241649,
"learning_rate": 1.6439494112227173e-07,
"loss": 0.0105,
"step": 1032
},
{
"epoch": 1.8380782918149468,
"grad_norm": 0.8050119700338149,
"learning_rate": 1.6085974459622567e-07,
"loss": 0.0284,
"step": 1033
},
{
"epoch": 1.8398576512455516,
"grad_norm": 0.47130464093732133,
"learning_rate": 1.573623518347517e-07,
"loss": 0.0199,
"step": 1034
},
{
"epoch": 1.8416370106761566,
"grad_norm": 0.46933869367395276,
"learning_rate": 1.5390279015977117e-07,
"loss": 0.0197,
"step": 1035
},
{
"epoch": 1.8434163701067616,
"grad_norm": 0.7134459326855846,
"learning_rate": 1.5048108659766693e-07,
"loss": 0.0223,
"step": 1036
},
{
"epoch": 1.8451957295373664,
"grad_norm": 0.37486560589355483,
"learning_rate": 1.470972678790711e-07,
"loss": 0.0191,
"step": 1037
},
{
"epoch": 1.8469750889679717,
"grad_norm": 0.4645108224608385,
"learning_rate": 1.437513604386559e-07,
"loss": 0.0185,
"step": 1038
},
{
"epoch": 1.8487544483985765,
"grad_norm": 0.5477065721799405,
"learning_rate": 1.404433904149266e-07,
"loss": 0.0183,
"step": 1039
},
{
"epoch": 1.8505338078291815,
"grad_norm": 0.47880509487119105,
"learning_rate": 1.3717338365001943e-07,
"loss": 0.0178,
"step": 1040
},
{
"epoch": 1.8523131672597866,
"grad_norm": 0.6747568623480683,
"learning_rate": 1.3394136568949834e-07,
"loss": 0.0384,
"step": 1041
},
{
"epoch": 1.8540925266903914,
"grad_norm": 0.4928805754441432,
"learning_rate": 1.307473617821553e-07,
"loss": 0.0154,
"step": 1042
},
{
"epoch": 1.8558718861209964,
"grad_norm": 0.5189820557912201,
"learning_rate": 1.275913968798137e-07,
"loss": 0.0218,
"step": 1043
},
{
"epoch": 1.8576512455516014,
"grad_norm": 0.42672301551604463,
"learning_rate": 1.2447349563713186e-07,
"loss": 0.0169,
"step": 1044
},
{
"epoch": 1.8594306049822062,
"grad_norm": 0.46175334104934335,
"learning_rate": 1.213936824114137e-07,
"loss": 0.0166,
"step": 1045
},
{
"epoch": 1.8612099644128115,
"grad_norm": 0.672827910918819,
"learning_rate": 1.1835198126241509e-07,
"loss": 0.0283,
"step": 1046
},
{
"epoch": 1.8629893238434163,
"grad_norm": 0.5211979526912517,
"learning_rate": 1.1534841595215617e-07,
"loss": 0.0264,
"step": 1047
},
{
"epoch": 1.8647686832740213,
"grad_norm": 0.48371581606375236,
"learning_rate": 1.1238300994473983e-07,
"loss": 0.0208,
"step": 1048
},
{
"epoch": 1.8665480427046264,
"grad_norm": 0.5239180269913135,
"learning_rate": 1.0945578640616183e-07,
"loss": 0.0293,
"step": 1049
},
{
"epoch": 1.8683274021352312,
"grad_norm": 0.4585786409071437,
"learning_rate": 1.0656676820413603e-07,
"loss": 0.0146,
"step": 1050
},
{
"epoch": 1.8701067615658364,
"grad_norm": 0.4591011903758282,
"learning_rate": 1.0371597790791166e-07,
"loss": 0.022,
"step": 1051
},
{
"epoch": 1.8718861209964412,
"grad_norm": 0.5671780677616047,
"learning_rate": 1.0090343778809908e-07,
"loss": 0.0221,
"step": 1052
},
{
"epoch": 1.8736654804270463,
"grad_norm": 0.42875188473901343,
"learning_rate": 9.812916981649433e-08,
"loss": 0.0132,
"step": 1053
},
{
"epoch": 1.8754448398576513,
"grad_norm": 0.5096209067259122,
"learning_rate": 9.539319566590766e-08,
"loss": 0.0211,
"step": 1054
},
{
"epoch": 1.8772241992882561,
"grad_norm": 0.5698569970340378,
"learning_rate": 9.269553670999743e-08,
"loss": 0.0263,
"step": 1055
},
{
"epoch": 1.8790035587188612,
