DRIP / llava /final /original /trainer_state.json
YusenPeng's picture
Upload folder using huggingface_hub
61af751 verified
Raw
History Blame Contribute Delete
212 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9992251189538466,
"eval_steps": 500,
"global_step": 1219,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0008197088752697675,
"grad_norm": 30.179790926168046,
"learning_rate": 5.405405405405406e-07,
"loss": 2.1007,
"step": 1
},
{
"epoch": 0.001639417750539535,
"grad_norm": 30.766150310742862,
"learning_rate": 1.0810810810810812e-06,
"loss": 2.0082,
"step": 2
},
{
"epoch": 0.0024591266258093022,
"grad_norm": 33.06541337025983,
"learning_rate": 1.6216216216216219e-06,
"loss": 2.1982,
"step": 3
},
{
"epoch": 0.00327883550107907,
"grad_norm": 30.935665517982905,
"learning_rate": 2.1621621621621623e-06,
"loss": 2.121,
"step": 4
},
{
"epoch": 0.0040985443763488375,
"grad_norm": 27.776915112701452,
"learning_rate": 2.702702702702703e-06,
"loss": 1.9566,
"step": 5
},
{
"epoch": 0.0049182532516186045,
"grad_norm": 25.84495634789921,
"learning_rate": 3.2432432432432437e-06,
"loss": 1.9059,
"step": 6
},
{
"epoch": 0.005737962126888372,
"grad_norm": 12.432965584383247,
"learning_rate": 3.7837837837837844e-06,
"loss": 1.425,
"step": 7
},
{
"epoch": 0.00655767100215814,
"grad_norm": 10.125040917843112,
"learning_rate": 4.324324324324325e-06,
"loss": 1.3635,
"step": 8
},
{
"epoch": 0.007377379877427907,
"grad_norm": 7.9978889586354915,
"learning_rate": 4.864864864864866e-06,
"loss": 1.3271,
"step": 9
},
{
"epoch": 0.008197088752697675,
"grad_norm": 3.6909796324091557,
"learning_rate": 5.405405405405406e-06,
"loss": 1.1511,
"step": 10
},
{
"epoch": 0.009016797627967442,
"grad_norm": 3.419581102405752,
"learning_rate": 5.945945945945947e-06,
"loss": 1.1845,
"step": 11
},
{
"epoch": 0.009836506503237209,
"grad_norm": 3.6261126759857576,
"learning_rate": 6.486486486486487e-06,
"loss": 1.1466,
"step": 12
},
{
"epoch": 0.010656215378506978,
"grad_norm": 3.6239512042390913,
"learning_rate": 7.027027027027028e-06,
"loss": 1.0555,
"step": 13
},
{
"epoch": 0.011475924253776745,
"grad_norm": 4.023475590475269,
"learning_rate": 7.567567567567569e-06,
"loss": 1.128,
"step": 14
},
{
"epoch": 0.012295633129046512,
"grad_norm": 4.42205564512199,
"learning_rate": 8.108108108108109e-06,
"loss": 1.0579,
"step": 15
},
{
"epoch": 0.01311534200431628,
"grad_norm": 3.887763791725328,
"learning_rate": 8.64864864864865e-06,
"loss": 1.0618,
"step": 16
},
{
"epoch": 0.013935050879586047,
"grad_norm": 2.982466995216937,
"learning_rate": 9.189189189189191e-06,
"loss": 1.0635,
"step": 17
},
{
"epoch": 0.014754759754855814,
"grad_norm": 2.928264200468093,
"learning_rate": 9.729729729729732e-06,
"loss": 1.0211,
"step": 18
},
{
"epoch": 0.015574468630125581,
"grad_norm": 3.0147536603967864,
"learning_rate": 1.027027027027027e-05,
"loss": 1.0717,
"step": 19
},
{
"epoch": 0.01639417750539535,
"grad_norm": 2.8194092179604318,
"learning_rate": 1.0810810810810812e-05,
"loss": 0.9395,
"step": 20
},
{
"epoch": 0.017213886380665115,
"grad_norm": 3.084827945374452,
"learning_rate": 1.1351351351351352e-05,
"loss": 0.9861,
"step": 21
},
{
"epoch": 0.018033595255934884,
"grad_norm": 3.409230973125943,
"learning_rate": 1.1891891891891894e-05,
"loss": 1.0025,
"step": 22
},
{
"epoch": 0.018853304131204653,
"grad_norm": 2.1813460639152162,
"learning_rate": 1.2432432432432433e-05,
"loss": 0.9077,
"step": 23
},
{
"epoch": 0.019673013006474418,
"grad_norm": 1.880037187273125,
"learning_rate": 1.2972972972972975e-05,
"loss": 0.9248,
"step": 24
},
{
"epoch": 0.020492721881744187,
"grad_norm": 2.049138558758981,
"learning_rate": 1.3513513513513515e-05,
"loss": 0.9251,
"step": 25
},
{
"epoch": 0.021312430757013955,
"grad_norm": 2.168069324043155,
"learning_rate": 1.4054054054054055e-05,
"loss": 0.8794,
"step": 26
},
{
"epoch": 0.02213213963228372,
"grad_norm": 2.2331426454485137,
"learning_rate": 1.4594594594594596e-05,
"loss": 0.9224,
"step": 27
},
{
"epoch": 0.02295184850755349,
"grad_norm": 1.471811736467114,
"learning_rate": 1.5135135135135138e-05,
"loss": 0.9205,
"step": 28
},
{
"epoch": 0.023771557382823258,
"grad_norm": 1.8074008214649548,
"learning_rate": 1.5675675675675676e-05,
"loss": 0.9298,
"step": 29
},
{
"epoch": 0.024591266258093023,
"grad_norm": 2.0091050868048734,
"learning_rate": 1.6216216216216218e-05,
"loss": 0.8866,
"step": 30
},
{
"epoch": 0.025410975133362792,
"grad_norm": 2.0793311814633872,
"learning_rate": 1.6756756756756757e-05,
"loss": 0.9063,
"step": 31
},
{
"epoch": 0.02623068400863256,
"grad_norm": 1.8768929995467092,
"learning_rate": 1.72972972972973e-05,
"loss": 0.9039,
"step": 32
},
{
"epoch": 0.027050392883902326,
"grad_norm": 2.0754249763027586,
"learning_rate": 1.783783783783784e-05,
"loss": 0.8843,
"step": 33
},
{
"epoch": 0.027870101759172095,
"grad_norm": 1.5858880843860839,
"learning_rate": 1.8378378378378383e-05,
"loss": 0.8448,
"step": 34
},
{
"epoch": 0.02868981063444186,
"grad_norm": 1.7108892037214618,
"learning_rate": 1.891891891891892e-05,
"loss": 0.874,
"step": 35
},
{
"epoch": 0.02950951950971163,
"grad_norm": 2.066992768178221,
"learning_rate": 1.9459459459459463e-05,
"loss": 0.8442,
"step": 36
},
{
"epoch": 0.030329228384981397,
"grad_norm": 1.7102591496135033,
"learning_rate": 2e-05,
"loss": 0.8933,
"step": 37
},
{
"epoch": 0.031148937260251162,
"grad_norm": 1.5567338554670231,
"learning_rate": 1.999996467887392e-05,
"loss": 0.8688,
"step": 38
},
{
"epoch": 0.03196864613552093,
"grad_norm": 1.2948090193308381,
"learning_rate": 1.9999858715745195e-05,
"loss": 0.8616,
"step": 39
},
{
"epoch": 0.0327883550107907,
"grad_norm": 1.5644818961289724,
"learning_rate": 1.9999682111362368e-05,
"loss": 0.8676,
"step": 40
},
{
"epoch": 0.03360806388606047,
"grad_norm": 1.5203193890130469,
"learning_rate": 1.9999434866973018e-05,
"loss": 0.8285,
"step": 41
},
{
"epoch": 0.03442777276133023,
"grad_norm": 1.3242737437748129,
"learning_rate": 1.999911698432373e-05,
"loss": 0.8656,
"step": 42
},
{
"epoch": 0.0352474816366,
"grad_norm": 1.5359667969972672,
"learning_rate": 1.9998728465660105e-05,
"loss": 0.8348,
"step": 43
},
{
"epoch": 0.03606719051186977,
"grad_norm": 1.7494905485086696,
"learning_rate": 1.9998269313726722e-05,
"loss": 0.8221,
"step": 44
},
{
"epoch": 0.036886899387139536,
"grad_norm": 1.682405417056653,
"learning_rate": 1.9997739531767132e-05,
"loss": 0.8359,
"step": 45
},
{
"epoch": 0.037706608262409305,
"grad_norm": 1.5063006612020524,
"learning_rate": 1.999713912352384e-05,
"loss": 0.8711,
"step": 46
},
{
"epoch": 0.038526317137679074,
"grad_norm": 2.4222314538853555,
"learning_rate": 1.9996468093238256e-05,
"loss": 0.8742,
"step": 47
},
{
"epoch": 0.039346026012948836,
"grad_norm": 1.9253939093745664,
"learning_rate": 1.9995726445650698e-05,
"loss": 0.7809,
"step": 48
},
{
"epoch": 0.040165734888218604,
"grad_norm": 1.7543954048539314,
"learning_rate": 1.999491418600033e-05,
"loss": 0.8104,
"step": 49
},
{
"epoch": 0.04098544376348837,
"grad_norm": 1.5722673780078342,
"learning_rate": 1.999403132002513e-05,
"loss": 0.8343,
"step": 50
},
{
"epoch": 0.04180515263875814,
"grad_norm": 1.729748964002606,
"learning_rate": 1.9993077853961874e-05,
"loss": 0.8534,
"step": 51
},
{
"epoch": 0.04262486151402791,
"grad_norm": 1.7106014776901128,
"learning_rate": 1.999205379454605e-05,
"loss": 0.8114,
"step": 52
},
{
"epoch": 0.04344457038929768,
"grad_norm": 1.4209419376765775,
"learning_rate": 1.999095914901185e-05,
"loss": 0.848,
"step": 53
},
{
"epoch": 0.04426427926456744,
"grad_norm": 1.566401132778228,
"learning_rate": 1.9989793925092092e-05,
"loss": 0.8124,
"step": 54
},
{
"epoch": 0.04508398813983721,
"grad_norm": 1.5535119497298269,
"learning_rate": 1.9988558131018188e-05,
"loss": 0.8208,
"step": 55
},
{
"epoch": 0.04590369701510698,
"grad_norm": 1.734794325945457,
"learning_rate": 1.9987251775520056e-05,
"loss": 0.8654,
"step": 56
},
{
"epoch": 0.04672340589037675,
"grad_norm": 1.816040936441046,
"learning_rate": 1.9985874867826095e-05,
"loss": 0.8038,
"step": 57
},
{
"epoch": 0.047543114765646516,
"grad_norm": 1.5719375289823379,
"learning_rate": 1.9984427417663085e-05,
"loss": 0.8721,
"step": 58
},
{
"epoch": 0.04836282364091628,
"grad_norm": 1.1560906825621926,
"learning_rate": 1.9982909435256143e-05,
"loss": 0.8083,
"step": 59
},
{
"epoch": 0.049182532516186046,
"grad_norm": 1.3052018898979492,
"learning_rate": 1.998132093132864e-05,
"loss": 0.8126,
"step": 60
},
{
"epoch": 0.050002241391455815,
"grad_norm": 1.4249853339922471,
"learning_rate": 1.9979661917102116e-05,
"loss": 0.8493,
"step": 61
},
{
"epoch": 0.050821950266725584,
"grad_norm": 1.7480925046739464,
"learning_rate": 1.997793240429623e-05,
"loss": 0.8011,
"step": 62
},
{
"epoch": 0.05164165914199535,
"grad_norm": 1.4518906151625661,
"learning_rate": 1.9976132405128646e-05,
"loss": 0.8395,
"step": 63
},
{
"epoch": 0.05246136801726512,
"grad_norm": 1.3474727838215297,
"learning_rate": 1.997426193231497e-05,
"loss": 0.8371,
"step": 64
},
{
"epoch": 0.05328107689253488,
"grad_norm": 1.3010387844893,
"learning_rate": 1.9972320999068636e-05,
"loss": 0.8169,
"step": 65
},
{
"epoch": 0.05410078576780465,
"grad_norm": 1.4592654991040828,
"learning_rate": 1.997030961910084e-05,
"loss": 0.855,
"step": 66
},
{
"epoch": 0.05492049464307442,
"grad_norm": 1.3540724517148188,
"learning_rate": 1.9968227806620413e-05,
"loss": 0.8169,
"step": 67
},
{
"epoch": 0.05574020351834419,
"grad_norm": 1.4238089589066887,
"learning_rate": 1.9966075576333757e-05,
"loss": 0.8162,
"step": 68
},
{
"epoch": 0.05655991239361396,
"grad_norm": 1.2863889124951176,
"learning_rate": 1.9963852943444703e-05,
"loss": 0.8385,
"step": 69
},
{
"epoch": 0.05737962126888372,
"grad_norm": 1.2900098060384804,
"learning_rate": 1.996155992365444e-05,
"loss": 0.8178,
"step": 70
},
{
"epoch": 0.05819933014415349,
"grad_norm": 1.2134760053051137,
"learning_rate": 1.995919653316137e-05,
"loss": 0.8103,
"step": 71
},
{
"epoch": 0.05901903901942326,
"grad_norm": 1.3018350176391125,
"learning_rate": 1.9956762788661018e-05,
"loss": 0.8264,
"step": 72
},
{
"epoch": 0.059838747894693026,
"grad_norm": 1.3194214577555265,
"learning_rate": 1.9954258707345903e-05,
"loss": 0.8204,
"step": 73
},
{
"epoch": 0.060658456769962794,
"grad_norm": 1.3149613648555498,
"learning_rate": 1.9951684306905418e-05,
"loss": 0.8002,
"step": 74
},
{
"epoch": 0.06147816564523256,
"grad_norm": 1.4805348390963582,
"learning_rate": 1.9949039605525705e-05,
"loss": 0.8158,
"step": 75
},
{
"epoch": 0.062297874520502325,
"grad_norm": 1.465213161966698,
"learning_rate": 1.9946324621889538e-05,
"loss": 0.8342,
"step": 76
},
{
"epoch": 0.0631175833957721,
"grad_norm": 2.0904994603827576,
"learning_rate": 1.9943539375176164e-05,
"loss": 0.8355,
"step": 77
},
{
"epoch": 0.06393729227104186,
"grad_norm": 1.4153393765052338,
"learning_rate": 1.99406838850612e-05,
"loss": 0.7919,
"step": 78
},
{
"epoch": 0.06475700114631162,
"grad_norm": 1.5713332461986556,
"learning_rate": 1.9937758171716468e-05,
"loss": 0.8043,
"step": 79
},
{
"epoch": 0.0655767100215814,
"grad_norm": 1.404917863748091,
"learning_rate": 1.9934762255809866e-05,
"loss": 0.8099,
"step": 80
},
{
"epoch": 0.06639641889685116,
"grad_norm": 1.4653173459292306,
"learning_rate": 1.993169615850522e-05,
"loss": 0.8277,
"step": 81
},
{
"epoch": 0.06721612777212094,
"grad_norm": 1.3572261026226928,
"learning_rate": 1.992855990146213e-05,
"loss": 0.813,
"step": 82
},
{
"epoch": 0.0680358366473907,
"grad_norm": 1.3754270735628051,
"learning_rate": 1.9925353506835827e-05,
"loss": 0.791,
"step": 83
},
{
"epoch": 0.06885554552266046,
"grad_norm": 1.5319259607332218,
"learning_rate": 1.9922076997277e-05,
"loss": 0.7573,
"step": 84
},
{
"epoch": 0.06967525439793024,
"grad_norm": 1.1847409870962453,
"learning_rate": 1.9918730395931648e-05,
"loss": 0.8091,
"step": 85
},
{
"epoch": 0.0704949632732,
"grad_norm": 1.4870926698444034,
"learning_rate": 1.9915313726440922e-05,
"loss": 0.7594,
"step": 86
},
{
"epoch": 0.07131467214846977,
"grad_norm": 1.447539434567867,
"learning_rate": 1.9911827012940944e-05,
"loss": 0.807,
"step": 87
},
{
"epoch": 0.07213438102373954,
"grad_norm": 1.2610605708125813,
"learning_rate": 1.9908270280062643e-05,
"loss": 0.8141,
"step": 88
},
{
"epoch": 0.07295408989900931,
"grad_norm": 1.5798821696581762,
"learning_rate": 1.9904643552931578e-05,
"loss": 0.7791,
"step": 89
},
{
"epoch": 0.07377379877427907,
"grad_norm": 1.5432883811841962,
"learning_rate": 1.9900946857167768e-05,
"loss": 0.817,
"step": 90
},
{
"epoch": 0.07459350764954883,
"grad_norm": 1.8975434471652286,
"learning_rate": 1.989718021888551e-05,
"loss": 0.7997,
"step": 91
},
{
"epoch": 0.07541321652481861,
"grad_norm": 1.3517603027652132,
"learning_rate": 1.9893343664693177e-05,
"loss": 0.7848,
"step": 92
},
{
"epoch": 0.07623292540008837,
"grad_norm": 1.7612241756262808,
"learning_rate": 1.9889437221693053e-05,
"loss": 0.7822,
"step": 93
},
{
"epoch": 0.07705263427535815,
"grad_norm": 1.8272659804358784,
"learning_rate": 1.9885460917481137e-05,
"loss": 0.8389,
"step": 94
},
{
"epoch": 0.07787234315062791,
"grad_norm": 1.3965452105268819,
"learning_rate": 1.988141478014693e-05,
"loss": 0.7495,
"step": 95
},
{
"epoch": 0.07869205202589767,
"grad_norm": 1.3856369443916108,
"learning_rate": 1.9877298838273263e-05,
"loss": 0.7873,
"step": 96
},
{
"epoch": 0.07951176090116745,
"grad_norm": 1.8193996238319035,
"learning_rate": 1.9873113120936074e-05,
"loss": 0.8076,
"step": 97
},
{
"epoch": 0.08033146977643721,
"grad_norm": 1.569837992958631,
"learning_rate": 1.9868857657704215e-05,
"loss": 0.7938,
"step": 98
},
{
"epoch": 0.08115117865170698,
"grad_norm": 1.6600411042461471,
"learning_rate": 1.9864532478639233e-05,
"loss": 0.8025,
"step": 99
},
{
"epoch": 0.08197088752697675,
"grad_norm": 1.7834826083049724,
"learning_rate": 1.986013761429517e-05,
"loss": 0.829,
"step": 100
},
{
"epoch": 0.08279059640224651,
"grad_norm": 1.3764538071872083,
"learning_rate": 1.9855673095718338e-05,
"loss": 0.7941,
"step": 101
},
{
"epoch": 0.08361030527751628,
"grad_norm": 1.2857559010310884,
"learning_rate": 1.9851138954447097e-05,
"loss": 0.765,
"step": 102
},
{
"epoch": 0.08443001415278605,
"grad_norm": 1.4262737963036518,
"learning_rate": 1.9846535222511648e-05,
"loss": 0.7948,
"step": 103
},
{
"epoch": 0.08524972302805582,
"grad_norm": 1.6124069305203288,
"learning_rate": 1.9841861932433784e-05,
"loss": 0.8138,
"step": 104
},
{
"epoch": 0.08606943190332558,
"grad_norm": 1.264116283649808,
"learning_rate": 1.983711911722669e-05,
"loss": 0.8081,
"step": 105
},
{
"epoch": 0.08688914077859536,
"grad_norm": 1.3286329568952977,
"learning_rate": 1.9832306810394665e-05,
"loss": 0.7994,
"step": 106
},
{
"epoch": 0.08770884965386512,
"grad_norm": 1.2720227149725776,
"learning_rate": 1.982742504593294e-05,
"loss": 0.7857,
"step": 107
},
{
"epoch": 0.08852855852913488,
"grad_norm": 1.4300065773886619,
"learning_rate": 1.9822473858327398e-05,
"loss": 0.826,
"step": 108
},
{
"epoch": 0.08934826740440466,
"grad_norm": 1.350947436974937,
"learning_rate": 1.9817453282554334e-05,
"loss": 0.7659,
"step": 109
},
{
"epoch": 0.09016797627967442,
"grad_norm": 1.4318006683411986,
"learning_rate": 1.981236335408024e-05,
"loss": 0.763,
"step": 110
},
{
"epoch": 0.0909876851549442,
"grad_norm": 1.2618856607772395,
"learning_rate": 1.98072041088615e-05,
"loss": 0.758,
"step": 111
},
{
"epoch": 0.09180739403021396,
"grad_norm": 1.2498215141801015,
"learning_rate": 1.98019755833442e-05,
"loss": 0.7655,
"step": 112
},
{
"epoch": 0.09262710290548372,
"grad_norm": 1.5779861261563592,
"learning_rate": 1.9796677814463812e-05,
"loss": 0.8063,
"step": 113
},
{
"epoch": 0.0934468117807535,