"grad_norm": 0.6033057286889466,
"learning_rate": 9.003621402309815e-08,
"loss": 0.0234,
"step": 1056
},
{
"epoch": 1.8807829181494662,
"grad_norm": 0.6296956645114715,
"learning_rate": 8.741524838005888e-08,
"loss": 0.022,
"step": 1057
},
{
"epoch": 1.8825622775800712,
"grad_norm": 0.5033125381619202,
"learning_rate": 8.483266025608061e-08,
"loss": 0.017,
"step": 1058
},
{
"epoch": 1.8843416370106763,
"grad_norm": 0.3762170537945546,
"learning_rate": 8.228846982655525e-08,
"loss": 0.014,
"step": 1059
},
{
"epoch": 1.886120996441281,
"grad_norm": 0.4355574070967754,
"learning_rate": 7.978269696691021e-08,
"loss": 0.017,
"step": 1060
},
{
"epoch": 1.887900355871886,
"grad_norm": 0.4949657971860853,
"learning_rate": 7.731536125244965e-08,
"loss": 0.0207,
"step": 1061
},
{
"epoch": 1.8896797153024911,
"grad_norm": 0.5441555031839033,
"learning_rate": 7.488648195820513e-08,
"loss": 0.0275,
"step": 1062
},
{
"epoch": 1.891459074733096,
"grad_norm": 0.5581946761357965,
"learning_rate": 7.249607805878245e-08,
"loss": 0.0204,
"step": 1063
},
{
"epoch": 1.8932384341637012,
"grad_norm": 0.5172288701480963,
"learning_rate": 7.014416822821557e-08,
"loss": 0.0188,
"step": 1064
},
{
"epoch": 1.895017793594306,
"grad_norm": 0.47786819136931424,
"learning_rate": 6.783077083981793e-08,
"loss": 0.018,
"step": 1065
},
{
"epoch": 1.896797153024911,
"grad_norm": 0.5462306130339647,
"learning_rate": 6.55559039660425e-08,
"loss": 0.025,
"step": 1066
},
{
"epoch": 1.898576512455516,
"grad_norm": 0.6184481195317907,
"learning_rate": 6.331958537833693e-08,
"loss": 0.0225,
"step": 1067
},
{
"epoch": 1.9003558718861209,
"grad_norm": 0.5246313355949664,
"learning_rate": 6.112183254700866e-08,
"loss": 0.0207,
"step": 1068
},
{
"epoch": 1.9021352313167261,
"grad_norm": 0.4997329636880393,
"learning_rate": 5.8962662641083856e-08,
"loss": 0.0138,
"step": 1069
},
{
"epoch": 1.903914590747331,
"grad_norm": 0.5042868411030565,
"learning_rate": 5.6842092528176516e-08,
"loss": 0.0202,
"step": 1070
},
{
"epoch": 1.905693950177936,
"grad_norm": 0.563766681120732,
"learning_rate": 5.476013877435626e-08,
"loss": 0.0231,
"step": 1071
},
{
"epoch": 1.907473309608541,
"grad_norm": 0.4757468657886632,
"learning_rate": 5.271681764401848e-08,
"loss": 0.0222,
"step": 1072
},
{
"epoch": 1.9092526690391458,
"grad_norm": 0.39344457114058556,
"learning_rate": 5.071214509975775e-08,
"loss": 0.0179,
"step": 1073
},
{
"epoch": 1.9110320284697508,
"grad_norm": 0.5821410028661035,
"learning_rate": 4.8746136802240716e-08,
"loss": 0.0243,
"step": 1074
},
{
"epoch": 1.9128113879003559,
"grad_norm": 0.7232314404691818,
"learning_rate": 4.6818808110087875e-08,
"loss": 0.0282,
"step": 1075
},
{
"epoch": 1.914590747330961,
"grad_norm": 0.6711138694102632,
"learning_rate": 4.493017407975087e-08,
"loss": 0.0255,
"step": 1076
},
{
"epoch": 1.916370106761566,
"grad_norm": 0.561536381761869,
"learning_rate": 4.308024946539424e-08,
"loss": 0.0222,
"step": 1077
},
{
"epoch": 1.9181494661921707,
"grad_norm": 0.4823470406112393,
"learning_rate": 4.1269048718783344e-08,
"loss": 0.0151,
"step": 1078
},
{
"epoch": 1.9199288256227758,
"grad_norm": 0.40633290350445944,
"learning_rate": 3.9496585989167726e-08,
"loss": 0.0173,
"step": 1079
},
{
"epoch": 1.9217081850533808,
"grad_norm": 0.5429308586558288,
"learning_rate": 3.776287512317345e-08,
"loss": 0.0234,
"step": 1080
},
{
"epoch": 1.9234875444839856,
"grad_norm": 0.45912863848209473,
"learning_rate": 3.606792966469375e-08,
"loss": 0.0171,
"step": 1081
},
{
"epoch": 1.9252669039145909,
"grad_norm": 0.5522933814196604,
"learning_rate": 3.4411762854782426e-08,
"loss": 0.0219,
"step": 1082
},
{
"epoch": 1.9270462633451957,
"grad_norm": 0.46754274562595394,
"learning_rate": 3.279438763155174e-08,
"loss": 0.0149,
"step": 1083
},
{
"epoch": 1.9288256227758007,
"grad_norm": 0.5444073320437323,
"learning_rate": 3.121581663007134e-08,
"loss": 0.0203,
"step": 1084
},
{
"epoch": 1.9306049822064058,
"grad_norm": 0.3435533232851744,
"learning_rate": 2.967606218226837e-08,
"loss": 0.0121,
"step": 1085
},
{
"epoch": 1.9323843416370106,
"grad_norm": 0.386309318238084,
"learning_rate": 2.8175136316832e-08,
"loss": 0.0146,
"step": 1086
},
{
"epoch": 1.9341637010676158,
"grad_norm": 0.549005267767948,
"learning_rate": 2.6713050759120117e-08,
"loss": 0.0261,
"step": 1087
},
{
"epoch": 1.9359430604982206,
"grad_norm": 0.4518538837947768,
"learning_rate": 2.528981693106558e-08,
"loss": 0.0151,
"step": 1088
},
{
"epoch": 1.9377224199288257,
"grad_norm": 0.5390565318932959,
"learning_rate": 2.3905445951089013e-08,
"loss": 0.023,
"step": 1089
},
{
"epoch": 1.9395017793594307,
"grad_norm": 0.48814057477910455,
"learning_rate": 2.2559948634011673e-08,
"loss": 0.015,
"step": 1090
},
{
"epoch": 1.9412811387900355,
"grad_norm": 0.5815594851582494,
"learning_rate": 2.125333549096942e-08,
"loss": 0.0241,
"step": 1091
},
{
"epoch": 1.9430604982206405,
"grad_norm": 0.47576177620450144,
"learning_rate": 1.9985616729332747e-08,
"loss": 0.0225,
"step": 1092
},
{
"epoch": 1.9448398576512456,
"grad_norm": 0.43576495337333404,
"learning_rate": 1.8756802252625773e-08,
"loss": 0.0168,
"step": 1093
},
{
"epoch": 1.9466192170818504,
"grad_norm": 0.34050834892228654,
"learning_rate": 1.75669016604485e-08,
"loss": 0.0105,
"step": 1094
},
{
"epoch": 1.9483985765124556,
"grad_norm": 3.18666887240072,
"learning_rate": 1.6415924248403547e-08,
"loss": 0.0378,
"step": 1095
},
{
"epoch": 1.9501779359430604,
"grad_norm": 0.40611945564968616,
"learning_rate": 1.5303879008021773e-08,
"loss": 0.0164,
"step": 1096
},
{
"epoch": 1.9519572953736655,
"grad_norm": 0.5219000260112633,
"learning_rate": 1.4230774626691756e-08,
"loss": 0.0166,
"step": 1097
},
{
"epoch": 1.9537366548042705,
"grad_norm": 0.4441972827621415,
"learning_rate": 1.3196619487594875e-08,
"loss": 0.0176,
"step": 1098
},
{
"epoch": 1.9555160142348753,
"grad_norm": 0.9922416915845649,
"learning_rate": 1.2201421669636448e-08,
"loss": 0.0348,
"step": 1099
},
{
"epoch": 1.9572953736654806,
"grad_norm": 0.4627641262857506,
"learning_rate": 1.1245188947384133e-08,
"loss": 0.0186,
"step": 1100
},
{
"epoch": 1.9590747330960854,
"grad_norm": 0.43893060465327755,
"learning_rate": 1.0327928791006858e-08,
"loss": 0.0129,
"step": 1101
},
{
"epoch": 1.9608540925266904,
"grad_norm": 0.3899754226584124,
"learning_rate": 9.449648366217645e-09,
"loss": 0.0142,
"step": 1102
},
{
"epoch": 1.9626334519572954,
"grad_norm": 0.4540912678834396,
"learning_rate": 8.61035453421588e-09,
"loss": 0.0163,