"grad_norm": 1.838226857058528,
"learning_rate": 1.9791310839644976e-05,
"loss": 0.7804,
"step": 114
},
{
"epoch": 0.09426652065602326,
"grad_norm": 1.4989872477349564,
"learning_rate": 1.97858746968012e-05,
"loss": 0.7842,
"step": 115
},
{
"epoch": 0.09508622953129303,
"grad_norm": 1.5368313044552222,
"learning_rate": 1.9780369424334633e-05,
"loss": 0.8593,
"step": 116
},
{
"epoch": 0.0959059384065628,
"grad_norm": 1.3564903659493959,
"learning_rate": 1.9774795061135754e-05,
"loss": 0.7764,
"step": 117
},
{
"epoch": 0.09672564728183256,
"grad_norm": 1.1596566338703373,
"learning_rate": 1.9769151646583122e-05,
"loss": 0.7621,
"step": 118
},
{
"epoch": 0.09754535615710233,
"grad_norm": 1.2172994750615027,
"learning_rate": 1.9763439220543084e-05,
"loss": 0.8485,
"step": 119
},
{
"epoch": 0.09836506503237209,
"grad_norm": 1.4907977725003783,
"learning_rate": 1.9757657823369508e-05,
"loss": 0.7774,
"step": 120
},
{
"epoch": 0.09918477390764187,
"grad_norm": 1.1459287702430172,
"learning_rate": 1.9751807495903484e-05,
"loss": 0.7987,
"step": 121
},
{
"epoch": 0.10000448278291163,
"grad_norm": 1.3718386781064213,
"learning_rate": 1.9745888279473046e-05,
"loss": 0.7289,
"step": 122
},
{
"epoch": 0.1008241916581814,
"grad_norm": 1.152621116625451,
"learning_rate": 1.973990021589287e-05,
"loss": 0.7531,
"step": 123
},
{
"epoch": 0.10164390053345117,
"grad_norm": 1.2745018824522787,
"learning_rate": 1.9733843347463982e-05,
"loss": 0.7887,
"step": 124
},
{
"epoch": 0.10246360940872093,
"grad_norm": 1.2821241089400977,
"learning_rate": 1.972771771697347e-05,
"loss": 0.7608,
"step": 125
},
{
"epoch": 0.1032833182839907,
"grad_norm": 1.4488253633573054,
"learning_rate": 1.9721523367694167e-05,
"loss": 0.8365,
"step": 126
},
{
"epoch": 0.10410302715926047,
"grad_norm": 1.2898276627395178,
"learning_rate": 1.971526034338435e-05,
"loss": 0.8164,
"step": 127
},
{
"epoch": 0.10492273603453024,
"grad_norm": 1.1959656433907844,
"learning_rate": 1.970892868828743e-05,
"loss": 0.7483,
"step": 128
},
{
"epoch": 0.1057424449098,
"grad_norm": 1.720309463436613,
"learning_rate": 1.9702528447131647e-05,
"loss": 0.8176,
"step": 129
},
{
"epoch": 0.10656215378506977,
"grad_norm": 1.2753185558497269,
"learning_rate": 1.969605966512975e-05,
"loss": 0.7824,
"step": 130
},
{
"epoch": 0.10738186266033954,
"grad_norm": 1.3835228106685054,
"learning_rate": 1.9689522387978666e-05,
"loss": 0.7832,
"step": 131
},
{
"epoch": 0.1082015715356093,
"grad_norm": 1.1983777929022796,
"learning_rate": 1.9682916661859197e-05,
"loss": 0.8016,
"step": 132
},
{
"epoch": 0.10902128041087908,
"grad_norm": 1.5367419648739533,
"learning_rate": 1.967624253343568e-05,
"loss": 0.7394,
"step": 133
},
{
"epoch": 0.10984098928614884,
"grad_norm": 1.2506416581349882,
"learning_rate": 1.9669500049855657e-05,
"loss": 0.7681,
"step": 134
},
{
"epoch": 0.1106606981614186,
"grad_norm": 1.3364031418400169,
"learning_rate": 1.9662689258749555e-05,
"loss": 0.8437,
"step": 135
},
{
"epoch": 0.11148040703668838,
"grad_norm": 1.3587715567625391,
"learning_rate": 1.9655810208230334e-05,
"loss": 0.7668,
"step": 136
},
{
"epoch": 0.11230011591195814,
"grad_norm": 1.7435578973716974,
"learning_rate": 1.9648862946893158e-05,
"loss": 0.78,
"step": 137
},
{
"epoch": 0.11311982478722792,
"grad_norm": 1.4203645128112254,
"learning_rate": 1.964184752381504e-05,
"loss": 0.7668,
"step": 138
},
{
"epoch": 0.11393953366249768,
"grad_norm": 1.283806821660531,
"learning_rate": 1.963476398855452e-05,
"loss": 0.7864,
"step": 139
},
{
"epoch": 0.11475924253776744,
"grad_norm": 1.3466411084236394,
"learning_rate": 1.9627612391151278e-05,
"loss": 0.7748,
"step": 140
},
{
"epoch": 0.11557895141303721,
"grad_norm": 1.7951232798175083,
"learning_rate": 1.962039278212581e-05,
"loss": 0.7711,
"step": 141
},
{
"epoch": 0.11639866028830698,
"grad_norm": 1.517672712526967,
"learning_rate": 1.961310521247906e-05,
"loss": 0.7883,
"step": 142
},
{
"epoch": 0.11721836916357675,
"grad_norm": 1.3477054489648024,
"learning_rate": 1.9605749733692063e-05,
"loss": 0.7965,
"step": 143
},
{
"epoch": 0.11803807803884651,
"grad_norm": 1.3495349470712177,
"learning_rate": 1.9598326397725577e-05,
"loss": 0.7986,
"step": 144
},
{
"epoch": 0.11885778691411629,
"grad_norm": 1.1177242113723307,
"learning_rate": 1.9590835257019715e-05,
"loss": 0.7922,
"step": 145
},
{
"epoch": 0.11967749578938605,
"grad_norm": 1.2970157842078793,
"learning_rate": 1.958327636449359e-05,
"loss": 0.7577,
"step": 146
},
{
"epoch": 0.12049720466465581,
"grad_norm": 1.3731551698076407,
"learning_rate": 1.9575649773544914e-05,
"loss": 0.8174,
"step": 147
},
{
"epoch": 0.12131691353992559,
"grad_norm": 1.2064802992358998,
"learning_rate": 1.9567955538049643e-05,
"loss": 0.7837,
"step": 148
},
{
"epoch": 0.12213662241519535,
"grad_norm": 1.1506629460178057,
"learning_rate": 1.9560193712361597e-05,
"loss": 0.7687,
"step": 149
},
{
"epoch": 0.12295633129046513,
"grad_norm": 1.2871919785250847,
"learning_rate": 1.9552364351312056e-05,
"loss": 0.7714,
"step": 150
},
{
"epoch": 0.12377604016573489,
"grad_norm": 1.326698454969782,
"learning_rate": 1.954446751020939e-05,
"loss": 0.7833,
"step": 151
},
{
"epoch": 0.12459574904100465,
"grad_norm": 1.4103500339112234,
"learning_rate": 1.953650324483866e-05,
"loss": 0.8166,
"step": 152
},
{
"epoch": 0.12541545791627443,
"grad_norm": 1.3875114825870618,
"learning_rate": 1.9528471611461235e-05,
"loss": 0.8002,
"step": 153
},
{
"epoch": 0.1262351667915442,
"grad_norm": 1.3369502030231106,
"learning_rate": 1.952037266681438e-05,
"loss": 0.7688,
"step": 154
},
{
"epoch": 0.12705487566681395,
"grad_norm": 1.350414346771441,
"learning_rate": 1.9512206468110863e-05,
"loss": 0.8198,
"step": 155
},
{
"epoch": 0.12787458454208372,
"grad_norm": 1.1267138205680407,
"learning_rate": 1.9503973073038554e-05,
"loss": 0.7861,
"step": 156
},
{
"epoch": 0.1286942934173535,
"grad_norm": 1.1777303042065168,
"learning_rate": 1.949567253976001e-05,
"loss": 0.7946,
"step": 157
},
{
"epoch": 0.12951400229262325,
"grad_norm": 1.038575956729356,
"learning_rate": 1.948730492691206e-05,
"loss": 0.7487,
"step": 158
},
{
"epoch": 0.13033371116789302,
"grad_norm": 1.22971469720092,
"learning_rate": 1.9478870293605416e-05,
"loss": 0.7631,
"step": 159
},
{
"epoch": 0.1311534200431628,
"grad_norm": 1.0785792529426395,
"learning_rate": 1.947036869942422e-05,
"loss": 0.7888,
"step": 160
},
{
"epoch": 0.13197312891843258,
"grad_norm": 1.614389323377392,
"learning_rate": 1.9461800204425653e-05,
"loss": 0.7489,
"step": 161
},
{
"epoch": 0.13279283779370232,
"grad_norm": 1.2443041618160595,
"learning_rate": 1.9453164869139488e-05,
"loss": 0.9025,
"step": 162
},
{
"epoch": 0.1336125466689721,
"grad_norm": 1.3670900677876743,
"learning_rate": 1.9444462754567682e-05,
"loss": 0.8123,
"step": 163
},
{
"epoch": 0.13443225554424187,
"grad_norm": 1.5658238045136883,
"learning_rate": 1.9435693922183935e-05,
"loss": 0.7674,
"step": 164
},
{
"epoch": 0.13525196441951162,
"grad_norm": 1.3536671136972342,
"learning_rate": 1.9426858433933248e-05,
"loss": 0.766,
"step": 165
},
{
"epoch": 0.1360716732947814,
"grad_norm": 1.4231629020132193,
"learning_rate": 1.94179563522315e-05,
"loss": 0.7952,
"step": 166
},
{
"epoch": 0.13689138217005117,
"grad_norm": 1.4978139492566758,
"learning_rate": 1.9408987739965006e-05,
"loss": 0.7589,
"step": 167
},
{
"epoch": 0.13771109104532092,
"grad_norm": 1.5968332310772775,
"learning_rate": 1.9399952660490057e-05,
"loss": 0.7748,
"step": 168
},
{
"epoch": 0.1385307999205907,
"grad_norm": 1.4111730772516582,
"learning_rate": 1.9390851177632496e-05,
"loss": 0.7635,
"step": 169
},
{
"epoch": 0.13935050879586047,
"grad_norm": 1.252969688606119,
"learning_rate": 1.9381683355687245e-05,
"loss": 0.765,
"step": 170
},
{
"epoch": 0.14017021767113025,
"grad_norm": 1.1527972148084198,
"learning_rate": 1.937244925941786e-05,
"loss": 0.8022,
"step": 171
},
{
"epoch": 0.1409899265464,
"grad_norm": 1.1663896053258958,
"learning_rate": 1.9363148954056077e-05,
"loss": 0.7876,
"step": 172
},
{
"epoch": 0.14180963542166977,
"grad_norm": 1.2812752733100474,
"learning_rate": 1.9353782505301352e-05,
"loss": 0.7698,
"step": 173
},
{
"epoch": 0.14262934429693955,
"grad_norm": 1.1413732379185346,
"learning_rate": 1.9344349979320386e-05,
"loss": 0.7849,
"step": 174
},
{
"epoch": 0.1434490531722093,
"grad_norm": 1.1920463077200083,
"learning_rate": 1.9334851442746665e-05,
"loss": 0.7804,
"step": 175
},
{
"epoch": 0.14426876204747907,
"grad_norm": 1.1715039605005428,
"learning_rate": 1.9325286962679993e-05,
"loss": 0.7864,
"step": 176
},
{
"epoch": 0.14508847092274885,
"grad_norm": 1.097979299635343,
"learning_rate": 1.9315656606686012e-05,
"loss": 0.7789,
"step": 177
},
{
"epoch": 0.14590817979801862,
"grad_norm": 1.1906344465845144,
"learning_rate": 1.930596044279572e-05,
"loss": 0.7559,
"step": 178
},
{
"epoch": 0.14672788867328837,
"grad_norm": 1.063000407175785,
"learning_rate": 1.9296198539505013e-05,
"loss": 0.7629,
"step": 179
},
{
"epoch": 0.14754759754855815,
"grad_norm": 1.1143786049124003,
"learning_rate": 1.9286370965774166e-05,
"loss": 0.7825,
"step": 180
},
{
"epoch": 0.14836730642382792,
"grad_norm": 1.184422288092832,
"learning_rate": 1.9276477791027374e-05,
"loss": 0.7627,
"step": 181
},
{
"epoch": 0.14918701529909767,
"grad_norm": 1.411996735986641,
"learning_rate": 1.9266519085152254e-05,
"loss": 0.785,
"step": 182
},
{
"epoch": 0.15000672417436745,
"grad_norm": 1.372038246706677,
"learning_rate": 1.9256494918499348e-05,
"loss": 0.7914,
"step": 183
},
{
"epoch": 0.15082643304963722,
"grad_norm": 1.3348589539694835,
"learning_rate": 1.9246405361881623e-05,
"loss": 0.749,
"step": 184
},
{
"epoch": 0.15164614192490697,
"grad_norm": 1.5415333806984186,
"learning_rate": 1.9236250486573978e-05,
"loss": 0.7651,
"step": 185
},
{
"epoch": 0.15246585080017674,
"grad_norm": 1.2557803970058343,
"learning_rate": 1.9226030364312747e-05,
"loss": 0.7819,
"step": 186
},
{
"epoch": 0.15328555967544652,
"grad_norm": 1.0940635592143146,
"learning_rate": 1.9215745067295168e-05,
"loss": 0.7484,
"step": 187
},
{
"epoch": 0.1541052685507163,
"grad_norm": 1.396897824369515,
"learning_rate": 1.92053946681789e-05,
"loss": 0.7698,
"step": 188
},
{
"epoch": 0.15492497742598604,
"grad_norm": 1.390551679174612,
"learning_rate": 1.919497924008149e-05,
"loss": 0.7453,
"step": 189
},
{
"epoch": 0.15574468630125582,
"grad_norm": 1.458155621850461,
"learning_rate": 1.918449885657987e-05,
"loss": 0.7259,
"step": 190
},
{
"epoch": 0.1565643951765256,
"grad_norm": 1.2486201905765646,
"learning_rate": 1.917395359170983e-05,
"loss": 0.7304,
"step": 191
},
{
"epoch": 0.15738410405179534,
"grad_norm": 1.2857773425779437,
"learning_rate": 1.9163343519965494e-05,
"loss": 0.7753,
"step": 192
},
{
"epoch": 0.15820381292706512,
"grad_norm": 1.585450755985716,
"learning_rate": 1.91526687162988e-05,
"loss": 0.797,
"step": 193
},
{
"epoch": 0.1590235218023349,
"grad_norm": 1.3883735779161643,
"learning_rate": 1.914192925611896e-05,
"loss": 0.7705,
"step": 194
},
{
"epoch": 0.15984323067760467,
"grad_norm": 1.1056871815340525,
"learning_rate": 1.913112521529195e-05,
"loss": 0.7713,
"step": 195
},
{
"epoch": 0.16066293955287442,
"grad_norm": 1.1780203201808228,
"learning_rate": 1.9120256670139942e-05,
"loss": 0.7443,
"step": 196
},
{
"epoch": 0.1614826484281442,
"grad_norm": 1.2437600398366402,
"learning_rate": 1.9109323697440782e-05,
"loss": 0.7584,
"step": 197
},
{
"epoch": 0.16230235730341397,
"grad_norm": 1.2758067099309862,
"learning_rate": 1.909832637442746e-05,
"loss": 0.7391,
"step": 198
},
{
"epoch": 0.16312206617868372,
"grad_norm": 1.307989702571116,
"learning_rate": 1.9087264778787534e-05,
"loss": 0.8058,
"step": 199
},
{
"epoch": 0.1639417750539535,
"grad_norm": 1.2522240361680288,
"learning_rate": 1.9076138988662615e-05,
"loss": 0.7726,
"step": 200
},
{
"epoch": 0.16476148392922327,
"grad_norm": 1.326536738376905,
"learning_rate": 1.9064949082647785e-05,
"loss": 0.7653,
"step": 201
},
{
"epoch": 0.16558119280449302,
"grad_norm": 1.2582509952602114,
"learning_rate": 1.9053695139791058e-05,
"loss": 0.7535,
"step": 202
},
{
"epoch": 0.1664009016797628,
"grad_norm": 1.325267565251269,
"learning_rate": 1.904237723959283e-05,
"loss": 0.7725,
"step": 203
},
{
"epoch": 0.16722061055503257,
"grad_norm": 1.2350320222463889,
"learning_rate": 1.903099546200529e-05,
"loss": 0.7603,
"step": 204
},
{
"epoch": 0.16804031943030234,
"grad_norm": 1.1295109418727558,
"learning_rate": 1.901954988743188e-05,
"loss": 0.7044,
"step": 205
},
{
"epoch": 0.1688600283055721,
"grad_norm": 1.5001298494910937,
"learning_rate": 1.9008040596726713e-05,
"loss": 0.7604,
"step": 206
},
{
"epoch": 0.16967973718084187,
"grad_norm": 1.048888405634734,
"learning_rate": 1.8996467671194017e-05,
"loss": 0.7374,
"step": 207
},
{
"epoch": 0.17049944605611164,
"grad_norm": 1.3129466681568118,
"learning_rate": 1.898483119258754e-05,
"loss": 0.765,
"step": 208
},
{
"epoch": 0.1713191549313814,
"grad_norm": 1.193236419546853,
"learning_rate": 1.8973131243109987e-05,
"loss": 0.7588,
"step": 209
},
{
"epoch": 0.17213886380665117,
"grad_norm": 1.1573803650943608,
"learning_rate": 1.896136790541244e-05,
"loss": 0.766,
"step": 210
},
{
"epoch": 0.17295857268192094,
"grad_norm": 1.319205011784051,
"learning_rate": 1.8949541262593764e-05,
"loss": 0.7654,
"step": 211
},
{
"epoch": 0.17377828155719072,
"grad_norm": 1.1670276218721964,
"learning_rate": 1.8937651398200027e-05,
"loss": 0.775,
"step": 212
},
{
"epoch": 0.17459799043246046,
"grad_norm": 1.3538137911239734,
"learning_rate": 1.892569839622391e-05,
"loss": 0.7565,
"step": 213
},
{
"epoch": 0.17541769930773024,
"grad_norm": 1.3378168256501808,
"learning_rate": 1.8913682341104107e-05,
"loss": 0.7591,
"step": 214
},
{
"epoch": 0.17623740818300002,
"grad_norm": 1.4001135240594023,
"learning_rate": 1.8901603317724742e-05,
"loss": 0.7671,
"step": 215
},
{
"epoch": 0.17705711705826976,
"grad_norm": 1.2330529865325488,
"learning_rate": 1.8889461411414754e-05,
"loss": 0.7409,
"step": 216
},
{
"epoch": 0.17787682593353954,
"grad_norm": 1.2036657517267269,
"learning_rate": 1.8877256707947308e-05,
"loss": 0.7579,
"step": 217
},
{
"epoch": 0.17869653480880932,
"grad_norm": 1.1096175989523498,
"learning_rate": 1.886498929353917e-05,
"loss": 0.7201,
"step": 218
},
{
"epoch": 0.17951624368407906,
"grad_norm": 1.1702620886507735,
"learning_rate": 1.8852659254850128e-05,
"loss": 0.7608,
"step": 219
},
{
"epoch": 0.18033595255934884,
"grad_norm": 1.1618543221505961,
"learning_rate": 1.8840266678982343e-05,
"loss": 0.7585,
"step": 220
},
{
"epoch": 0.18115566143461861,
"grad_norm": 1.129628587487089,
"learning_rate": 1.882781165347977e-05,
"loss": 0.7427,
"step": 221
},
{
"epoch": 0.1819753703098884,
"grad_norm": 1.134165412978348,
"learning_rate": 1.8815294266327507e-05,
"loss": 0.78,
"step": 222
},
{
"epoch": 0.18279507918515814,
"grad_norm": 1.2428018037444357,
"learning_rate": 1.88027146059512e-05,
"loss": 0.7346,
"step": 223
},
{
"epoch": 0.1836147880604279,
"grad_norm": 1.3489999363819276,
"learning_rate": 1.8790072761216406e-05,
"loss": 0.786,
"step": 224
},
{
"epoch": 0.1844344969356977,
"grad_norm": 1.3208315778175526,
"learning_rate": 1.8777368821427954e-05,
"loss": 0.7495,
"step": 225
},
{
"epoch": 0.18525420581096744,
"grad_norm": 1.4840730999361789,
"learning_rate": 1.8764602876329346e-05,
"loss": 0.7661,
"step": 226
},
{
"epoch": 0.1860739146862372,
"grad_norm": 1.2152631644856686,
"learning_rate": 1.8751775016102087e-05,
"loss": 0.7603,
"step": 227
},
{
"epoch": 0.186893623561507,
"grad_norm": 1.4516865857108312,
"learning_rate": 1.8738885331365073e-05,
"loss": 0.7269,
"step": 228
},
{
"epoch": 0.18771333243677676,
"grad_norm": 1.284318664669841,
"learning_rate": 1.872593391317394e-05,
"loss": 0.7378,
"step": 229
},
{
"epoch": 0.1885330413120465,
"grad_norm": 1.1111212091170015,
"learning_rate": 1.871292085302042e-05,
"loss": 0.7376,
"step": 230
},
{
"epoch": 0.1893527501873163,
"grad_norm": 1.3327879378631937,
"learning_rate": 1.8699846242831707e-05,
"loss": 0.7626,
"step": 231
},
{
"epoch": 0.19017245906258606,
"grad_norm": 1.3950256729152688,
"learning_rate": 1.8686710174969786e-05,
"loss": 0.7837,
"step": 232
},
{
"epoch": 0.1909921679378558,
"grad_norm": 1.109586554294812,
"learning_rate": 1.86735127422308e-05,
"loss": 0.7543,
"step": 233
},
{
"epoch": 0.1918118768131256,
"grad_norm": 1.6284101907473822,
"learning_rate": 1.866025403784439e-05,
"loss": 0.7584,
"step": 234
},
{
"epoch": 0.19263158568839536,
"grad_norm": 1.3472252155582047,
"learning_rate": 1.8646934155473025e-05,
"loss": 0.7508,
"step": 235
},
{
"epoch": 0.1934512945636651,
"grad_norm": 1.2433191287827514,
"learning_rate": 1.8633553189211353e-05,
"loss": 0.7541,
"step": 236
},
{
"epoch": 0.19427100343893489,
"grad_norm": 1.1569625781773563,
"learning_rate": 1.862011123358554e-05,
"loss": 0.721,
"step": 237
},
{
"epoch": 0.19509071231420466,
"grad_norm": 1.0464539639356059,
"learning_rate": 1.8606608383552583e-05,
"loss": 0.77,
"step": 238
},
{
"epoch": 0.19591042118947444,
"grad_norm": 1.199746020608613,
"learning_rate": 1.8593044734499657e-05,
"loss": 0.7584,
"step": 239
},
{
"epoch": 0.19673013006474419,
"grad_norm": 1.251450934233781,
"learning_rate": 1.8579420382243433e-05,
"loss": 0.7233,
"step": 240
},
{
"epoch": 0.19754983894001396,
"grad_norm": 1.1084218164211046,
"learning_rate": 1.8565735423029406e-05,
"loss": 0.7788,
"step": 241
},
{
"epoch": 0.19836954781528374,
"grad_norm": 1.2347361201870448,
"learning_rate": 1.8551989953531204e-05,
"loss": 0.7978,
"step": 242
},
{
"epoch": 0.19918925669055348,
"grad_norm": 1.1402144542644108,
"learning_rate": 1.8538184070849926e-05,
"loss": 0.7616,
"step": 243
},
{
"epoch": 0.20000896556582326,
"grad_norm": 1.1662737994181693,
"learning_rate": 1.8524317872513434e-05,
"loss": 0.7181,
"step": 244
},
{
"epoch": 0.20082867444109304,
"grad_norm": 1.1177301882337476,
"learning_rate": 1.8510391456475674e-05,
"loss": 0.7352,
"step": 245
},
{
"epoch": 0.2016483833163628,
"grad_norm": 1.599475127420748,
"learning_rate": 1.8496404921115993e-05,
"loss": 0.7414,
"step": 246
},
{
"epoch": 0.20246809219163256,
"grad_norm": 1.106299509246655,
"learning_rate": 1.8482358365238414e-05,
"loss": 0.7339,
"step": 247
},
{
"epoch": 0.20328780106690233,
"grad_norm": 1.555565751879419,
"learning_rate": 1.8468251888070982e-05,
"loss": 0.7666,
"step": 248
},
{
"epoch": 0.2041075099421721,
"grad_norm": 1.2207504805308689,
"learning_rate": 1.845408558926502e-05,
"loss": 0.741,
"step": 249
},
{
"epoch": 0.20492721881744186,
"grad_norm": 1.18082604015543,
"learning_rate": 1.8439859568894464e-05,
"loss": 0.7829,
"step": 250
},
{
"epoch": 0.20574692769271163,
"grad_norm": 1.1151358100509305,
"learning_rate": 1.842557392745512e-05,
"loss": 0.757,
"step": 251
},
{
"epoch": 0.2065666365679814,
"grad_norm": 1.0916201198843662,
"learning_rate": 1.8411228765863973e-05,
"loss": 0.7548,
"step": 252
},
{
"epoch": 0.20738634544325116,
"grad_norm": 1.1185666568619232,
"learning_rate": 1.839682418545848e-05,
"loss": 0.7163,
"step": 253
},
{
"epoch": 0.20820605431852093,
"grad_norm": 1.167511943178696,
"learning_rate": 1.838236028799584e-05,
"loss": 0.7514,
"step": 254
},
{
"epoch": 0.2090257631937907,
"grad_norm": 1.1750726836533263,
"learning_rate": 1.8367837175652284e-05,
"loss": 0.7515,
"step": 255
},
{
"epoch": 0.20984547206906048,
"grad_norm": 1.0678585750169112,
"learning_rate": 1.8353254951022342e-05,
"loss": 0.7602,
"step": 256
},
{
"epoch": 0.21066518094433023,
"grad_norm": 1.0232358718785153,
"learning_rate": 1.833861371711814e-05,
"loss": 0.7391,
"step": 257
},
{
"epoch": 0.2114848898196,
"grad_norm": 1.1125091966481833,
"learning_rate": 1.832391357736865e-05,
"loss": 0.7383,
"step": 258
},
{
"epoch": 0.21230459869486978,
"grad_norm": 1.0726945879923688,
"learning_rate": 1.8309154635618967e-05,
"loss": 0.7779,
"step": 259
},
{
"epoch": 0.21312430757013953,
"grad_norm": 1.1895019083488085,
"learning_rate": 1.829433699612958e-05,
"loss": 0.7861,
"step": 260
},
{
"epoch": 0.2139440164454093,
"grad_norm": 1.3760744106270784,
"learning_rate": 1.827946076357564e-05,
"loss": 0.7524,
"step": 261
},
{
"epoch": 0.21476372532067908,
"grad_norm": 1.3957491151421677,
"learning_rate": 1.8264526043046193e-05,
"loss": 0.7678,
"step": 262
},
{
"epoch": 0.21558343419594886,
"grad_norm": 1.2975658092679558,
"learning_rate": 1.824953294004347e-05,
"loss": 0.7448,
"step": 263
},
{
"epoch": 0.2164031430712186,
"grad_norm": 1.2131336289076444,
"learning_rate": 1.8234481560482136e-05,
"loss": 0.7847,
"step": 264
},
{
"epoch": 0.21722285194648838,
"grad_norm": 1.3527270869946275,
"learning_rate": 1.8219372010688516e-05,
"loss": 0.7359,
"step": 265
},
{
"epoch": 0.21804256082175816,
"grad_norm": 1.4407084614444403,
"learning_rate": 1.8204204397399875e-05,
"loss": 0.7545,
"step": 266
},
{
"epoch": 0.2188622696970279,
"grad_norm": 1.169809335746366,
"learning_rate": 1.8188978827763654e-05,
"loss": 0.7177,
"step": 267
},
{
"epoch": 0.21968197857229768,
"grad_norm": 1.100792465173885,
"learning_rate": 1.8173695409336703e-05,
"loss": 0.7388,
"step": 268
},
{
"epoch": 0.22050168744756746,
"grad_norm": 1.4330282110430612,
"learning_rate": 1.815835425008453e-05,
"loss": 0.7442,
"step": 269
},
{
"epoch": 0.2213213963228372,
"grad_norm": 1.308338161462054,
"learning_rate": 1.814295545838054e-05,
"loss": 0.7304,
"step": 270
},
{
"epoch": 0.22214110519810698,
"grad_norm": 1.062023124889891,
"learning_rate": 1.8127499143005266e-05,
"loss": 0.7622,
"step": 271
},
{
"epoch": 0.22296081407337676,
"grad_norm": 1.500406025215467,
"learning_rate": 1.81119854131456e-05,
"loss": 0.707,
"step": 272
},
{
"epoch": 0.22378052294864653,
"grad_norm": 1.2571243546224666,
"learning_rate": 1.809641437839403e-05,
"loss": 0.7425,
"step": 273
},
{
"epoch": 0.22460023182391628,
"grad_norm": 1.4111561156495562,
"learning_rate": 1.8080786148747842e-05,
"loss": 0.7372,
"step": 274
},
{
"epoch": 0.22541994069918606,
"grad_norm": 1.1056793027165572,
"learning_rate": 1.8065100834608378e-05,
"loss": 0.7954,
"step": 275
},
{
"epoch": 0.22623964957445583,
"grad_norm": 1.5147248282727566,
"learning_rate": 1.8049358546780227e-05,
"loss": 0.73,
"step": 276
},
{
"epoch": 0.22705935844972558,
"grad_norm": 1.3343439411064637,
"learning_rate": 1.8033559396470455e-05,
"loss": 0.7186,
"step": 277
},
{
"epoch": 0.22787906732499535,
"grad_norm": 1.2461133896878915,
"learning_rate": 1.8017703495287815e-05,
"loss": 0.7368,
"step": 278
},
{
"epoch": 0.22869877620026513,
"grad_norm": 1.4322049495661178,
"learning_rate": 1.8001790955241972e-05,
"loss": 0.7257,
"step": 279
},
{
"epoch": 0.22951848507553488,
"grad_norm": 1.3600448553141125,
"learning_rate": 1.7985821888742687e-05,
"loss": 0.7414,
"step": 280
},
{
"epoch": 0.23033819395080465,
"grad_norm": 1.3273390863293366,
"learning_rate": 1.7969796408599042e-05,
"loss": 0.7371,
"step": 281
},
{
"epoch": 0.23115790282607443,
"grad_norm": 1.5358373624995807,
"learning_rate": 1.795371462801864e-05,
"loss": 0.7416,
"step": 282
},
{
"epoch": 0.2319776117013442,
"grad_norm": 1.303859023890605,
"learning_rate": 1.79375766606068e-05,
"loss": 0.7185,
"step": 283
},
{
"epoch": 0.23279732057661395,
"grad_norm": 1.0884984487972824,
"learning_rate": 1.7921382620365757e-05,
"loss": 0.7229,
"step": 284
},
{
"epoch": 0.23361702945188373,
"grad_norm": 1.0712476566733449,
"learning_rate": 1.7905132621693862e-05,
"loss": 0.7128,
"step": 285
},
{
"epoch": 0.2344367383271535,
"grad_norm": 1.0795078605302708,
"learning_rate": 1.788882677938476e-05,
"loss": 0.7323,
"step": 286
},
{
"epoch": 0.23525644720242325,
"grad_norm": 1.2579723423519416,
"learning_rate": 1.78724652086266e-05,
"loss": 0.7507,
"step": 287
},
{
"epoch": 0.23607615607769303,
"grad_norm": 1.0776344487859562,
"learning_rate": 1.7856048025001195e-05,
"loss": 0.7515,
"step": 288
},
{
"epoch": 0.2368958649529628,
"grad_norm": 1.0389458353142942,
"learning_rate": 1.7839575344483237e-05,
"loss": 0.7534,
"step": 289
},
{
"epoch": 0.23771557382823258,
"grad_norm": 1.0379302135531618,
"learning_rate": 1.7823047283439444e-05,
"loss": 0.7197,
"step": 290
},
{
"epoch": 0.23853528270350233,
"grad_norm": 1.010455612578165,
"learning_rate": 1.7806463958627765e-05,
"loss": 0.7305,
"step": 291
},
{
"epoch": 0.2393549915787721,
"grad_norm": 1.4438268918227726,
"learning_rate": 1.7789825487196538e-05,
"loss": 0.784,
"step": 292
},
{
"epoch": 0.24017470045404188,
"grad_norm": 1.2763943377985623,
"learning_rate": 1.7773131986683675e-05,
"loss": 0.7542,
"step": 293
},
{
"epoch": 0.24099440932931163,
"grad_norm": 1.243466110953911,
"learning_rate": 1.775638357501582e-05,
"loss": 0.7439,
"step": 294
},
{
"epoch": 0.2418141182045814,
"grad_norm": 1.2615651265111143,
"learning_rate": 1.7739580370507533e-05,
"loss": 0.7333,
"step": 295
},
{
"epoch": 0.24263382707985118,
"grad_norm": 1.129495176581153,
"learning_rate": 1.7722722491860427e-05,
"loss": 0.7265,
"step": 296
},
{
"epoch": 0.24345353595512093,
"grad_norm": 1.0602472392591882,
"learning_rate": 1.7705810058162354e-05,
"loss": 0.7081,
"step": 297
},
{
"epoch": 0.2442732448303907,
"grad_norm": 1.2176403243614142,
"learning_rate": 1.768884318888656e-05,
"loss": 0.7373,
"step": 298
},
{
"epoch": 0.24509295370566048,
"grad_norm": 1.3496176663326471,
"learning_rate": 1.7671822003890825e-05,
"loss": 0.7137,
"step": 299
},
{
"epoch": 0.24591266258093025,
"grad_norm": 1.6236403686173448,
"learning_rate": 1.7654746623416637e-05,
"loss": 0.7432,
"step": 300
},
{
"epoch": 0.2467323714562,
"grad_norm": 1.3284610983953054,
"learning_rate": 1.7637617168088327e-05,
"loss": 0.7218,
"step": 301
},
{
"epoch": 0.24755208033146978,
"grad_norm": 1.301817695632009,
"learning_rate": 1.762043375891223e-05,
"loss": 0.7345,
"step": 302
},
{
"epoch": 0.24837178920673955,
"grad_norm": 1.1172675319808283,
"learning_rate": 1.7603196517275808e-05,
"loss": 0.7304,
"step": 303
},
{
"epoch": 0.2491914980820093,
"grad_norm": 1.3293871724347375,
"learning_rate": 1.758590556494683e-05,
"loss": 0.7237,
"step": 304
},
{
"epoch": 0.2500112069572791,
"grad_norm": 1.2588955646488025,
"learning_rate": 1.756856102407247e-05,
"loss": 0.757,
"step": 305
},
{
"epoch": 0.25083091583254885,
"grad_norm": 1.1437116421002054,
"learning_rate": 1.7551163017178473e-05,
"loss": 0.7316,
"step": 306
},
{
"epoch": 0.2516506247078186,
"grad_norm": 1.3843672734729005,
"learning_rate": 1.753371166716828e-05,
"loss": 0.7428,
"step": 307
},
{
"epoch": 0.2524703335830884,
"grad_norm": 1.4719769071099351,
"learning_rate": 1.7516207097322154e-05,
"loss": 0.7453,
"step": 308
},
{
"epoch": 0.2532900424583581,
"grad_norm": 1.2425882734913514,
"learning_rate": 1.749864943129632e-05,
"loss": 0.7439,
"step": 309
},
{
"epoch": 0.2541097513336279,
"grad_norm": 1.1062741360580992,
"learning_rate": 1.748103879312209e-05,
"loss": 0.7181,
"step": 310
},
{
"epoch": 0.2549294602088977,
"grad_norm": 1.2078126311989654,
"learning_rate": 1.746337530720497e-05,
"loss": 0.7782,
"step": 311
},
{
"epoch": 0.25574916908416745,
"grad_norm": 1.1175977498695826,
"learning_rate": 1.7445659098323807e-05,
"loss": 0.743,
"step": 312
},
{
"epoch": 0.2565688779594372,
"grad_norm": 1.3982891802536677,
"learning_rate": 1.7427890291629895e-05,
"loss": 0.7436,
"step": 313
},
{
"epoch": 0.257388586834707,
"grad_norm": 1.0557561304713365,
"learning_rate": 1.7410069012646076e-05,
"loss": 0.789,
"step": 314
},
{
"epoch": 0.2582082957099768,
"grad_norm": 1.0662740874323304,
"learning_rate": 1.7392195387265888e-05,
"loss": 0.7838,
"step": 315
},
{
"epoch": 0.2590280045852465,
"grad_norm": 1.009777078461186,
"learning_rate": 1.737426954175264e-05,
"loss": 0.7333,
"step": 316
},
{
"epoch": 0.25984771346051627,
"grad_norm": 1.0418727834469572,
"learning_rate": 1.7356291602738542e-05,
"loss": 0.7285,
"step": 317
},
{
"epoch": 0.26066742233578605,
"grad_norm": 1.0886150256944194,
"learning_rate": 1.733826169722381e-05,
"loss": 0.7306,
"step": 318
},
{
"epoch": 0.2614871312110558,
"grad_norm": 1.1958960910904468,
"learning_rate": 1.732017995257575e-05,
"loss": 0.7278,
"step": 319
},
{
"epoch": 0.2623068400863256,
"grad_norm": 1.1390713799207537,
"learning_rate": 1.7302046496527882e-05,
"loss": 0.7632,
"step": 320
},
{
"epoch": 0.2631265489615954,
"grad_norm": 1.1640918173335415,
"learning_rate": 1.7283861457179022e-05,
"loss": 0.7781,
"step": 321
},
{
"epoch": 0.26394625783686515,
"grad_norm": 1.2921689832623544,
"learning_rate": 1.726562496299239e-05,
"loss": 0.7381,
"step": 322
},
{
"epoch": 0.26476596671213487,
"grad_norm": 1.0051942452859242,
"learning_rate": 1.724733714279468e-05,
"loss": 0.729,
"step": 323
},
{
"epoch": 0.26558567558740465,
"grad_norm": 1.1352614057081478,
"learning_rate": 1.7228998125775175e-05,
"loss": 0.7227,
"step": 324
},
{
"epoch": 0.2664053844626744,
"grad_norm": 1.1108632627846275,
"learning_rate": 1.721060804148482e-05,
"loss": 0.7574,
"step": 325
},
{
"epoch": 0.2672250933379442,
"grad_norm": 1.2925573596152042,
"learning_rate": 1.7192167019835314e-05,
"loss": 0.7918,
"step": 326
},
{
"epoch": 0.268044802213214,
"grad_norm": 1.0615952710554621,
"learning_rate": 1.717367519109819e-05,
"loss": 0.7065,
"step": 327
},
{
"epoch": 0.26886451108848375,
"grad_norm": 1.1032084129174033,
"learning_rate": 1.7155132685903888e-05,
"loss": 0.7402,
"step": 328
},
{
"epoch": 0.2696842199637535,
"grad_norm": 1.247191970959756,
"learning_rate": 1.713653963524084e-05,
"loss": 0.7553,
"step": 329
},
{
"epoch": 0.27050392883902324,
"grad_norm": 0.9216973158791136,
"learning_rate": 1.7117896170454542e-05,
"loss": 0.7467,
"step": 330
},
{
"epoch": 0.271323637714293,
"grad_norm": 1.0233508095130186,
"learning_rate": 1.7099202423246632e-05,
"loss": 0.7446,
"step": 331
},
{
"epoch": 0.2721433465895628,
"grad_norm": 0.9393837898276846,
"learning_rate": 1.7080458525673948e-05,
"loss": 0.7121,
"step": 332
},
{
"epoch": 0.27296305546483257,
"grad_norm": 1.101687910379085,
"learning_rate": 1.7061664610147605e-05,
"loss": 0.7119,
"step": 333
},
{
"epoch": 0.27378276434010235,
"grad_norm": 1.1252230476839957,
"learning_rate": 1.7042820809432057e-05,
"loss": 0.6956,
"step": 334
},
{
"epoch": 0.2746024732153721,
"grad_norm": 1.1476380056838789,
"learning_rate": 1.702392725664415e-05,
"loss": 0.7913,
"step": 335
},
{
"epoch": 0.27542218209064184,
"grad_norm": 1.231225623673207,
"learning_rate": 1.70049840852522e-05,
"loss": 0.7607,
"step": 336
},
{
"epoch": 0.2762418909659116,
"grad_norm": 1.1990083735582506,
"learning_rate": 1.6985991429075038e-05,
"loss": 0.7036,
"step": 337
},
{
"epoch": 0.2770615998411814,
"grad_norm": 1.3760329408743888,
"learning_rate": 1.6966949422281058e-05,
"loss": 0.7242,
"step": 338
},
{
"epoch": 0.27788130871645117,
"grad_norm": 1.6758031223580756,
"learning_rate": 1.6947858199387296e-05,
"loss": 0.7657,
"step": 339
},
{
"epoch": 0.27870101759172095,
"grad_norm": 1.1348149301895436,
"learning_rate": 1.692871789525844e-05,
"loss": 0.7293,
"step": 340
},
{
"epoch": 0.2795207264669907,
"grad_norm": 1.076254923302961,
"learning_rate": 1.6909528645105908e-05,
"loss": 0.6945,
"step": 341
},
{
"epoch": 0.2803404353422605,
"grad_norm": 1.155962968992227,
"learning_rate": 1.6890290584486892e-05,
"loss": 0.7234,
"step": 342
},
{
"epoch": 0.2811601442175302,
"grad_norm": 1.1154109966936674,
"learning_rate": 1.687100384930338e-05,
"loss": 0.7395,
"step": 343
},
{
"epoch": 0.2819798530928,
"grad_norm": 1.1642366412904992,
"learning_rate": 1.685166857580122e-05,