"step": 1103
},
{
"epoch": 1.9644128113879002,
"grad_norm": 0.6288860850384945,
"learning_rate": 7.81005385163458e-09,
"loss": 0.024,
"step": 1104
},
{
"epoch": 1.9661921708185055,
"grad_norm": 0.34247484393969974,
"learning_rate": 7.048752570488205e-09,
"loss": 0.0171,
"step": 1105
},
{
"epoch": 1.9679715302491103,
"grad_norm": 0.5984067970125618,
"learning_rate": 6.326456638125478e-09,
"loss": 0.0315,
"step": 1106
},
{
"epoch": 1.9697508896797153,
"grad_norm": 0.49087237864292027,
"learning_rate": 5.643171697183314e-09,
"loss": 0.0203,
"step": 1107
},
{
"epoch": 1.9715302491103204,
"grad_norm": 0.6310058570699796,
"learning_rate": 4.998903085539075e-09,
"loss": 0.0291,
"step": 1108
},
{
"epoch": 1.9733096085409252,
"grad_norm": 0.4032243767726734,
"learning_rate": 4.393655836272825e-09,
"loss": 0.0167,
"step": 1109
},
{
"epoch": 1.9750889679715302,
"grad_norm": 0.6543751428521375,
"learning_rate": 3.8274346776262514e-09,
"loss": 0.0251,
"step": 1110
},
{
"epoch": 1.9768683274021353,
"grad_norm": 0.39974353787119493,
"learning_rate": 3.300244032966582e-09,
"loss": 0.0141,
"step": 1111
},
{
"epoch": 1.97864768683274,
"grad_norm": 0.36978620489834124,
"learning_rate": 2.8120880207493928e-09,
"loss": 0.0135,
"step": 1112
},
{
"epoch": 1.9804270462633453,
"grad_norm": 0.44909601310141894,
"learning_rate": 2.362970454491409e-09,
"loss": 0.0173,
"step": 1113
},
{
"epoch": 1.9822064056939501,
"grad_norm": 0.5250010055401794,
"learning_rate": 1.952894842735531e-09,
"loss": 0.0192,
"step": 1114
},
{
"epoch": 1.9839857651245552,
"grad_norm": 0.3376889042548764,
"learning_rate": 1.5818643890258555e-09,
"loss": 0.0139,
"step": 1115
},
{
"epoch": 1.9857651245551602,
"grad_norm": 0.6314007780398342,
"learning_rate": 1.2498819918843609e-09,
"loss": 0.0253,
"step": 1116
},
{
"epoch": 1.987544483985765,
"grad_norm": 0.611903309182949,
"learning_rate": 9.569502447837053e-10,
"loss": 0.028,
"step": 1117
},
{
"epoch": 1.9893238434163703,
"grad_norm": 0.5311788648774444,
"learning_rate": 7.03071436131686e-10,
"loss": 0.0231,
"step": 1118
},
{
"epoch": 1.991103202846975,
"grad_norm": 0.41500559812152865,
"learning_rate": 4.882475492506977e-10,
"loss": 0.0181,
"step": 1119
},
{
"epoch": 1.99288256227758,
"grad_norm": 0.7566039345567596,
"learning_rate": 3.124802623627465e-10,
"loss": 0.0304,
"step": 1120
},
{
"epoch": 1.9946619217081851,
"grad_norm": 0.6557379960860823,
"learning_rate": 1.7577094857557097e-10,
"loss": 0.0222,
"step": 1121
},
{
"epoch": 1.99644128113879,
"grad_norm": 0.48628525276169826,
"learning_rate": 7.812067587487093e-11,
"loss": 0.0189,
"step": 1122
},
{
"epoch": 1.998220640569395,
"grad_norm": 0.43543362029851085,
"learning_rate": 1.9530207111539967e-11,
"loss": 0.016,
"step": 1123
},
{
"epoch": 2.0,
"grad_norm": 0.3169781021038276,
"learning_rate": 0.0,
"loss": 0.0121,
"step": 1124
},
{
"epoch": 2.0,
"step": 1124,
"total_flos": 5008217800704.0,
"train_loss": 0.04604034536365698,
"train_runtime": 858.435,
"train_samples_per_second": 10.468,
"train_steps_per_second": 1.309
}
],
"logging_steps": 1,
"max_steps": 1124,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 50000000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5008217800704.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}