"loss": 0.7242,
"step": 344
},
{
"epoch": 0.28279956196806977,
"grad_norm": 1.0878357397294232,
"learning_rate": 1.683228490056913e-05,
"loss": 0.7671,
"step": 345
},
{
"epoch": 0.28361927084333954,
"grad_norm": 1.2109130277932583,
"learning_rate": 1.6812852960537763e-05,
"loss": 0.74,
"step": 346
},
{
"epoch": 0.2844389797186093,
"grad_norm": 1.2554887399525227,
"learning_rate": 1.6793372892978716e-05,
"loss": 0.7612,
"step": 347
},
{
"epoch": 0.2852586885938791,
"grad_norm": 1.117259274783754,
"learning_rate": 1.6773844835503574e-05,
"loss": 0.7123,
"step": 348
},
{
"epoch": 0.28607839746914887,
"grad_norm": 1.0652151784160404,
"learning_rate": 1.6754268926062936e-05,
"loss": 0.7123,
"step": 349
},
{
"epoch": 0.2868981063444186,
"grad_norm": 1.0049815147078915,
"learning_rate": 1.6734645302945434e-05,
"loss": 0.7089,
"step": 350
},
{
"epoch": 0.28771781521968837,
"grad_norm": 1.3790835665028844,
"learning_rate": 1.671497410477676e-05,
"loss": 0.725,
"step": 351
},
{
"epoch": 0.28853752409495814,
"grad_norm": 1.1036671134262883,
"learning_rate": 1.6695255470518687e-05,
"loss": 0.7307,
"step": 352
},
{
"epoch": 0.2893572329702279,
"grad_norm": 0.9816946084723515,
"learning_rate": 1.6675489539468094e-05,
"loss": 0.7504,
"step": 353
},
{
"epoch": 0.2901769418454977,
"grad_norm": 1.1060156546071784,
"learning_rate": 1.665567645125596e-05,
"loss": 0.7376,
"step": 354
},
{
"epoch": 0.29099665072076747,
"grad_norm": 1.008749999873428,
"learning_rate": 1.6635816345846413e-05,
"loss": 0.775,
"step": 355
},
{
"epoch": 0.29181635959603724,
"grad_norm": 1.2283221913345734,
"learning_rate": 1.66159093635357e-05,
"loss": 0.7233,
"step": 356
},
{
"epoch": 0.29263606847130696,
"grad_norm": 1.5910908498224698,
"learning_rate": 1.659595564495124e-05,
"loss": 0.7691,
"step": 357
},
{
"epoch": 0.29345577734657674,
"grad_norm": 1.2587854652473174,
"learning_rate": 1.6575955331050585e-05,
"loss": 0.7699,
"step": 358
},
{
"epoch": 0.2942754862218465,
"grad_norm": 1.5467901971390223,
"learning_rate": 1.6555908563120457e-05,
"loss": 0.7375,
"step": 359
},
{
"epoch": 0.2950951950971163,
"grad_norm": 1.3693777821298927,
"learning_rate": 1.6535815482775745e-05,
"loss": 0.7552,
"step": 360
},
{
"epoch": 0.29591490397238607,
"grad_norm": 1.0598058518008313,
"learning_rate": 1.6515676231958488e-05,
"loss": 0.7377,
"step": 361
},
{
"epoch": 0.29673461284765584,
"grad_norm": 1.138653423427796,
"learning_rate": 1.6495490952936898e-05,
"loss": 0.7033,
"step": 362
},
{
"epoch": 0.2975543217229256,
"grad_norm": 0.9692694133176057,
"learning_rate": 1.647525978830432e-05,
"loss": 0.7042,
"step": 363
},
{
"epoch": 0.29837403059819534,
"grad_norm": 1.0418397682960119,
"learning_rate": 1.645498288097827e-05,
"loss": 0.7602,
"step": 364
},
{
"epoch": 0.2991937394734651,
"grad_norm": 1.0537794177478244,
"learning_rate": 1.6434660374199377e-05,
"loss": 0.7354,
"step": 365
},
{
"epoch": 0.3000134483487349,
"grad_norm": 1.0946083916458766,
"learning_rate": 1.641429241153041e-05,
"loss": 0.7152,
"step": 366
},
{
"epoch": 0.30083315722400467,
"grad_norm": 1.259061612531398,
"learning_rate": 1.6393879136855247e-05,
"loss": 0.7376,
"step": 367
},
{
"epoch": 0.30165286609927444,
"grad_norm": 1.247543588849218,
"learning_rate": 1.6373420694377857e-05,
"loss": 0.7606,
"step": 368
},
{
"epoch": 0.3024725749745442,
"grad_norm": 1.0714211644352196,
"learning_rate": 1.6352917228621284e-05,
"loss": 0.7574,
"step": 369
},
{
"epoch": 0.30329228384981394,
"grad_norm": 1.1195959087092493,
"learning_rate": 1.633236888442663e-05,
"loss": 0.6844,
"step": 370
},
{
"epoch": 0.3041119927250837,
"grad_norm": 0.9818790789263798,
"learning_rate": 1.631177580695202e-05,
"loss": 0.7023,
"step": 371
},
{
"epoch": 0.3049317016003535,
"grad_norm": 1.057011665382933,
"learning_rate": 1.6291138141671598e-05,
"loss": 0.711,
"step": 372
},
{
"epoch": 0.30575141047562326,
"grad_norm": 1.2925296159779756,
"learning_rate": 1.6270456034374477e-05,
"loss": 0.7362,
"step": 373
},
{
"epoch": 0.30657111935089304,
"grad_norm": 0.9813332914224199,
"learning_rate": 1.6249729631163718e-05,
"loss": 0.7143,
"step": 374
},
{
"epoch": 0.3073908282261628,
"grad_norm": 1.1694282928349735,
"learning_rate": 1.6228959078455306e-05,
"loss": 0.6965,
"step": 375
},
{
"epoch": 0.3082105371014326,
"grad_norm": 1.0368946510682566,
"learning_rate": 1.62081445229771e-05,
"loss": 0.7323,
"step": 376
},
{
"epoch": 0.3090302459767023,
"grad_norm": 1.4302047097764494,
"learning_rate": 1.6187286111767812e-05,
"loss": 0.7463,
"step": 377
},
{
"epoch": 0.3098499548519721,
"grad_norm": 1.2112825873347262,
"learning_rate": 1.616638399217595e-05,
"loss": 0.7544,
"step": 378
},
{
"epoch": 0.31066966372724186,
"grad_norm": 1.2841071119830416,
"learning_rate": 1.61454383118588e-05,
"loss": 0.7211,
"step": 379
},
{
"epoch": 0.31148937260251164,
"grad_norm": 1.1373190945003797,
"learning_rate": 1.6124449218781358e-05,
"loss": 0.704,
"step": 380
},
{
"epoch": 0.3123090814777814,
"grad_norm": 1.1361884990490227,
"learning_rate": 1.6103416861215314e-05,
"loss": 0.7737,
"step": 381
},
{
"epoch": 0.3131287903530512,
"grad_norm": 1.2427299010715174,
"learning_rate": 1.608234138773797e-05,
"loss": 0.7166,
"step": 382
},
{
"epoch": 0.31394849922832097,
"grad_norm": 1.0872388025166924,
"learning_rate": 1.6061222947231224e-05,
"loss": 0.6976,
"step": 383
},
{
"epoch": 0.3147682081035907,
"grad_norm": 1.1508774421630683,
"learning_rate": 1.6040061688880494e-05,
"loss": 0.7914,
"step": 384
},
{
"epoch": 0.31558791697886046,
"grad_norm": 1.2405915582919167,
"learning_rate": 1.6018857762173672e-05,
"loss": 0.7317,
"step": 385
},
{
"epoch": 0.31640762585413024,
"grad_norm": 1.1867806270398218,
"learning_rate": 1.5997611316900075e-05,
"loss": 0.7186,
"step": 386
},
{
"epoch": 0.3172273347294,
"grad_norm": 1.2223676338157003,
"learning_rate": 1.5976322503149373e-05,
"loss": 0.7507,
"step": 387
},
{
"epoch": 0.3180470436046698,
"grad_norm": 1.0670232704518172,
"learning_rate": 1.5954991471310542e-05,
"loss": 0.7388,
"step": 388
},
{
"epoch": 0.31886675247993956,
"grad_norm": 1.056135120064279,
"learning_rate": 1.5933618372070805e-05,
"loss": 0.7534,
"step": 389
},
{
"epoch": 0.31968646135520934,
"grad_norm": 0.9367539562277986,
"learning_rate": 1.5912203356414532e-05,
"loss": 0.7461,
"step": 390
},
{
"epoch": 0.32050617023047906,
"grad_norm": 1.042778382318825,
"learning_rate": 1.589074657562223e-05,
"loss": 0.7055,
"step": 391
},
{
"epoch": 0.32132587910574884,
"grad_norm": 1.0128242483834669,
"learning_rate": 1.5869248181269427e-05,
"loss": 0.7411,
"step": 392
},
{
"epoch": 0.3221455879810186,
"grad_norm": 0.9644057186534029,
"learning_rate": 1.5847708325225618e-05,
"loss": 0.7602,
"step": 393
},
{
"epoch": 0.3229652968562884,
"grad_norm": 1.0544545552539342,
"learning_rate": 1.5826127159653203e-05,
"loss": 0.7133,
"step": 394
},
{
"epoch": 0.32378500573155816,
"grad_norm": 1.4470618499019512,
"learning_rate": 1.5804504837006396e-05,
"loss": 0.6837,
"step": 395
},
{
"epoch": 0.32460471460682794,
"grad_norm": 1.1382072505416,
"learning_rate": 1.578284151003015e-05,
"loss": 0.7217,
"step": 396
},
{
"epoch": 0.3254244234820977,
"grad_norm": 1.0699112411097602,
"learning_rate": 1.5761137331759084e-05,
"loss": 0.7051,
"step": 397
},
{
"epoch": 0.32624413235736743,
"grad_norm": 1.096759612689205,
"learning_rate": 1.5739392455516408e-05,
"loss": 0.7073,
"step": 398
},
{
"epoch": 0.3270638412326372,
"grad_norm": 1.3152249754831804,
"learning_rate": 1.571760703491282e-05,
"loss": 0.7375,
"step": 399
},
{
"epoch": 0.327883550107907,
"grad_norm": 1.2525579297408636,
"learning_rate": 1.5695781223845442e-05,
"loss": 0.7356,
"step": 400
},
{
"epoch": 0.32870325898317676,
"grad_norm": 1.1809465219797077,
"learning_rate": 1.5673915176496716e-05,
"loss": 0.7283,
"step": 401
},
{
"epoch": 0.32952296785844654,
"grad_norm": 1.1298080924142333,
"learning_rate": 1.5652009047333322e-05,
"loss": 0.7453,
"step": 402
},
{
"epoch": 0.3303426767337163,
"grad_norm": 1.3591366731971994,
"learning_rate": 1.56300629911051e-05,
"loss": 0.7246,
"step": 403
},
{
"epoch": 0.33116238560898603,
"grad_norm": 1.2500207753380739,
"learning_rate": 1.560807716284392e-05,
"loss": 0.7495,
"step": 404
},
{
"epoch": 0.3319820944842558,
"grad_norm": 1.2332080123216231,
"learning_rate": 1.5586051717862634e-05,
"loss": 0.742,
"step": 405
},
{
"epoch": 0.3328018033595256,
"grad_norm": 1.0704537953512225,
"learning_rate": 1.5563986811753948e-05,
"loss": 0.7241,
"step": 406
},
{
"epoch": 0.33362151223479536,
"grad_norm": 1.1774525207234647,
"learning_rate": 1.554188260038932e-05,
"loss": 0.7377,
"step": 407
},
{
"epoch": 0.33444122111006513,
"grad_norm": 1.2635720249942302,
"learning_rate": 1.551973923991788e-05,
"loss": 0.7601,
"step": 408
},
{
"epoch": 0.3352609299853349,
"grad_norm": 1.1081384843738813,
"learning_rate": 1.5497556886765316e-05,
"loss": 0.7475,
"step": 409
},
{
"epoch": 0.3360806388606047,
"grad_norm": 1.2963322835483428,
"learning_rate": 1.5475335697632768e-05,
"loss": 0.7525,
"step": 410
},
{
"epoch": 0.3369003477358744,
"grad_norm": 1.0914692496421534,
"learning_rate": 1.545307582949571e-05,
"loss": 0.7005,
"step": 411
},
{
"epoch": 0.3377200566111442,
"grad_norm": 0.998863450079788,
"learning_rate": 1.5430777439602875e-05,
"loss": 0.7265,
"step": 412
},
{
"epoch": 0.33853976548641396,
"grad_norm": 1.0225725072009355,
"learning_rate": 1.540844068547511e-05,
"loss": 0.7564,
"step": 413
},
{
"epoch": 0.33935947436168373,
"grad_norm": 0.9450553290666434,
"learning_rate": 1.5386065724904273e-05,
"loss": 0.7184,
"step": 414
},
{
"epoch": 0.3401791832369535,
"grad_norm": 1.3828770917184838,
"learning_rate": 1.536365271595212e-05,
"loss": 0.757,
"step": 415
},
{
"epoch": 0.3409988921122233,
"grad_norm": 1.156215045557116,
"learning_rate": 1.5341201816949208e-05,
"loss": 0.729,
"step": 416
},
{
"epoch": 0.34181860098749306,
"grad_norm": 1.3233452079305341,
"learning_rate": 1.5318713186493736e-05,
"loss": 0.7236,
"step": 417
},
{
"epoch": 0.3426383098627628,
"grad_norm": 1.0703750607992697,
"learning_rate": 1.529618698345045e-05,
"loss": 0.7317,
"step": 418
},
{
"epoch": 0.34345801873803256,
"grad_norm": 1.0948427095856885,
"learning_rate": 1.5273623366949525e-05,
"loss": 0.6771,
"step": 419
},
{
"epoch": 0.34427772761330233,
"grad_norm": 1.0148078626943782,
"learning_rate": 1.5251022496385433e-05,
"loss": 0.6882,
"step": 420
},
{
"epoch": 0.3450974364885721,
"grad_norm": 1.1019967233780132,
"learning_rate": 1.5228384531415809e-05,
"loss": 0.7551,
"step": 421
},
{
"epoch": 0.3459171453638419,
"grad_norm": 1.2538128755726932,
"learning_rate": 1.5205709631960337e-05,
"loss": 0.7325,
"step": 422
},
{
"epoch": 0.34673685423911166,
"grad_norm": 1.0612942700734946,
"learning_rate": 1.5182997958199617e-05,
"loss": 0.7346,
"step": 423
},
{
"epoch": 0.34755656311438143,
"grad_norm": 1.068959511040185,
"learning_rate": 1.5160249670574026e-05,
"loss": 0.7425,
"step": 424
},
{
"epoch": 0.34837627198965115,
"grad_norm": 1.047264249993212,
"learning_rate": 1.5137464929782586e-05,
"loss": 0.7061,
"step": 425
},
{
"epoch": 0.34919598086492093,
"grad_norm": 1.0413520073606914,
"learning_rate": 1.5114643896781844e-05,
"loss": 0.73,
"step": 426
},
{
"epoch": 0.3500156897401907,
"grad_norm": 0.9705733158335844,
"learning_rate": 1.5091786732784717e-05,
"loss": 0.6757,
"step": 427
},
{
"epoch": 0.3508353986154605,
"grad_norm": 1.1007334341363684,
"learning_rate": 1.5068893599259354e-05,
"loss": 0.6998,
"step": 428
},
{
"epoch": 0.35165510749073026,
"grad_norm": 1.03885618227098,
"learning_rate": 1.5045964657928006e-05,
"loss": 0.7338,
"step": 429
},
{
"epoch": 0.35247481636600003,
"grad_norm": 1.1040582135246244,
"learning_rate": 1.5023000070765886e-05,
"loss": 0.7332,
"step": 430
},
{
"epoch": 0.3532945252412698,
"grad_norm": 1.0698177820243113,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.7201,
"step": 431
},
{
"epoch": 0.35411423411653953,
"grad_norm": 1.2287799786184035,
"learning_rate": 1.4976964608108038e-05,
"loss": 0.7061,
"step": 432
},
{
"epoch": 0.3549339429918093,
"grad_norm": 1.379622814875012,
"learning_rate": 1.495389405781719e-05,
"loss": 0.6768,
"step": 433
},
{
"epoch": 0.3557536518670791,
"grad_norm": 1.288557826507751,
"learning_rate": 1.4930788512103018e-05,
"loss": 0.731,
"step": 434
},
{
"epoch": 0.35657336074234885,
"grad_norm": 1.3496778584976903,
"learning_rate": 1.4907648134188304e-05,
"loss": 0.7016,
"step": 435
},
{
"epoch": 0.35739306961761863,
"grad_norm": 1.0377968497133676,
"learning_rate": 1.488447308754189e-05,
"loss": 0.6875,
"step": 436
},
{
"epoch": 0.3582127784928884,
"grad_norm": 0.9544487446483688,
"learning_rate": 1.486126353587752e-05,
"loss": 0.7289,
"step": 437
},
{
"epoch": 0.3590324873681581,
"grad_norm": 1.327671738412972,
"learning_rate": 1.4838019643152699e-05,
"loss": 0.6986,
"step": 438
},
{
"epoch": 0.3598521962434279,
"grad_norm": 0.9857386807710959,
"learning_rate": 1.4814741573567514e-05,
"loss": 0.6971,
"step": 439
},
{
"epoch": 0.3606719051186977,
"grad_norm": 1.0963187554224276,
"learning_rate": 1.47914294915635e-05,
"loss": 0.711,
"step": 440
},
{
"epoch": 0.36149161399396745,
"grad_norm": 1.0973676845659175,
"learning_rate": 1.4768083561822451e-05,
"loss": 0.7151,
"step": 441
},
{
"epoch": 0.36231132286923723,
"grad_norm": 1.2357046291244502,
"learning_rate": 1.4744703949265268e-05,
"loss": 0.7311,
"step": 442
},
{
"epoch": 0.363131031744507,
"grad_norm": 1.0000924948654448,
"learning_rate": 1.4721290819050804e-05,
"loss": 0.6926,
"step": 443
},
{
"epoch": 0.3639507406197768,
"grad_norm": 1.0463831766160467,
"learning_rate": 1.4697844336574685e-05,
"loss": 0.7083,
"step": 444
},
{
"epoch": 0.3647704494950465,
"grad_norm": 0.994289955476413,
"learning_rate": 1.467436466746814e-05,
"loss": 0.6831,
"step": 445
},
{
"epoch": 0.3655901583703163,
"grad_norm": 1.600132874655478,
"learning_rate": 1.465085197759684e-05,
"loss": 0.7377,
"step": 446
},
{
"epoch": 0.36640986724558605,
"grad_norm": 1.5691064241497346,
"learning_rate": 1.4627306433059724e-05,
"loss": 0.7116,
"step": 447
},
{
"epoch": 0.3672295761208558,
"grad_norm": 1.3392318283113458,
"learning_rate": 1.4603728200187823e-05,
"loss": 0.7035,
"step": 448
},
{
"epoch": 0.3680492849961256,
"grad_norm": 1.1393594973183825,
"learning_rate": 1.4580117445543077e-05,
"loss": 0.734,
"step": 449
},
{
"epoch": 0.3688689938713954,
"grad_norm": 1.0381400907889826,
"learning_rate": 1.4556474335917181e-05,
"loss": 0.7458,
"step": 450
},
{
"epoch": 0.36968870274666515,
"grad_norm": 1.0532620172520957,
"learning_rate": 1.4532799038330385e-05,
"loss": 0.722,
"step": 451
},
{
"epoch": 0.3705084116219349,
"grad_norm": 1.1928327761126667,
"learning_rate": 1.4509091720030323e-05,
"loss": 0.7319,
"step": 452
},
{
"epoch": 0.37132812049720465,
"grad_norm": 1.0095580106273234,
"learning_rate": 1.4485352548490827e-05,
"loss": 0.7196,
"step": 453
},
{
"epoch": 0.3721478293724744,
"grad_norm": 1.221030658019384,
"learning_rate": 1.4461581691410757e-05,
"loss": 0.7158,
"step": 454
},
{
"epoch": 0.3729675382477442,
"grad_norm": 1.2553458495613763,
"learning_rate": 1.4437779316712797e-05,
"loss": 0.6961,
"step": 455
},
{
"epoch": 0.373787247123014,
"grad_norm": 1.0329543683911575,
"learning_rate": 1.4413945592542282e-05,
"loss": 0.6891,
"step": 456
},
{
"epoch": 0.37460695599828375,
"grad_norm": 1.198227097224669,
"learning_rate": 1.4390080687266013e-05,
"loss": 0.7142,
"step": 457
},
{
"epoch": 0.37542666487355353,
"grad_norm": 1.1463065128190106,
"learning_rate": 1.4366184769471053e-05,
"loss": 0.7672,
"step": 458
},
{
"epoch": 0.37624637374882325,
"grad_norm": 1.1086539261670978,
"learning_rate": 1.4342258007963541e-05,
"loss": 0.7039,
"step": 459
},
{
"epoch": 0.377066082624093,
"grad_norm": 1.2117053117791163,
"learning_rate": 1.4318300571767514e-05,
"loss": 0.6743,
"step": 460
},
{
"epoch": 0.3778857914993628,
"grad_norm": 1.1985378668403752,
"learning_rate": 1.4294312630123699e-05,
"loss": 0.8615,
"step": 461
},
{
"epoch": 0.3787055003746326,
"grad_norm": 0.9419297745006495,
"learning_rate": 1.4270294352488316e-05,
"loss": 0.7265,
"step": 462
},
{
"epoch": 0.37952520924990235,
"grad_norm": 1.3522357538492233,
"learning_rate": 1.4246245908531883e-05,
"loss": 0.7513,
"step": 463
},
{
"epoch": 0.3803449181251721,
"grad_norm": 0.9890932602178704,
"learning_rate": 1.4222167468138034e-05,
"loss": 0.701,
"step": 464
},
{
"epoch": 0.38116462700044185,
"grad_norm": 1.1622386680794563,
"learning_rate": 1.4198059201402288e-05,
"loss": 0.7151,
"step": 465
},
{
"epoch": 0.3819843358757116,
"grad_norm": 1.037810876017797,
"learning_rate": 1.4173921278630874e-05,
"loss": 0.7147,
"step": 466
},
{
"epoch": 0.3828040447509814,
"grad_norm": 1.071454924560903,
"learning_rate": 1.4149753870339509e-05,
"loss": 0.7348,
"step": 467
},
{
"epoch": 0.3836237536262512,
"grad_norm": 1.1537723474189674,
"learning_rate": 1.4125557147252215e-05,
"loss": 0.7154,
"step": 468
},
{
"epoch": 0.38444346250152095,
"grad_norm": 0.9441167342399105,
"learning_rate": 1.410133128030009e-05,
"loss": 0.7263,
"step": 469
},
{
"epoch": 0.3852631713767907,
"grad_norm": 1.0767147405105144,
"learning_rate": 1.4077076440620112e-05,
"loss": 0.705,
"step": 470
},
{
"epoch": 0.3860828802520605,
"grad_norm": 1.070075967534455,
"learning_rate": 1.4052792799553935e-05,
"loss": 0.7431,
"step": 471
},
{
"epoch": 0.3869025891273302,
"grad_norm": 1.3173189498027416,
"learning_rate": 1.4028480528646669e-05,
"loss": 0.7221,
"step": 472
},
{
"epoch": 0.3877222980026,
"grad_norm": 1.0613708733624905,
"learning_rate": 1.4004139799645669e-05,
"loss": 0.7172,
"step": 473
},
{
"epoch": 0.38854200687786977,
"grad_norm": 1.1060941503102721,
"learning_rate": 1.3979770784499325e-05,
"loss": 0.7128,
"step": 474
},
{
"epoch": 0.38936171575313955,
"grad_norm": 1.064631264322929,
"learning_rate": 1.3955373655355852e-05,
"loss": 0.7264,
"step": 475
},
{
"epoch": 0.3901814246284093,
"grad_norm": 1.0721942171043368,
"learning_rate": 1.3930948584562062e-05,
"loss": 0.7029,
"step": 476
},
{
"epoch": 0.3910011335036791,
"grad_norm": 1.0885903157919428,
"learning_rate": 1.3906495744662159e-05,
"loss": 0.7201,
"step": 477
},
{
"epoch": 0.3918208423789489,
"grad_norm": 1.1726976562618674,
"learning_rate": 1.3882015308396508e-05,
"loss": 0.7325,
"step": 478
},
{
"epoch": 0.3926405512542186,
"grad_norm": 1.0206048482549173,
"learning_rate": 1.3857507448700425e-05,
"loss": 0.7431,
"step": 479
},
{
"epoch": 0.39346026012948837,
"grad_norm": 0.9882098152339169,
"learning_rate": 1.3832972338702954e-05,
"loss": 0.6943,
"step": 480
},
{
"epoch": 0.39427996900475815,
"grad_norm": 1.1797496796272384,
"learning_rate": 1.3808410151725633e-05,
"loss": 0.7051,
"step": 481
},
{
"epoch": 0.3950996778800279,
"grad_norm": 1.246758560192774,
"learning_rate": 1.3783821061281285e-05,
"loss": 0.6913,
"step": 482
},
{
"epoch": 0.3959193867552977,
"grad_norm": 1.0452680181708345,
"learning_rate": 1.3759205241072782e-05,
"loss": 0.7489,
"step": 483
},
{
"epoch": 0.3967390956305675,
"grad_norm": 1.096858942768496,
"learning_rate": 1.3734562864991823e-05,
"loss": 0.7055,
"step": 484
},
{
"epoch": 0.39755880450583725,
"grad_norm": 1.1584082057417153,
"learning_rate": 1.3709894107117698e-05,
"loss": 0.7554,
"step": 485
},
{
"epoch": 0.39837851338110697,
"grad_norm": 1.3310229795859712,
"learning_rate": 1.3685199141716074e-05,
"loss": 0.7218,
"step": 486
},
{
"epoch": 0.39919822225637674,
"grad_norm": 1.221487094994542,
"learning_rate": 1.3660478143237748e-05,
"loss": 0.6972,
"step": 487
},
{
"epoch": 0.4000179311316465,
"grad_norm": 1.1342788826015309,
"learning_rate": 1.3635731286317415e-05,
"loss": 0.713,
"step": 488
},
{
"epoch": 0.4008376400069163,
"grad_norm": 1.1707179724526333,
"learning_rate": 1.3610958745772456e-05,
"loss": 0.7203,
"step": 489
},
{
"epoch": 0.40165734888218607,
"grad_norm": 1.1418211578661164,
"learning_rate": 1.3586160696601667e-05,
"loss": 0.7181,
"step": 490
},
{
"epoch": 0.40247705775745585,
"grad_norm": 1.0885662510936218,
"learning_rate": 1.3561337313984053e-05,
"loss": 0.7123,
"step": 491
},
{
"epoch": 0.4032967666327256,
"grad_norm": 0.9572672378818216,
"learning_rate": 1.3536488773277586e-05,
"loss": 0.7264,
"step": 492
},
{
"epoch": 0.40411647550799534,
"grad_norm": 1.4467547526435829,
"learning_rate": 1.3511615250017948e-05,
"loss": 0.6915,
"step": 493
},
{
"epoch": 0.4049361843832651,
"grad_norm": 1.2046562212650662,
"learning_rate": 1.3486716919917313e-05,
"loss": 0.716,
"step": 494
},
{
"epoch": 0.4057558932585349,
"grad_norm": 1.1761064081197452,
"learning_rate": 1.3461793958863087e-05,
"loss": 0.7102,
"step": 495
},
{
"epoch": 0.40657560213380467,
"grad_norm": 0.9496835749934628,
"learning_rate": 1.3436846542916686e-05,
"loss": 0.6764,
"step": 496
},
{
"epoch": 0.40739531100907445,
"grad_norm": 1.073249972390515,
"learning_rate": 1.3411874848312274e-05,
"loss": 0.7183,
"step": 497
},
{
"epoch": 0.4082150198843442,
"grad_norm": 1.0833826384214176,
"learning_rate": 1.338687905145552e-05,
"loss": 0.7527,
"step": 498
},
{
"epoch": 0.40903472875961394,
"grad_norm": 1.0721966717766631,
"learning_rate": 1.3361859328922368e-05,
"loss": 0.7633,
"step": 499
},
{
"epoch": 0.4098544376348837,
"grad_norm": 0.9657897212304841,
"learning_rate": 1.3336815857457772e-05,
"loss": 0.7066,
"step": 500
},
{
"epoch": 0.4106741465101535,
"grad_norm": 1.033126720091575,
"learning_rate": 1.3311748813974454e-05,
"loss": 0.686,
"step": 501
},
{
"epoch": 0.41149385538542327,
"grad_norm": 1.101350197873293,
"learning_rate": 1.3286658375551654e-05,
"loss": 0.7052,
"step": 502
},
{
"epoch": 0.41231356426069304,
"grad_norm": 1.127236249529374,
"learning_rate": 1.3261544719433878e-05,
"loss": 0.7256,
"step": 503
},
{
"epoch": 0.4131332731359628,
"grad_norm": 1.0159927807299238,
"learning_rate": 1.3236408023029655e-05,
"loss": 0.7448,
"step": 504
},
{
"epoch": 0.4139529820112326,
"grad_norm": 1.169555181448038,
"learning_rate": 1.3211248463910263e-05,
"loss": 0.7458,
"step": 505
},
{
"epoch": 0.4147726908865023,
"grad_norm": 1.095328585418129,
"learning_rate": 1.3186066219808498e-05,
"loss": 0.6653,
"step": 506
},
{
"epoch": 0.4155923997617721,
"grad_norm": 1.132025145078908,
"learning_rate": 1.3160861468617402e-05,
"loss": 0.7021,
"step": 507
},
{
"epoch": 0.41641210863704187,
"grad_norm": 1.162464053424148,
"learning_rate": 1.3135634388389016e-05,
"loss": 0.6927,
"step": 508
},
{
"epoch": 0.41723181751231164,
"grad_norm": 1.0164921340939768,
"learning_rate": 1.311038515733311e-05,
"loss": 0.7169,
"step": 509
},
{
"epoch": 0.4180515263875814,
"grad_norm": 1.1213101750485268,
"learning_rate": 1.3085113953815948e-05,
"loss": 0.7156,
"step": 510
},
{
"epoch": 0.4188712352628512,
"grad_norm": 1.073482301891583,
"learning_rate": 1.3059820956358998e-05,
"loss": 0.7077,
"step": 511
},
{
"epoch": 0.41969094413812097,
"grad_norm": 1.1474422263901856,
"learning_rate": 1.3034506343637687e-05,
"loss": 0.7308,
"step": 512
},
{
"epoch": 0.4205106530133907,
"grad_norm": 1.2080854938120422,
"learning_rate": 1.3009170294480149e-05,
"loss": 0.71,
"step": 513
},
{
"epoch": 0.42133036188866047,
"grad_norm": 0.9381437623811617,
"learning_rate": 1.2983812987865936e-05,
"loss": 0.6977,
"step": 514
},
{
"epoch": 0.42215007076393024,
"grad_norm": 1.068721862187115,
"learning_rate": 1.2958434602924771e-05,
"loss": 0.7132,
"step": 515
},
{
"epoch": 0.4229697796392,
"grad_norm": 1.0593167874016132,
"learning_rate": 1.2933035318935285e-05,
"loss": 0.7446,
"step": 516
},
{
"epoch": 0.4237894885144698,
"grad_norm": 1.1294801183222807,
"learning_rate": 1.290761531532374e-05,
"loss": 0.7165,
"step": 517
},
{
"epoch": 0.42460919738973957,
"grad_norm": 1.0182893534811215,
"learning_rate": 1.2882174771662765e-05,
"loss": 0.73,
"step": 518
},
{
"epoch": 0.42542890626500934,
"grad_norm": 1.0343997053674674,
"learning_rate": 1.2856713867670089e-05,
"loss": 0.7089,
"step": 519
},
{
"epoch": 0.42624861514027906,
"grad_norm": 1.1280462157017102,
"learning_rate": 1.2831232783207278e-05,
"loss": 0.7396,
"step": 520
},
{
"epoch": 0.42706832401554884,
"grad_norm": 1.0644375793654632,
"learning_rate": 1.2805731698278442e-05,
"loss": 0.6935,
"step": 521
},
{
"epoch": 0.4278880328908186,
"grad_norm": 1.1224294993327997,
"learning_rate": 1.2780210793028994e-05,
"loss": 0.7076,
"step": 522
},
{
"epoch": 0.4287077417660884,
"grad_norm": 0.974849894596135,
"learning_rate": 1.2754670247744353e-05,
"loss": 0.7331,
"step": 523
},
{
"epoch": 0.42952745064135817,
"grad_norm": 1.2124639967918331,
"learning_rate": 1.272911024284869e-05,
"loss": 0.6856,
"step": 524
},
{
"epoch": 0.43034715951662794,
"grad_norm": 1.2960107337846904,
"learning_rate": 1.2703530958903631e-05,
"loss": 0.7123,
"step": 525
},
{
"epoch": 0.4311668683918977,
"grad_norm": 0.9301159236067108,
"learning_rate": 1.2677932576606998e-05,
"loss": 0.7187,
"step": 526
},
{
"epoch": 0.43198657726716744,
"grad_norm": 1.0085943291064037,
"learning_rate": 1.265231527679153e-05,
"loss": 0.6681,
"step": 527
},
{
"epoch": 0.4328062861424372,
"grad_norm": 0.9536487518119403,
"learning_rate": 1.2626679240423605e-05,
"loss": 0.6813,
"step": 528
},
{
"epoch": 0.433625995017707,
"grad_norm": 1.0926402375297084,
"learning_rate": 1.260102464860195e-05,
"loss": 0.6748,
"step": 529
},
{
"epoch": 0.43444570389297676,
"grad_norm": 1.2047240039450702,
"learning_rate": 1.2575351682556387e-05,
"loss": 0.7069,
"step": 530
},
{
"epoch": 0.43526541276824654,
"grad_norm": 1.0125001513416305,
"learning_rate": 1.2549660523646527e-05,
"loss": 0.68,
"step": 531
},
{
"epoch": 0.4360851216435163,
"grad_norm": 1.1938817437891787,
"learning_rate": 1.2523951353360504e-05,
"loss": 0.6773,
"step": 532
},
{
"epoch": 0.43690483051878604,
"grad_norm": 1.1863599690404465,
"learning_rate": 1.2498224353313684e-05,
"loss": 0.709,
"step": 533
},
{
"epoch": 0.4377245393940558,
"grad_norm": 0.9964771138837594,
"learning_rate": 1.2472479705247393e-05,
"loss": 0.669,
"step": 534
},
{
"epoch": 0.4385442482693256,
"grad_norm": 1.0718179496212026,
"learning_rate": 1.2446717591027624e-05,
"loss": 0.705,
"step": 535
},
{
"epoch": 0.43936395714459536,
"grad_norm": 1.2127511954058237,
"learning_rate": 1.2420938192643746e-05,
"loss": 0.6944,
"step": 536
},
{
"epoch": 0.44018366601986514,
"grad_norm": 0.9849046194450556,
"learning_rate": 1.2395141692207244e-05,
"loss": 0.7192,
"step": 537
},
{
"epoch": 0.4410033748951349,
"grad_norm": 1.0639972588219935,
"learning_rate": 1.2369328271950404e-05,
"loss": 0.7446,
"step": 538
},
{
"epoch": 0.4418230837704047,
"grad_norm": 0.9942621319464566,
"learning_rate": 1.2343498114225038e-05,
"loss": 0.7205,
"step": 539
},
{
"epoch": 0.4426427926456744,
"grad_norm": 1.0652388102796444,
"learning_rate": 1.2317651401501201e-05,
"loss": 0.7053,
"step": 540
},
{
"epoch": 0.4434625015209442,
"grad_norm": 1.0202292684389789,
"learning_rate": 1.229178831636589e-05,
"loss": 0.6956,
"step": 541
},
{
"epoch": 0.44428221039621396,
"grad_norm": 1.0115158167926785,
"learning_rate": 1.2265909041521764e-05,
"loss": 0.6624,
"step": 542
},
{
"epoch": 0.44510191927148374,
"grad_norm": 1.2787568234777853,
"learning_rate": 1.2240013759785849e-05,
"loss": 0.6643,
"step": 543
},
{
"epoch": 0.4459216281467535,
"grad_norm": 1.0345417083547934,
"learning_rate": 1.2214102654088248e-05,
"loss": 0.7032,
"step": 544
},
{
"epoch": 0.4467413370220233,
"grad_norm": 1.0815626325605556,
"learning_rate": 1.2188175907470847e-05,
"loss": 0.6893,
"step": 545
},
{
"epoch": 0.44756104589729306,
"grad_norm": 0.9083592191038633,
"learning_rate": 1.2162233703086024e-05,
"loss": 0.6871,
"step": 546
},
{
"epoch": 0.4483807547725628,
"grad_norm": 1.2528554901847724,
"learning_rate": 1.2136276224195349e-05,
"loss": 0.6891,
"step": 547
},
{
"epoch": 0.44920046364783256,
"grad_norm": 1.1952511452298105,
"learning_rate": 1.2110303654168305e-05,
"loss": 0.7172,
"step": 548
},
{
"epoch": 0.45002017252310234,
"grad_norm": 1.2956138729963762,
"learning_rate": 1.2084316176480972e-05,
"loss": 0.7096,
"step": 549
},
{
"epoch": 0.4508398813983721,
"grad_norm": 1.14468316880153,
"learning_rate": 1.2058313974714746e-05,
"loss": 0.7551,
"step": 550
},
{
"epoch": 0.4516595902736419,
"grad_norm": 1.0266099334680534,
"learning_rate": 1.2032297232555039e-05,
"loss": 0.6896,
"step": 551
},
{
"epoch": 0.45247929914891166,
"grad_norm": 1.090663886902925,
"learning_rate": 1.2006266133789975e-05,
"loss": 0.6972,
"step": 552
},
{
"epoch": 0.45329900802418144,
"grad_norm": 1.1406596892836078,
"learning_rate": 1.1980220862309097e-05,
"loss": 0.6963,
"step": 553
},
{
"epoch": 0.45411871689945116,
"grad_norm": 0.951359337153633,
"learning_rate": 1.1954161602102068e-05,
"loss": 0.7028,
"step": 554
},
{
"epoch": 0.45493842577472093,
"grad_norm": 0.9438613115171647,
"learning_rate": 1.1928088537257376e-05,
"loss": 0.6953,
"step": 555
},
{
"epoch": 0.4557581346499907,
"grad_norm": 1.2032290232783127,
"learning_rate": 1.1902001851961019e-05,
"loss": 0.7261,
"step": 556
},
{
"epoch": 0.4565778435252605,
"grad_norm": 1.1280373153297252,
"learning_rate": 1.1875901730495215e-05,
"loss": 0.7241,
"step": 557
},
{
"epoch": 0.45739755240053026,
"grad_norm": 1.0199931637594866,
"learning_rate": 1.1849788357237108e-05,
"loss": 0.689,
"step": 558
},
{
"epoch": 0.45821726127580004,
"grad_norm": 1.050127178800676,
"learning_rate": 1.1823661916657441e-05,
"loss": 0.737,
"step": 559
},
{
"epoch": 0.45903697015106976,
"grad_norm": 1.1639881702304666,
"learning_rate": 1.1797522593319277e-05,
"loss": 0.6752,
"step": 560
},
{
"epoch": 0.45985667902633953,
"grad_norm": 1.1225891456156893,
"learning_rate": 1.1771370571876681e-05,
"loss": 0.7389,
"step": 561
},
{
"epoch": 0.4606763879016093,
"grad_norm": 1.145675771789765,
"learning_rate": 1.1745206037073426e-05,
"loss": 0.6957,
"step": 562
},
{
"epoch": 0.4614960967768791,
"grad_norm": 0.9900975009183012,
"learning_rate": 1.1719029173741675e-05,
"loss": 0.73,
"step": 563
},
{
"epoch": 0.46231580565214886,
"grad_norm": 1.1179182042088796,
"learning_rate": 1.1692840166800688e-05,
"loss": 0.7477,
"step": 564
},
{
"epoch": 0.46313551452741863,
"grad_norm": 1.2353497325632112,
"learning_rate": 1.1666639201255507e-05,
"loss": 0.7212,
"step": 565
},
{
"epoch": 0.4639552234026884,
"grad_norm": 1.0998616345774277,
"learning_rate": 1.1640426462195654e-05,
"loss": 0.6798,
"step": 566
},
{
"epoch": 0.46477493227795813,
"grad_norm": 1.0255033503662383,
"learning_rate": 1.1614202134793823e-05,
"loss": 0.674,
"step": 567
},
{
"epoch": 0.4655946411532279,
"grad_norm": 1.1465881373881053,
"learning_rate": 1.1587966404304565e-05,
"loss": 0.6703,
"step": 568
},
{
"epoch": 0.4664143500284977,
"grad_norm": 1.0633908386224562,
"learning_rate": 1.1561719456062995e-05,
"loss": 0.7123,
"step": 569
},
{
"epoch": 0.46723405890376746,
"grad_norm": 1.0679978753304626,
"learning_rate": 1.1535461475483461e-05,
"loss": 0.7066,
"step": 570
},
{
"epoch": 0.46805376777903723,
"grad_norm": 0.9197821920130222,
"learning_rate": 1.150919264805825e-05,
"loss": 0.6983,
"step": 571
},
{
"epoch": 0.468873476654307,
"grad_norm": 0.9537906364854416,
"learning_rate": 1.1482913159356281e-05,
"loss": 0.7299,
"step": 572
},
{
"epoch": 0.4696931855295768,
"grad_norm": 1.0723672683236365,
"learning_rate": 1.1456623195021778e-05,
"loss": 0.6981,
"step": 573
},
{
"epoch": 0.4705128944048465,
"grad_norm": 1.100623221762658,
"learning_rate": 1.143032294077297e-05,
"loss": 0.7547,
"step": 574
},
{
"epoch": 0.4713326032801163,
"grad_norm": 0.8873104634369195,
"learning_rate": 1.1404012582400778e-05,
"loss": 0.6514,
"step": 575
},
{
"epoch": 0.47215231215538606,
"grad_norm": 0.9630009266493926,
"learning_rate": 1.1377692305767496e-05,
"loss": 0.717,
"step": 576
},
{
"epoch": 0.47297202103065583,
"grad_norm": 0.9227461757036858,
"learning_rate": 1.1351362296805487e-05,
"loss": 0.7323,
"step": 577
},
{
"epoch": 0.4737917299059256,
"grad_norm": 1.019737287468102,
"learning_rate": 1.1325022741515865e-05,
"loss": 0.7109,
"step": 578
},
{
"epoch": 0.4746114387811954,
"grad_norm": 1.1187383235686352,
"learning_rate": 1.1298673825967184e-05,
"loss": 0.7108,
"step": 579
},
{
"epoch": 0.47543114765646516,
"grad_norm": 1.126892262142564,
"learning_rate": 1.1272315736294108e-05,
"loss": 0.7599,
"step": 580
},
{
"epoch": 0.4762508565317349,
"grad_norm": 1.1174529718996196,
"learning_rate": 1.1245948658696127e-05,
"loss": 0.692,
"step": 581
},
{
"epoch": 0.47707056540700465,
"grad_norm": 1.0511738145306786,
"learning_rate": 1.1219572779436215e-05,
"loss": 0.6628,
"step": 582
},
{
"epoch": 0.47789027428227443,
"grad_norm": 1.124488559527458,
"learning_rate": 1.1193188284839518e-05,
"loss": 0.6837,
"step": 583
},
{
"epoch": 0.4787099831575442,
"grad_norm": 0.9036206888378797,
"learning_rate": 1.1166795361292055e-05,
"loss": 0.6873,
"step": 584
},
{
"epoch": 0.479529692032814,
"grad_norm": 1.08171561236448,
"learning_rate": 1.1140394195239376e-05,
"loss": 0.7091,
"step": 585
},
{
"epoch": 0.48034940090808376,
"grad_norm": 0.9866557061499747,
"learning_rate": 1.1113984973185268e-05,
"loss": 0.7049,
"step": 586
},
{
"epoch": 0.48116910978335353,
"grad_norm": 1.0361904367397607,
"learning_rate": 1.1087567881690422e-05,
"loss": 0.6661,
"step": 587
},
{
"epoch": 0.48198881865862325,
"grad_norm": 1.1157901054278567,
"learning_rate": 1.1061143107371122e-05,
"loss": 0.7439,
"step": 588
},
{
"epoch": 0.48280852753389303,
"grad_norm": 1.5040388215394276,
"learning_rate": 1.1034710836897922e-05,
"loss": 0.7327,
"step": 589
},
{
"epoch": 0.4836282364091628,
"grad_norm": 1.3390697244474046,
"learning_rate": 1.1008271256994338e-05,
"loss": 0.7548,
"step": 590
},
{
"epoch": 0.4844479452844326,
"grad_norm": 1.0147009628435193,
"learning_rate": 1.0981824554435518e-05,
"loss": 0.7145,
"step": 591
},
{
"epoch": 0.48526765415970236,
"grad_norm": 1.3136263563345296,
"learning_rate": 1.095537091604692e-05,
"loss": 0.7075,
"step": 592
},
{
"epoch": 0.48608736303497213,
"grad_norm": 1.2276064002237606,
"learning_rate": 1.0928910528703007e-05,
"loss": 0.6956,
"step": 593
},
{
"epoch": 0.48690707191024185,
"grad_norm": 1.0117999739470787,
"learning_rate": 1.0902443579325914e-05,
"loss": 0.6847,
"step": 594
},
{
"epoch": 0.4877267807855116,
"grad_norm": 1.0417431733710114,
"learning_rate": 1.087597025488413e-05,
"loss": 0.6982,
"step": 595
},
{
"epoch": 0.4885464896607814,
"grad_norm": 1.1273032439741224,
"learning_rate": 1.0849490742391184e-05,
"loss": 0.7231,
"step": 596
},
{
"epoch": 0.4893661985360512,
"grad_norm": 0.9423530224610465,
"learning_rate": 1.0823005228904315e-05,
"loss": 0.6419,
"step": 597
},
{
"epoch": 0.49018590741132095,
"grad_norm": 1.0787011367132358,
"learning_rate": 1.0796513901523156e-05,
"loss": 0.6857,
"step": 598
},
{
"epoch": 0.49100561628659073,
"grad_norm": 0.9826720981379837,
"learning_rate": 1.0770016947388407e-05,
"loss": 0.6954,
"step": 599
},
{
"epoch": 0.4918253251618605,
"grad_norm": 1.1728395885732517,
"learning_rate": 1.0743514553680521e-05,
"loss": 0.6937,
"step": 600
},
{
"epoch": 0.4926450340371302,
"grad_norm": 1.0469881022998453,
"learning_rate": 1.0717006907618377e-05,
"loss": 0.7001,
"step": 601
},
{
"epoch": 0.4934647429124,
"grad_norm": 1.0338506575284014,
"learning_rate": 1.0690494196457954e-05,
"loss": 0.6696,
"step": 602
},
{
"epoch": 0.4942844517876698,
"grad_norm": 0.931968678456202,
"learning_rate": 1.066397660749102e-05,
"loss": 0.7455,
"step": 603
},
{
"epoch": 0.49510416066293955,
"grad_norm": 0.9788740184125867,
"learning_rate": 1.0637454328043792e-05,
"loss": 0.7278,
"step": 604
},
{
"epoch": 0.4959238695382093,
"grad_norm": 1.0107018677975004,
"learning_rate": 1.0610927545475624e-05,
"loss": 0.6733,
"step": 605
},
{
"epoch": 0.4967435784134791,
"grad_norm": 0.9785269075804016,
"learning_rate": 1.0584396447177682e-05,
"loss": 0.6844,
"step": 606
},
{
"epoch": 0.4975632872887489,
"grad_norm": 1.4123480525557235,
"learning_rate": 1.0557861220571626e-05,
"loss": 0.7245,
"step": 607
},
{
"epoch": 0.4983829961640186,
"grad_norm": 1.0169986661751373,
"learning_rate": 1.0531322053108268e-05,
"loss": 0.7083,
"step": 608
},
{
"epoch": 0.4992027050392884,
"grad_norm": 1.0165623072548116,
"learning_rate": 1.050477913226626e-05,
"loss": 0.6881,
"step": 609
},
{
"epoch": 0.5000224139145582,
"grad_norm": 0.9765396371767978,
"learning_rate": 1.0478232645550784e-05,
"loss": 0.6874,
"step": 610
},
{
"epoch": 0.5008421227898279,
"grad_norm": 1.2056109097327445,
"learning_rate": 1.045168278049219e-05,
"loss": 0.662,
"step": 611
},
{
"epoch": 0.5016618316650977,
"grad_norm": 1.1334242779823047,
"learning_rate": 1.042512972464471e-05,
"loss": 0.75,
"step": 612
},
{
"epoch": 0.5024815405403674,
"grad_norm": 1.4577782954010892,
"learning_rate": 1.0398573665585105e-05,
"loss": 0.7535,
"step": 613
},
{
"epoch": 0.5033012494156373,
"grad_norm": 0.9060546888057004,
"learning_rate": 1.0372014790911365e-05,
"loss": 0.6872,
"step": 614
},
{
"epoch": 0.504120958290907,
"grad_norm": 1.0810384765147207,
"learning_rate": 1.0345453288241356e-05,
"loss": 0.7116,
"step": 615
},
{
"epoch": 0.5049406671661768,
"grad_norm": 0.999276369583175,
"learning_rate": 1.0318889345211519e-05,
"loss": 0.6634,
"step": 616
},
{
"epoch": 0.5057603760414465,
"grad_norm": 1.003285235817525,
"learning_rate": 1.0292323149475527e-05,
"loss": 0.7134,
"step": 617
},
{
"epoch": 0.5065800849167162,
"grad_norm": 1.068311941554304,
"learning_rate": 1.0265754888702972e-05,
"loss": 0.6644,
"step": 618
},
{
"epoch": 0.5073997937919861,
"grad_norm": 0.9176165791561176,
"learning_rate": 1.023918475057803e-05,
"loss": 0.7127,
"step": 619
},
{
"epoch": 0.5082195026672558,
"grad_norm": 1.0628172596775136,
"learning_rate": 1.0212612922798143e-05,
"loss": 0.6936,
"step": 620
},
{
"epoch": 0.5090392115425256,
"grad_norm": 1.1304424610837809,
"learning_rate": 1.0186039593072685e-05,
"loss": 0.72,
"step": 621
},
{
"epoch": 0.5098589204177953,
"grad_norm": 1.1201022968724756,
"learning_rate": 1.0159464949121642e-05,
"loss": 0.6772,
"step": 622
},
{
"epoch": 0.5106786292930652,
"grad_norm": 1.0988296304769767,
"learning_rate": 1.0132889178674283e-05,
"loss": 0.6663,
"step": 623
},
{
"epoch": 0.5114983381683349,
"grad_norm": 0.9153599329158012,
"learning_rate": 1.0106312469467841e-05,
"loss": 0.6759,
"step": 624
},
{
"epoch": 0.5123180470436046,
"grad_norm": 1.0183437832958995,
"learning_rate": 1.0079735009246168e-05,
"loss": 0.6931,
"step": 625
},
{
"epoch": 0.5131377559188744,
"grad_norm": 1.0881963927723373,
"learning_rate": 1.0053156985758435e-05,
"loss": 0.6805,
"step": 626
},
{
"epoch": 0.5139574647941442,
"grad_norm": 0.9708717120960165,
"learning_rate": 1.0026578586757778e-05,
"loss": 0.7037,
"step": 627
},
{
"epoch": 0.514777173669414,
"grad_norm": 1.0282492371108638,
"learning_rate": 1e-05,
"loss": 0.6831,
"step": 628
},
{
"epoch": 0.5155968825446837,
"grad_norm": 1.0518760957622026,
"learning_rate": 9.973421413242224e-06,
"loss": 0.6936,
"step": 629
},
{
"epoch": 0.5164165914199536,
"grad_norm": 0.9487845121780655,
"learning_rate": 9.946843014241572e-06,
"loss": 0.6654,
"step": 630
},
{
"epoch": 0.5172363002952233,
"grad_norm": 0.8593256158500326,
"learning_rate": 9.920264990753837e-06,
"loss": 0.7281,
"step": 631
},
{
"epoch": 0.518056009170493,
"grad_norm": 0.8237719636962215,
"learning_rate": 9.89368753053216e-06,
"loss": 0.7009,
"step": 632
},
{
"epoch": 0.5188757180457628,
"grad_norm": 1.0043839022182508,
"learning_rate": 9.867110821325717e-06,
"loss": 0.6837,
"step": 633
},
{
"epoch": 0.5196954269210325,
"grad_norm": 1.0275218242267354,
"learning_rate": 9.84053505087836e-06,
"loss": 0.7395,
"step": 634
},
{
"epoch": 0.5205151357963024,
"grad_norm": 0.9795606254636967,
"learning_rate": 9.813960406927318e-06,
"loss": 0.7423,
"step": 635
},
{
"epoch": 0.5213348446715721,
"grad_norm": 1.2622397994461323,
"learning_rate": 9.78738707720186e-06,
"loss": 0.6947,
"step": 636
},
{
"epoch": 0.5221545535468419,
"grad_norm": 1.0996490766157556,
"learning_rate": 9.760815249421973e-06,
"loss": 0.6751,
"step": 637
},
{
"epoch": 0.5229742624221116,
"grad_norm": 1.2231394645226992,
"learning_rate": 9.734245111297033e-06,
"loss": 0.7026,
"step": 638
},
{
"epoch": 0.5237939712973814,
"grad_norm": 1.7833524618893288,
"learning_rate": 9.707676850524473e-06,
"loss": 0.7126,
"step": 639
},
{
"epoch": 0.5246136801726512,
"grad_norm": 1.192628241875747,
"learning_rate": 9.681110654788483e-06,
"loss": 0.6929,
"step": 640
},
{
"epoch": 0.5254333890479209,
"grad_norm": 1.0330619310619844,
"learning_rate": 9.654546711758646e-06,
"loss": 0.6575,
"step": 641
},
{
"epoch": 0.5262530979231907,
"grad_norm": 0.9239513761063383,
"learning_rate": 9.627985209088639e-06,
"loss": 0.6711,
"step": 642
},
{
"epoch": 0.5270728067984605,
"grad_norm": 1.0144675307782336,
"learning_rate": 9.601426334414898e-06,
"loss": 0.7053,
"step": 643
},
{
"epoch": 0.5278925156737303,
"grad_norm": 1.2595984500957957,
"learning_rate": 9.574870275355295e-06,
"loss": 0.702,
"step": 644
},
{
"epoch": 0.528712224549,
"grad_norm": 1.08856803562517,
"learning_rate": 9.548317219507815e-06,
"loss": 0.6985,
"step": 645
},
{
"epoch": 0.5295319334242697,
"grad_norm": 0.9734725317105003,
"learning_rate": 9.521767354449218e-06,
"loss": 0.7079,
"step": 646
},
{
"epoch": 0.5303516422995396,
"grad_norm": 1.0247830003720153,
"learning_rate": 9.49522086773374e-06,
"loss": 0.713,
"step": 647
},
{
"epoch": 0.5311713511748093,
"grad_norm": 1.9146752209513695,
"learning_rate": 9.468677946891735e-06,
"loss": 0.6724,
"step": 648
},
{
"epoch": 0.5319910600500791,
"grad_norm": 1.1437028803071208,
"learning_rate": 9.442138779428376e-06,
"loss": 0.7193,
"step": 649
},
{
"epoch": 0.5328107689253488,
"grad_norm": 0.9951202847046366,
"learning_rate": 9.41560355282232e-06,
"loss": 0.704,
"step": 650
},
{
"epoch": 0.5336304778006187,
"grad_norm": 1.080157243599063,
"learning_rate": 9.389072454524381e-06,
"loss": 0.6917,
"step": 651
},
{
"epoch": 0.5344501866758884,
"grad_norm": 1.0096622733277012,
"learning_rate": 9.362545671956213e-06,
"loss": 0.7109,
"step": 652
},
{
"epoch": 0.5352698955511581,
"grad_norm": 1.234478638279633,
"learning_rate": 9.33602339250898e-06,
"loss": 0.6889,
"step": 653
},
{
"epoch": 0.536089604426428,
"grad_norm": 0.8987363282650567,
"learning_rate": 9.309505803542046e-06,
"loss": 0.6954,
"step": 654
},
{
"epoch": 0.5369093133016977,
"grad_norm": 1.132888864515375,
"learning_rate": 9.282993092381626e-06,
"loss": 0.7041,
"step": 655
},
{
"epoch": 0.5377290221769675,
"grad_norm": 1.0228655979819052,
"learning_rate": 9.256485446319482e-06,
"loss": 0.6761,
"step": 656
},
{
"epoch": 0.5385487310522372,
"grad_norm": 0.9674738807216214,
"learning_rate": 9.229983052611598e-06,
"loss": 0.7006,
"step": 657
},
{
"epoch": 0.539368439927507,
"grad_norm": 1.0090848905277268,
"learning_rate": 9.20348609847685e-06,
"loss": 0.6695,
"step": 658
},
{
"epoch": 0.5401881488027768,
"grad_norm": 1.0882159904816862,
"learning_rate": 9.176994771095687e-06,
"loss": 0.7085,
"step": 659
},
{
"epoch": 0.5410078576780465,
"grad_norm": 1.0653611932734905,
"learning_rate": 9.150509257608816e-06,
"loss": 0.6923,
"step": 660
},
{
"epoch": 0.5418275665533163,
"grad_norm": 0.9326086407344134,
"learning_rate": 9.124029745115872e-06,
"loss": 0.7002,
"step": 661
},
{
"epoch": 0.542647275428586,
"grad_norm": 1.049494692626326,
"learning_rate": 9.097556420674089e-06,
"loss": 0.6957,
"step": 662
},
{
"epoch": 0.5434669843038559,
"grad_norm": 0.8593314023373697,
"learning_rate": 9.071089471296996e-06,
"loss": 0.6946,
"step": 663
},
{
"epoch": 0.5442866931791256,
"grad_norm": 1.2057133873232884,
"learning_rate": 9.044629083953082e-06,
"loss": 0.7143,
"step": 664
},
{
"epoch": 0.5451064020543954,
"grad_norm": 0.9476407795254262,
"learning_rate": 9.018175445564485e-06,
"loss": 0.6822,
"step": 665
},
{
"epoch": 0.5459261109296651,
"grad_norm": 0.9875695610787744,
"learning_rate": 8.991728743005664e-06,
"loss": 0.6549,
"step": 666
},
{
"epoch": 0.5467458198049349,
"grad_norm": 1.0642714918626957,
"learning_rate": 8.96528916310208e-06,
"loss": 0.6739,
"step": 667
},
{
"epoch": 0.5475655286802047,
"grad_norm": 1.0879437506636844,
"learning_rate": 8.938856892628881e-06,
"loss": 0.7073,
"step": 668
},
{
"epoch": 0.5483852375554744,
"grad_norm": 1.0806928017917943,
"learning_rate": 8.912432118309581e-06,
"loss": 0.6954,
"step": 669
},
{
"epoch": 0.5492049464307442,
"grad_norm": 1.0229445641913895,
"learning_rate": 8.886015026814736e-06,
"loss": 0.6771,
"step": 670
},
{
"epoch": 0.550024655306014,
"grad_norm": 1.1054447572451949,
"learning_rate": 8.859605804760626e-06,
"loss": 0.7363,
"step": 671
},
{
"epoch": 0.5508443641812837,
"grad_norm": 0.898787200253909,
"learning_rate": 8.833204638707948e-06,
"loss": 0.6652,
"step": 672
},
{
"epoch": 0.5516640730565535,
"grad_norm": 0.9210649015472893,
"learning_rate": 8.806811715160485e-06,
"loss": 0.6662,
"step": 673
},
{
"epoch": 0.5524837819318232,
"grad_norm": 0.8767379258899219,
"learning_rate": 8.780427220563788e-06,
"loss": 0.676,
"step": 674
},
{
"epoch": 0.5533034908070931,
"grad_norm": 0.9673040084654836,
"learning_rate": 8.754051341303875e-06,
"loss": 0.6855,
"step": 675
},
{
"epoch": 0.5541231996823628,
"grad_norm": 1.240373695374397,
"learning_rate": 8.727684263705896e-06,
"loss": 0.6686,
"step": 676
},
{
"epoch": 0.5549429085576326,
"grad_norm": 1.001710109377683,
"learning_rate": 8.701326174032821e-06,
"loss": 0.667,
"step": 677
},
{
"epoch": 0.5557626174329023,
"grad_norm": 1.1373894634495587,
"learning_rate": 8.674977258484136e-06,
"loss": 0.6785,
"step": 678
},
{
"epoch": 0.5565823263081721,
"grad_norm": 0.9875809033934482,
"learning_rate": 8.648637703194515e-06,
"loss": 0.6631,
"step": 679
},
{
"epoch": 0.5574020351834419,
"grad_norm": 0.992000234953506,
"learning_rate": 8.622307694232508e-06,
"loss": 0.6815,
"step": 680
},
{
"epoch": 0.5582217440587116,
"grad_norm": 1.1239235374517815,
"learning_rate": 8.595987417599225e-06,
"loss": 0.6877,
"step": 681
},
{
"epoch": 0.5590414529339814,
"grad_norm": 1.0154213881641487,
"learning_rate": 8.569677059227033e-06,
"loss": 0.6764,
"step": 682
},
{
"epoch": 0.5598611618092512,
"grad_norm": 0.8581221132256242,
"learning_rate": 8.543376804978223e-06,
"loss": 0.6478,
"step": 683
},
{
"epoch": 0.560680870684521,
"grad_norm": 0.9484056664865104,
"learning_rate": 8.517086840643722e-06,
"loss": 0.6942,
"step": 684
},
{
"epoch": 0.5615005795597907,
"grad_norm": 0.9386502452035249,
"learning_rate": 8.490807351941753e-06,
"loss": 0.7249,
"step": 685
},
{
"epoch": 0.5623202884350604,
"grad_norm": 1.0918869846767094,
"learning_rate": 8.464538524516546e-06,
"loss": 0.7323,
"step": 686
},
{
"epoch": 0.5631399973103303,
"grad_norm": 0.9628198052048128,
"learning_rate": 8.438280543937012e-06,
"loss": 0.6916,
"step": 687
},
{
"epoch": 0.5639597061856,
"grad_norm": 0.911943367219312,
"learning_rate": 8.412033595695436e-06,
"loss": 0.6731,
"step": 688
},
{
"epoch": 0.5647794150608698,
"grad_norm": 1.0799106548105355,
"learning_rate": 8.385797865206178e-06,
"loss": 0.7008,
"step": 689
},
{
"epoch": 0.5655991239361395,
"grad_norm": 0.8773939343725733,
"learning_rate": 8.359573537804347e-06,
"loss": 0.6633,
"step": 690
},
{
"epoch": 0.5664188328114094,
"grad_norm": 1.0911977353802045,
"learning_rate": 8.333360798744496e-06,
"loss": 0.6728,
"step": 691
},
{
"epoch": 0.5672385416866791,
"grad_norm": 0.9142826927820591,
"learning_rate": 8.307159833199317e-06,
"loss": 0.7008,
"step": 692
},
{
"epoch": 0.5680582505619488,
"grad_norm": 1.0742436019486135,
"learning_rate": 8.28097082625833e-06,
"loss": 0.699,
"step": 693
},
{
"epoch": 0.5688779594372186,
"grad_norm": 1.0674381195589975,
"learning_rate": 8.254793962926579e-06,
"loss": 0.6863,
"step": 694
},
{
"epoch": 0.5696976683124884,
"grad_norm": 1.024423131903941,
"learning_rate": 8.228629428123319e-06,
"loss": 0.726,
"step": 695
},
{
"epoch": 0.5705173771877582,
"grad_norm": 1.4902271117546952,
"learning_rate": 8.202477406680725e-06,
"loss": 0.6668,
"step": 696
},
{
"epoch": 0.5713370860630279,
"grad_norm": 1.0668800671622416,
"learning_rate": 8.17633808334256e-06,
"loss": 0.7002,
"step": 697
},
{
"epoch": 0.5721567949382977,
"grad_norm": 1.1261776653087388,
"learning_rate": 8.150211642762895e-06,
"loss": 0.6763,
"step": 698
},
{
"epoch": 0.5729765038135675,
"grad_norm": 0.9310930024282916,
"learning_rate": 8.124098269504786e-06,
"loss": 0.6865,
"step": 699
},
{
"epoch": 0.5737962126888372,
"grad_norm": 0.9492696082307697,
"learning_rate": 8.097998148038986e-06,
"loss": 0.6843,
"step": 700
},
{
"epoch": 0.574615921564107,
"grad_norm": 1.0701249639752783,
"learning_rate": 8.07191146274263e-06,
"loss": 0.6754,
"step": 701
},
{
"epoch": 0.5754356304393767,
"grad_norm": 1.1438199720321645,
"learning_rate": 8.045838397897932e-06,
"loss": 0.738,
"step": 702
},
{
"epoch": 0.5762553393146466,
"grad_norm": 1.0398125273091405,
"learning_rate": 8.019779137690906e-06,
"loss": 0.6892,
"step": 703
},
{
"epoch": 0.5770750481899163,
"grad_norm": 1.0276513920942751,
"learning_rate": 7.993733866210029e-06,
"loss": 0.6894,
"step": 704
},
{
"epoch": 0.5778947570651861,
"grad_norm": 0.9324102551034779,
"learning_rate": 7.967702767444963e-06,
"loss": 0.7013,
"step": 705
},
{
"epoch": 0.5787144659404558,
"grad_norm": 1.3355239761097153,
"learning_rate": 7.941686025285257e-06,
"loss": 0.6956,
"step": 706
},
{
"epoch": 0.5795341748157256,
"grad_norm": 1.1022923517020726,
"learning_rate": 7.915683823519031e-06,
"loss": 0.6774,
"step": 707
},
{
"epoch": 0.5803538836909954,
"grad_norm": 1.1951237712356495,
"learning_rate": 7.889696345831696e-06,
"loss": 0.7127,
"step": 708
},
{
"epoch": 0.5811735925662651,
"grad_norm": 0.9939980131028814,
"learning_rate": 7.863723775804651e-06,
"loss": 0.7227,
"step": 709
},
{
"epoch": 0.5819933014415349,
"grad_norm": 0.983910972463096,
"learning_rate": 7.83776629691398e-06,
"loss": 0.7139,
"step": 710
},
{
"epoch": 0.5828130103168047,
"grad_norm": 0.9968046926785361,
"learning_rate": 7.811824092529155e-06,
"loss": 0.695,
"step": 711
},
{
"epoch": 0.5836327191920745,
"grad_norm": 1.3112639396219898,
"learning_rate": 7.785897345911755e-06,
"loss": 0.7067,
"step": 712
},
{
"epoch": 0.5844524280673442,
"grad_norm": 1.0513245518332954,
"learning_rate": 7.759986240214155e-06,
"loss": 0.6881,
"step": 713
},
{
"epoch": 0.5852721369426139,
"grad_norm": 1.0353161019970707,
"learning_rate": 7.734090958478241e-06,
"loss": 0.7268,
"step": 714
},
{
"epoch": 0.5860918458178838,
"grad_norm": 0.9607479332004892,
"learning_rate": 7.708211683634112e-06,
"loss": 0.6684,
"step": 715
},
{
"epoch": 0.5869115546931535,
"grad_norm": 0.9427199040007197,
"learning_rate": 7.6823485984988e-06,
"loss": 0.6604,
"step": 716
},
{
"epoch": 0.5877312635684233,
"grad_norm": 0.9582958403279314,
"learning_rate": 7.656501885774963e-06,
"loss": 0.6835,
"step": 717
},
{
"epoch": 0.588550972443693,
"grad_norm": 0.9513029667741213,
"learning_rate": 7.6306717280496e-06,
"loss": 0.6985,
"step": 718
},
{
"epoch": 0.5893706813189629,
"grad_norm": 0.9149548095937733,
"learning_rate": 7.604858307792759e-06,
"loss": 0.6622,
"step": 719
},
{
"epoch": 0.5901903901942326,
"grad_norm": 1.1614419157754394,
"learning_rate": 7.579061807356256e-06,
"loss": 0.6525,
"step": 720
},
{
"epoch": 0.5910100990695023,
"grad_norm": 0.9316502143112609,
"learning_rate": 7.553282408972382e-06,
"loss": 0.6918,
"step": 721
},
{
"epoch": 0.5918298079447721,
"grad_norm": 1.211756806292709,
"learning_rate": 7.527520294752607e-06,
"loss": 0.7328,
"step": 722
},
{
"epoch": 0.5926495168200419,
"grad_norm": 1.0041507584009146,
"learning_rate": 7.501775646686316e-06,
"loss": 0.7105,
"step": 723
},
{
"epoch": 0.5934692256953117,
"grad_norm": 1.0080843057993603,
"learning_rate": 7.476048646639498e-06,
"loss": 0.6924,
"step": 724
},
{
"epoch": 0.5942889345705814,
"grad_norm": 1.0684532031686043,
"learning_rate": 7.450339476353475e-06,
"loss": 0.6968,
"step": 725
},
{
"epoch": 0.5951086434458512,
"grad_norm": 0.8548563797005055,
"learning_rate": 7.424648317443616e-06,
"loss": 0.6786,
"step": 726
},
{
"epoch": 0.595928352321121,
"grad_norm": 0.9140628981409149,
"learning_rate": 7.398975351398054e-06,
"loss": 0.6497,
"step": 727
},
{
"epoch": 0.5967480611963907,
"grad_norm": 1.0193177196185355,
"learning_rate": 7.3733207595764015e-06,
"loss": 0.6599,
"step": 728
},
{
"epoch": 0.5975677700716605,
"grad_norm": 0.9739217343448399,
"learning_rate": 7.347684723208471e-06,
"loss": 0.6733,
"step": 729
},
{
"epoch": 0.5983874789469302,
"grad_norm": 0.9970314899912143,
"learning_rate": 7.322067423393002e-06,
"loss": 0.6831,
"step": 730
},
{
"epoch": 0.5992071878222001,
"grad_norm": 1.0388363452219995,
"learning_rate": 7.2964690410963705e-06,
"loss": 0.6992,
"step": 731
},
{
"epoch": 0.6000268966974698,
"grad_norm": 0.9131563803785976,
"learning_rate": 7.2708897571513116e-06,
"loss": 0.6728,
"step": 732
},
{
"epoch": 0.6008466055727396,
"grad_norm": 0.9017740729080634,
"learning_rate": 7.2453297522556475e-06,
"loss": 0.6988,
"step": 733
},
{
"epoch": 0.6016663144480093,
"grad_norm": 1.1335048481538907,
"learning_rate": 7.2197892069710105e-06,
"loss": 0.7225,
"step": 734
},
{
"epoch": 0.602486023323279,
"grad_norm": 0.9210798811250238,
"learning_rate": 7.194268301721563e-06,
"loss": 0.6785,
"step": 735
},
{
"epoch": 0.6033057321985489,
"grad_norm": 1.143878864251672,
"learning_rate": 7.168767216792725e-06,
"loss": 0.6702,
"step": 736
},
{
"epoch": 0.6041254410738186,
"grad_norm": 0.9773102242856955,
"learning_rate": 7.143286132329912e-06,
"loss": 0.6843,
"step": 737
},
{
"epoch": 0.6049451499490884,
"grad_norm": 1.228200242480035,
"learning_rate": 7.117825228337236e-06,
"loss": 0.6886,
"step": 738
},
{
"epoch": 0.6057648588243582,
"grad_norm": 0.9596065725527181,
"learning_rate": 7.092384684676263e-06,
"loss": 0.6708,
"step": 739
},
{
"epoch": 0.6065845676996279,
"grad_norm": 1.0151810760870799,
"learning_rate": 7.066964681064716e-06,
"loss": 0.7387,
"step": 740
},
{
"epoch": 0.6074042765748977,
"grad_norm": 0.8997936926440689,
"learning_rate": 7.041565397075232e-06,
"loss": 0.6771,
"step": 741
},
{
"epoch": 0.6082239854501674,
"grad_norm": 1.3944158171998005,
"learning_rate": 7.016187012134069e-06,
"loss": 0.6377,
"step": 742
},
{
"epoch": 0.6090436943254373,
"grad_norm": 0.9839385786513499,
"learning_rate": 6.990829705519853e-06,
"loss": 0.684,
"step": 743
},
{
"epoch": 0.609863403200707,
"grad_norm": 0.9329448014775885,
"learning_rate": 6.965493656362314e-06,
"loss": 0.6824,
"step": 744
},
{
"epoch": 0.6106831120759768,
"grad_norm": 1.2995982052172474,
"learning_rate": 6.940179043641005e-06,
"loss": 0.6434,
"step": 745
},
{
"epoch": 0.6115028209512465,
"grad_norm": 1.0534163129670344,
"learning_rate": 6.914886046184055e-06,
"loss": 0.6492,
"step": 746
},
{
"epoch": 0.6123225298265162,
"grad_norm": 1.293855173902471,
"learning_rate": 6.889614842666892e-06,
"loss": 0.6489,
"step": 747
},
{
"epoch": 0.6131422387017861,
"grad_norm": 1.0874547219945876,
"learning_rate": 6.86436561161099e-06,
"loss": 0.6646,
"step": 748
},
{
"epoch": 0.6139619475770558,
"grad_norm": 0.9378126974707442,
"learning_rate": 6.8391385313826035e-06,
"loss": 0.6635,
"step": 749
},
{
"epoch": 0.6147816564523256,
"grad_norm": 0.9426661175100548,
"learning_rate": 6.813933780191504e-06,
"loss": 0.6648,
"step": 750
},
{
"epoch": 0.6156013653275954,
"grad_norm": 0.9856066248897091,
"learning_rate": 6.788751536089739e-06,
"loss": 0.6709,
"step": 751
},
{
"epoch": 0.6164210742028652,
"grad_norm": 1.0012369829940488,
"learning_rate": 6.763591976970347e-06,
"loss": 0.6691,
"step": 752
},
{
"epoch": 0.6172407830781349,
"grad_norm": 0.9959105513704459,
"learning_rate": 6.738455280566124e-06,
"loss": 0.6527,
"step": 753
},
{
"epoch": 0.6180604919534046,
"grad_norm": 0.8990460955435333,
"learning_rate": 6.713341624448351e-06,
"loss": 0.6863,
"step": 754
},
{
"epoch": 0.6188802008286745,
"grad_norm": 1.0797557639687205,
"learning_rate": 6.6882511860255505e-06,
"loss": 0.6829,
"step": 755
},
{
"epoch": 0.6196999097039442,
"grad_norm": 1.0981200859249032,
"learning_rate": 6.663184142542233e-06,
"loss": 0.7087,
"step": 756
},
{
"epoch": 0.620519618579214,
"grad_norm": 0.895548029637809,
"learning_rate": 6.638140671077633e-06,
"loss": 0.7186,
"step": 757
},
{
"epoch": 0.6213393274544837,
"grad_norm": 0.9129227581083621,
"learning_rate": 6.613120948544482e-06,
"loss": 0.6608,
"step": 758
},
{
"epoch": 0.6221590363297536,
"grad_norm": 1.088980184944204,
"learning_rate": 6.58812515168773e-06,
"loss": 0.702,
"step": 759
},
{
"epoch": 0.6229787452050233,
"grad_norm": 1.014738441984619,
"learning_rate": 6.563153457083315e-06,
"loss": 0.6525,
"step": 760
},
{
"epoch": 0.623798454080293,
"grad_norm": 1.2491143694173545,
"learning_rate": 6.538206041136915e-06,
"loss": 0.6114,
"step": 761
},
{
"epoch": 0.6246181629555628,
"grad_norm": 0.9450941313470351,
"learning_rate": 6.513283080082692e-06,
"loss": 0.6798,
"step": 762
},
{
"epoch": 0.6254378718308325,
"grad_norm": 0.9826901970277533,
"learning_rate": 6.488384749982054e-06,
"loss": 0.692,
"step": 763
},
{
"epoch": 0.6262575807061024,
"grad_norm": 1.2727397700829735,
"learning_rate": 6.463511226722416e-06,
"loss": 0.7307,
"step": 764
},
{
"epoch": 0.6270772895813721,
"grad_norm": 1.1266809817446504,
"learning_rate": 6.438662686015947e-06,
"loss": 0.7133,
"step": 765
},
{
"epoch": 0.6278969984566419,
"grad_norm": 1.0226392944373364,
"learning_rate": 6.413839303398337e-06,
"loss": 0.6466,
"step": 766
},
{
"epoch": 0.6287167073319117,
"grad_norm": 0.9289006036277805,
"learning_rate": 6.389041254227548e-06,
"loss": 0.663,
"step": 767
},
{
"epoch": 0.6295364162071814,
"grad_norm": 1.1109549707701603,
"learning_rate": 6.364268713682586e-06,
"loss": 0.7394,
"step": 768
},
{
"epoch": 0.6303561250824512,
"grad_norm": 1.2070927890557621,
"learning_rate": 6.339521856762254e-06,
"loss": 0.7084,
"step": 769
},
{
"epoch": 0.6311758339577209,
"grad_norm": 0.9390459015190175,
"learning_rate": 6.3148008582839284e-06,
"loss": 0.6472,
"step": 770
},
{
"epoch": 0.6319955428329908,
"grad_norm": 1.2324966049328243,
"learning_rate": 6.290105892882303e-06,
"loss": 0.6863,
"step": 771
},
{
"epoch": 0.6328152517082605,
"grad_norm": 0.9768240592532421,
"learning_rate": 6.265437135008181e-06,
"loss": 0.7051,
"step": 772
},
{
"epoch": 0.6336349605835303,
"grad_norm": 1.0530278624197773,
"learning_rate": 6.240794758927222e-06,
"loss": 0.6997,
"step": 773
},
{
"epoch": 0.6344546694588,
"grad_norm": 1.0131544645719084,
"learning_rate": 6.21617893871872e-06,
"loss": 0.6647,
"step": 774
},
{
"epoch": 0.6352743783340697,
"grad_norm": 1.0096440337266168,
"learning_rate": 6.191589848274369e-06,
"loss": 0.6491,
"step": 775
},
{
"epoch": 0.6360940872093396,
"grad_norm": 0.934033869895687,
"learning_rate": 6.167027661297049e-06,
"loss": 0.6522,
"step": 776
},
{
"epoch": 0.6369137960846093,
"grad_norm": 0.8898612318177457,
"learning_rate": 6.142492551299577e-06,
"loss": 0.6902,
"step": 777
},
{
"epoch": 0.6377335049598791,
"grad_norm": 0.8433415587571697,
"learning_rate": 6.117984691603493e-06,
"loss": 0.6715,
"step": 778
},
{
"epoch": 0.6385532138351488,
"grad_norm": 1.0765449584649374,
"learning_rate": 6.093504255337844e-06,
"loss": 0.6875,
"step": 779
},
{
"epoch": 0.6393729227104187,
"grad_norm": 1.117251979209498,
"learning_rate": 6.069051415437941e-06,
"loss": 0.6749,
"step": 780
},
{
"epoch": 0.6401926315856884,
"grad_norm": 0.906725823705208,
"learning_rate": 6.044626344644151e-06,
"loss": 0.6637,
"step": 781
},
{
"epoch": 0.6410123404609581,
"grad_norm": 0.9375437570439383,
"learning_rate": 6.020229215500677e-06,
"loss": 0.6836,
"step": 782
},
{
"epoch": 0.641832049336228,
"grad_norm": 0.995656349774453,
"learning_rate": 5.995860200354335e-06,
"loss": 0.6574,
"step": 783
},
{
"epoch": 0.6426517582114977,
"grad_norm": 1.1507271693887398,
"learning_rate": 5.971519471353335e-06,
"loss": 0.7091,
"step": 784
},
{
"epoch": 0.6434714670867675,
"grad_norm": 1.0155394628074705,
"learning_rate": 5.9472072004460665e-06,
"loss": 0.7005,
"step": 785
},
{
"epoch": 0.6442911759620372,
"grad_norm": 0.9676791996881187,
"learning_rate": 5.922923559379892e-06,
"loss": 0.6626,
"step": 786
},
{
"epoch": 0.645110884837307,
"grad_norm": 0.9947218278373859,
"learning_rate": 5.898668719699914e-06,
"loss": 0.6963,
"step": 787
},
{
"epoch": 0.6459305937125768,
"grad_norm": 1.0470661134698638,
"learning_rate": 5.874442852747788e-06,
"loss": 0.6926,
"step": 788
},
{
"epoch": 0.6467503025878465,
"grad_norm": 1.023967713835746,
"learning_rate": 5.850246129660494e-06,
"loss": 0.6798,
"step": 789
},
{
"epoch": 0.6475700114631163,
"grad_norm": 0.8525684574020227,
"learning_rate": 5.82607872136913e-06,
"loss": 0.6541,
"step": 790
},
{
"epoch": 0.648389720338386,
"grad_norm": 0.8552419741701156,
"learning_rate": 5.801940798597716e-06,
"loss": 0.6653,
"step": 791
},
{
"epoch": 0.6492094292136559,
"grad_norm": 0.985966522679896,
"learning_rate": 5.7778325318619665e-06,
"loss": 0.6979,
"step": 792
},
{
"epoch": 0.6500291380889256,
"grad_norm": 1.071421825526807,
"learning_rate": 5.753754091468116e-06,
"loss": 0.6875,
"step": 793
},
{
"epoch": 0.6508488469641954,
"grad_norm": 0.8292940677618507,
"learning_rate": 5.729705647511688e-06,
"loss": 0.6539,
"step": 794
},
{
"epoch": 0.6516685558394651,
"grad_norm": 0.9237781449160415,
"learning_rate": 5.7056873698763035e-06,
"loss": 0.6229,
"step": 795
},
{
"epoch": 0.6524882647147349,
"grad_norm": 0.8520241093407978,
"learning_rate": 5.681699428232488e-06,
"loss": 0.6786,
"step": 796
},
{
"epoch": 0.6533079735900047,
"grad_norm": 1.3173133150824567,
"learning_rate": 5.657741992036463e-06,
"loss": 0.6815,
"step": 797
},
{
"epoch": 0.6541276824652744,
"grad_norm": 0.9772163882515668,
"learning_rate": 5.633815230528954e-06,
"loss": 0.6412,
"step": 798
},
{
"epoch": 0.6549473913405442,
"grad_norm": 0.7825867263400921,
"learning_rate": 5.609919312733987e-06,
"loss": 0.6588,
"step": 799
},
{
"epoch": 0.655767100215814,
"grad_norm": 1.0207171520366223,
"learning_rate": 5.586054407457717e-06,
"loss": 0.7252,
"step": 800
},
{
"epoch": 0.6565868090910837,
"grad_norm": 0.9476473610823085,
"learning_rate": 5.562220683287205e-06,
"loss": 0.7107,
"step": 801
},
{
"epoch": 0.6574065179663535,
"grad_norm": 1.1414861248509818,
"learning_rate": 5.538418308589246e-06,
"loss": 0.683,
"step": 802
},
{
"epoch": 0.6582262268416232,
"grad_norm": 0.9302141594072603,
"learning_rate": 5.514647451509175e-06,
"loss": 0.6995,
"step": 803
},
{
"epoch": 0.6590459357168931,
"grad_norm": 0.8439935910633803,
"learning_rate": 5.490908279969682e-06,
"loss": 0.6625,
"step": 804
},
{
"epoch": 0.6598656445921628,
"grad_norm": 1.176388521809215,
"learning_rate": 5.467200961669619e-06,
"loss": 0.6489,
"step": 805
},
{
"epoch": 0.6606853534674326,
"grad_norm": 1.019297078499716,
"learning_rate": 5.443525664082818e-06,
"loss": 0.6554,
"step": 806
},
{
"epoch": 0.6615050623427023,
"grad_norm": 0.8702134395960319,
"learning_rate": 5.419882554456923e-06,
"loss": 0.6468,
"step": 807
},
{
"epoch": 0.6623247712179721,
"grad_norm": 0.9766741983468309,
"learning_rate": 5.39627179981218e-06,
"loss": 0.6574,
"step": 808
},
{
"epoch": 0.6631444800932419,
"grad_norm": 0.8649051577871882,
"learning_rate": 5.372693566940277e-06,
"loss": 0.7409,
"step": 809
},
{
"epoch": 0.6639641889685116,
"grad_norm": 1.4853528209007765,
"learning_rate": 5.349148022403162e-06,
"loss": 0.6336,
"step": 810
},
{
"epoch": 0.6647838978437814,
"grad_norm": 1.0541085370053918,
"learning_rate": 5.325635332531864e-06,
"loss": 0.6805,
"step": 811
},
{
"epoch": 0.6656036067190512,
"grad_norm": 1.2649909499699947,
"learning_rate": 5.302155663425319e-06,
"loss": 0.6991,
"step": 812
},
{
"epoch": 0.666423315594321,
"grad_norm": 1.063516993146653,
"learning_rate": 5.278709180949195e-06,
"loss": 0.6947,
"step": 813
},
{
"epoch": 0.6672430244695907,
"grad_norm": 1.0501654791658581,
"learning_rate": 5.255296050734733e-06,
"loss": 0.6519,
"step": 814
},
{
"epoch": 0.6680627333448604,
"grad_norm": 1.0676792868997675,
"learning_rate": 5.231916438177552e-06,
"loss": 0.6393,
"step": 815
},
{
"epoch": 0.6688824422201303,
"grad_norm": 0.973480173895216,
"learning_rate": 5.208570508436501e-06,
"loss": 0.692,
"step": 816
},
{
"epoch": 0.6697021510954,
"grad_norm": 0.9197382995498412,
"learning_rate": 5.1852584264324864e-06,
"loss": 0.6368,
"step": 817
},
{
"epoch": 0.6705218599706698,
"grad_norm": 0.8867930075260909,
"learning_rate": 5.161980356847307e-06,
"loss": 0.673,
"step": 818
},
{
"epoch": 0.6713415688459395,
"grad_norm": 0.8532905570544348,
"learning_rate": 5.138736464122485e-06,
"loss": 0.661,
"step": 819
},
{
"epoch": 0.6721612777212094,
"grad_norm": 0.9559861144767411,
"learning_rate": 5.115526912458113e-06,
"loss": 0.6772,
"step": 820
},
{
"epoch": 0.6729809865964791,
"grad_norm": 1.0069055402799345,
"learning_rate": 5.092351865811698e-06,
"loss": 0.692,
"step": 821
},
{
"epoch": 0.6738006954717488,
"grad_norm": 1.001091881118724,
"learning_rate": 5.0692114878969845e-06,
"loss": 0.6809,
"step": 822
},
{
"epoch": 0.6746204043470186,
"grad_norm": 0.9372084095329728,
"learning_rate": 5.046105942182815e-06,
"loss": 0.6828,
"step": 823
},
{
"epoch": 0.6754401132222884,
"grad_norm": 0.9579692319752076,
"learning_rate": 5.023035391891966e-06,
"loss": 0.7144,
"step": 824
},
{
"epoch": 0.6762598220975582,
"grad_norm": 0.9366214444691355,
"learning_rate": 5.000000000000003e-06,
"loss": 0.6756,
"step": 825
},
{
"epoch": 0.6770795309728279,
"grad_norm": 0.9046731654434792,
"learning_rate": 4.97699992923412e-06,
"loss": 0.6796,
"step": 826
},
{
"epoch": 0.6778992398480977,
"grad_norm": 0.9846386504206198,
"learning_rate": 4.954035342071994e-06,
"loss": 0.6995,
"step": 827
},
{
"epoch": 0.6787189487233675,
"grad_norm": 0.9725971365610542,
"learning_rate": 4.9311064007406494e-06,
"loss": 0.6993,
"step": 828
},
{
"epoch": 0.6795386575986372,
"grad_norm": 0.84552367685698,
"learning_rate": 4.908213267215287e-06,
"loss": 0.6636,
"step": 829
},
{
"epoch": 0.680358366473907,
"grad_norm": 0.8433804293636122,
"learning_rate": 4.885356103218159e-06,
"loss": 0.663,
"step": 830
},
{
"epoch": 0.6811780753491767,
"grad_norm": 1.001372337248185,
"learning_rate": 4.862535070217416e-06,
"loss": 0.6736,
"step": 831
},
{
"epoch": 0.6819977842244466,
"grad_norm": 1.0271257213532776,
"learning_rate": 4.839750329425979e-06,
"loss": 0.6933,
"step": 832
},
{
"epoch": 0.6828174930997163,
"grad_norm": 0.905414757851656,
"learning_rate": 4.8170020418003884e-06,
"loss": 0.7005,
"step": 833
},
{
"epoch": 0.6836372019749861,
"grad_norm": 0.8876300144932402,
"learning_rate": 4.794290368039664e-06,
"loss": 0.6903,
"step": 834
},
{
"epoch": 0.6844569108502558,
"grad_norm": 0.8447852583736626,
"learning_rate": 4.771615468584194e-06,
"loss": 0.6689,
"step": 835
},
{
"epoch": 0.6852766197255256,
"grad_norm": 1.1570672171860963,
"learning_rate": 4.7489775036145715e-06,
"loss": 0.6815,
"step": 836
},
{
"epoch": 0.6860963286007954,
"grad_norm": 1.056935486556543,
"learning_rate": 4.726376633050479e-06,
"loss": 0.6837,
"step": 837
},
{
"epoch": 0.6869160374760651,
"grad_norm": 0.902045492313076,
"learning_rate": 4.703813016549555e-06,
"loss": 0.6732,
"step": 838
},
{
"epoch": 0.6877357463513349,
"grad_norm": 0.8536604097815113,
"learning_rate": 4.681286813506271e-06,
"loss": 0.6633,
"step": 839
},
{
"epoch": 0.6885554552266047,
"grad_norm": 0.9483276482486834,
"learning_rate": 4.658798183050796e-06,
"loss": 0.702,
"step": 840
},
{
"epoch": 0.6893751641018745,
"grad_norm": 0.8931899793543349,
"learning_rate": 4.636347284047878e-06,
"loss": 0.6867,
"step": 841
},
{
"epoch": 0.6901948729771442,
"grad_norm": 0.917909820854939,
"learning_rate": 4.613934275095729e-06,
"loss": 0.6675,
"step": 842
},
{
"epoch": 0.6910145818524139,
"grad_norm": 0.9699408102330255,
"learning_rate": 4.5915593145248926e-06,
"loss": 0.6723,
"step": 843
},
{
"epoch": 0.6918342907276838,
"grad_norm": 1.0098755391591705,
"learning_rate": 4.569222560397126e-06,
"loss": 0.6911,
"step": 844
},
{
"epoch": 0.6926539996029535,
"grad_norm": 0.9835264345278736,
"learning_rate": 4.546924170504292e-06,
"loss": 0.6735,
"step": 845
},
{
"epoch": 0.6934737084782233,
"grad_norm": 0.9271232973484277,
"learning_rate": 4.524664302367239e-06,
"loss": 0.6576,
"step": 846
},
{
"epoch": 0.694293417353493,
"grad_norm": 0.9451374813106167,
"learning_rate": 4.502443113234688e-06,
"loss": 0.6743,
"step": 847
},
{
"epoch": 0.6951131262287629,
"grad_norm": 0.9249293079901136,
"learning_rate": 4.48026076008212e-06,
"loss": 0.6395,
"step": 848
},
{
"epoch": 0.6959328351040326,
"grad_norm": 0.9639127165317318,
"learning_rate": 4.458117399610682e-06,
"loss": 0.6593,
"step": 849
},
{
"epoch": 0.6967525439793023,
"grad_norm": 1.022789321901033,
"learning_rate": 4.436013188246056e-06,
"loss": 0.6555,
"step": 850
},
{
"epoch": 0.6975722528545721,
"grad_norm": 1.012946754243605,
"learning_rate": 4.4139482821373665e-06,
"loss": 0.6819,
"step": 851
},
{
"epoch": 0.6983919617298419,
"grad_norm": 1.0703301422327582,
"learning_rate": 4.391922837156082e-06,
"loss": 0.6727,
"step": 852
},
{
"epoch": 0.6992116706051117,
"grad_norm": 1.0470926524880477,
"learning_rate": 4.369937008894906e-06,
"loss": 0.6924,
"step": 853
},
{
"epoch": 0.7000313794803814,
"grad_norm": 0.9632460097564534,
"learning_rate": 4.3479909526666775e-06,
"loss": 0.6782,
"step": 854
},
{
"epoch": 0.7008510883556512,
"grad_norm": 0.8882440552764685,
"learning_rate": 4.326084823503287e-06,
"loss": 0.6406,
"step": 855
},
{
"epoch": 0.701670797230921,
"grad_norm": 0.9658184565213099,
"learning_rate": 4.30421877615456e-06,
"loss": 0.6765,
"step": 856
},
{
"epoch": 0.7024905061061907,
"grad_norm": 0.9096741636391663,
"learning_rate": 4.282392965087182e-06,
"loss": 0.6663,
"step": 857
},
{
"epoch": 0.7033102149814605,
"grad_norm": 0.888341707914917,
"learning_rate": 4.260607544483596e-06,
"loss": 0.677,
"step": 858
},
{
"epoch": 0.7041299238567302,
"grad_norm": 1.0483459336696939,
"learning_rate": 4.23886266824092e-06,
"loss": 0.6978,
"step": 859
},
{
"epoch": 0.7049496327320001,
"grad_norm": 0.9646417906985328,
"learning_rate": 4.217158489969857e-06,
"loss": 0.672,
"step": 860
},
{
"epoch": 0.7057693416072698,
"grad_norm": 0.9710201960324607,
"learning_rate": 4.195495162993607e-06,
"loss": 0.719,
"step": 861
},
{
"epoch": 0.7065890504825396,
"grad_norm": 1.0163465618393006,
"learning_rate": 4.173872840346798e-06,
"loss": 0.6923,
"step": 862
},
{
"epoch": 0.7074087593578093,
"grad_norm": 0.8472860337564002,
"learning_rate": 4.152291674774384e-06,
"loss": 0.684,
"step": 863
},
{
"epoch": 0.7082284682330791,
"grad_norm": 1.139846056573477,
"learning_rate": 4.130751818730578e-06,
"loss": 0.694,
"step": 864
},
{
"epoch": 0.7090481771083489,
"grad_norm": 0.9589674936132251,
"learning_rate": 4.109253424377773e-06,
"loss": 0.6932,
"step": 865
},
{
"epoch": 0.7098678859836186,
"grad_norm": 0.9890302476720341,
"learning_rate": 4.08779664358547e-06,
"loss": 0.6636,
"step": 866
},
{
"epoch": 0.7106875948588884,
"grad_norm": 0.9797226669901986,
"learning_rate": 4.066381627929202e-06,
"loss": 0.6402,
"step": 867
},
{
"epoch": 0.7115073037341582,
"grad_norm": 0.8947338348199839,
"learning_rate": 4.045008528689457e-06,
"loss": 0.6502,
"step": 868
},
{
"epoch": 0.7123270126094279,
"grad_norm": 1.009428172288083,
"learning_rate": 4.02367749685063e-06,
"loss": 0.69,
"step": 869
},
{
"epoch": 0.7131467214846977,
"grad_norm": 1.0023515381036499,
"learning_rate": 4.00238868309993e-06,
"loss": 0.6457,
"step": 870
},
{
"epoch": 0.7139664303599674,
"grad_norm": 1.023735182603929,
"learning_rate": 3.981142237826332e-06,
"loss": 0.6748,
"step": 871
},
{
"epoch": 0.7147861392352373,
"grad_norm": 0.8361097453343612,
"learning_rate": 3.959938311119511e-06,
"loss": 0.691,
"step": 872
},
{
"epoch": 0.715605848110507,
"grad_norm": 0.9786579923183001,
"learning_rate": 3.9387770527687795e-06,
"loss": 0.708,
"step": 873
},
{
"epoch": 0.7164255569857768,
"grad_norm": 0.9330758573016007,
"learning_rate": 3.917658612262033e-06,
"loss": 0.686,
"step": 874
},
{
"epoch": 0.7172452658610465,
"grad_norm": 0.8921297033457234,
"learning_rate": 3.8965831387846885e-06,
"loss": 0.6373,
"step": 875
},
{
"epoch": 0.7180649747363163,
"grad_norm": 0.9191039031265371,
"learning_rate": 3.875550781218644e-06,
"loss": 0.6759,
"step": 876
},
{
"epoch": 0.7188846836115861,
"grad_norm": 0.9193988849946894,
"learning_rate": 3.854561688141205e-06,
"loss": 0.6226,
"step": 877
},
{
"epoch": 0.7197043924868558,
"grad_norm": 0.9055072813490278,
"learning_rate": 3.833616007824054e-06,
"loss": 0.7186,
"step": 878
},
{
"epoch": 0.7205241013621256,
"grad_norm": 0.923764002674944,
"learning_rate": 3.8127138882321937e-06,
"loss": 0.6966,
"step": 879
},
{
"epoch": 0.7213438102373954,
"grad_norm": 0.8819016775616979,
"learning_rate": 3.791855477022903e-06,
"loss": 0.6962,
"step": 880
},
{
"epoch": 0.7221635191126652,
"grad_norm": 0.8219556932607174,
"learning_rate": 3.7710409215446986e-06,
"loss": 0.623,
"step": 881
},
{
"epoch": 0.7229832279879349,
"grad_norm": 0.9587660698003851,
"learning_rate": 3.750270368836283e-06,
"loss": 0.6532,
"step": 882
},
{
"epoch": 0.7238029368632046,
"grad_norm": 0.8268237223404371,
"learning_rate": 3.729543965625526e-06,
"loss": 0.6632,
"step": 883
},
{
"epoch": 0.7246226457384745,
"grad_norm": 0.9180065638779031,
"learning_rate": 3.7088618583284054e-06,
"loss": 0.6839,
"step": 884
},
{
"epoch": 0.7254423546137442,
"grad_norm": 0.8692664765227096,
"learning_rate": 3.6882241930479824e-06,
"loss": 0.6432,
"step": 885
},
{
"epoch": 0.726262063489014,
"grad_norm": 0.9445773689567405,
"learning_rate": 3.6676311155733745e-06,
"loss": 0.7083,
"step": 886
},
{
"epoch": 0.7270817723642837,
"grad_norm": 0.9717524368536932,
"learning_rate": 3.6470827713787195e-06,
"loss": 0.6923,
"step": 887
},
{
"epoch": 0.7279014812395536,
"grad_norm": 0.9772506676896406,
"learning_rate": 3.6265793056221465e-06,
"loss": 0.6483,
"step": 888
},
{
"epoch": 0.7287211901148233,
"grad_norm": 0.866387488629291,
"learning_rate": 3.606120863144753e-06,
"loss": 0.6555,
"step": 889
},
{
"epoch": 0.729540898990093,
"grad_norm": 1.1061791442769922,
"learning_rate": 3.5857075884695915e-06,
"loss": 0.7394,
"step": 890
},
{
"epoch": 0.7303606078653628,
"grad_norm": 1.0132977251230517,
"learning_rate": 3.5653396258006266e-06,
"loss": 0.7186,
"step": 891
},
{
"epoch": 0.7311803167406326,
"grad_norm": 0.9219493998371818,
"learning_rate": 3.5450171190217364e-06,
"loss": 0.6661,
"step": 892
},
{
"epoch": 0.7320000256159024,
"grad_norm": 0.8874598881661916,
"learning_rate": 3.524740211695683e-06,
"loss": 0.6654,
"step": 893
},
{
"epoch": 0.7328197344911721,
"grad_norm": 0.8425285268988214,
"learning_rate": 3.504509047063108e-06,
"loss": 0.7035,
"step": 894
},
{
"epoch": 0.7336394433664419,
"grad_norm": 1.0143172753012069,
"learning_rate": 3.4843237680415153e-06,
"loss": 0.701,
"step": 895
},
{
"epoch": 0.7344591522417117,
"grad_norm": 0.9405135231254385,
"learning_rate": 3.4641845172242573e-06,
"loss": 0.6425,
"step": 896
},
{
"epoch": 0.7352788611169814,
"grad_norm": 1.0041800243468666,
"learning_rate": 3.444091436879545e-06,
"loss": 0.6752,
"step": 897
},
{
"epoch": 0.7360985699922512,
"grad_norm": 1.009570777532334,
"learning_rate": 3.42404466894942e-06,
"loss": 0.6402,
"step": 898
},
{
"epoch": 0.7369182788675209,
"grad_norm": 1.1709725988747424,
"learning_rate": 3.4040443550487645e-06,
"loss": 0.6433,
"step": 899
},
{
"epoch": 0.7377379877427908,
"grad_norm": 0.8143116658984199,
"learning_rate": 3.384090636464301e-06,
"loss": 0.6544,
"step": 900
},
{
"epoch": 0.7385576966180605,
"grad_norm": 1.0042305767977582,
"learning_rate": 3.364183654153592e-06,
"loss": 0.722,
"step": 901
},
{
"epoch": 0.7393774054933303,
"grad_norm": 0.8906529801709184,
"learning_rate": 3.344323548744044e-06,
"loss": 0.6982,
"step": 902
},
{
"epoch": 0.7401971143686,
"grad_norm": 0.9294602112848915,
"learning_rate": 3.32451046053191e-06,
"loss": 0.657,
"step": 903
},
{
"epoch": 0.7410168232438697,
"grad_norm": 0.9192229109826785,
"learning_rate": 3.304744529481315e-06,
"loss": 0.6908,
"step": 904
},
{
"epoch": 0.7418365321191396,
"grad_norm": 0.8548627717596332,
"learning_rate": 3.285025895223244e-06,
"loss": 0.6853,
"step": 905
},
{
"epoch": 0.7426562409944093,
"grad_norm": 0.9689746663359722,
"learning_rate": 3.26535469705457e-06,
"loss": 0.6798,
"step": 906
},
{
"epoch": 0.7434759498696791,
"grad_norm": 1.042187551928837,
"learning_rate": 3.2457310739370684e-06,
"loss": 0.6865,
"step": 907
},
{
"epoch": 0.7442956587449489,
"grad_norm": 0.872938211904099,
"learning_rate": 3.2261551644964305e-06,
"loss": 0.7,
"step": 908
},
{
"epoch": 0.7451153676202187,
"grad_norm": 1.0529342518600908,
"learning_rate": 3.2066271070212873e-06,
"loss": 0.6322,
"step": 909
},
{
"epoch": 0.7459350764954884,
"grad_norm": 1.0212165414941,
"learning_rate": 3.1871470394622407e-06,
"loss": 0.6355,
"step": 910
},
{
"epoch": 0.7467547853707581,
"grad_norm": 1.2562936325345417,
"learning_rate": 3.167715099430874e-06,
"loss": 0.6317,
"step": 911
},
{
"epoch": 0.747574494246028,
"grad_norm": 1.0486310081277779,
"learning_rate": 3.148331424198784e-06,
"loss": 0.6476,
"step": 912
},
{
"epoch": 0.7483942031212977,
"grad_norm": 0.7723408807633589,
"learning_rate": 3.1289961506966217e-06,
"loss": 0.6478,
"step": 913
},
{
"epoch": 0.7492139119965675,
"grad_norm": 0.9624358151003661,
"learning_rate": 3.1097094155131122e-06,
"loss": 0.6678,
"step": 914
},
{
"epoch": 0.7500336208718372,
"grad_norm": 1.0382726069196302,
"learning_rate": 3.0904713548940936e-06,
"loss": 0.6464,
"step": 915
},
{
"epoch": 0.7508533297471071,
"grad_norm": 0.9680940578884094,
"learning_rate": 3.0712821047415655e-06,
"loss": 0.6739,
"step": 916
},
{
"epoch": 0.7516730386223768,
"grad_norm": 0.8535751200726183,
"learning_rate": 3.052141800612709e-06,
"loss": 0.6641,
"step": 917
},
{
"epoch": 0.7524927474976465,
"grad_norm": 1.0048582490398859,
"learning_rate": 3.0330505777189433e-06,
"loss": 0.6654,
"step": 918
},
{
"epoch": 0.7533124563729163,
"grad_norm": 1.047129920644304,
"learning_rate": 3.0140085709249666e-06,
"loss": 0.7326,
"step": 919
},
{
"epoch": 0.754132165248186,
"grad_norm": 0.9669343743859953,
"learning_rate": 2.9950159147478043e-06,
"loss": 0.678,
"step": 920
},
{
"epoch": 0.7549518741234559,
"grad_norm": 0.9294315680107723,
"learning_rate": 2.976072743355852e-06,
"loss": 0.6667,
"step": 921
},
{
"epoch": 0.7557715829987256,
"grad_norm": 0.8265140472873846,
"learning_rate": 2.957179190567947e-06,
"loss": 0.66,
"step": 922
},
{
"epoch": 0.7565912918739954,
"grad_norm": 0.8608793585657268,
"learning_rate": 2.938335389852397e-06,
"loss": 0.6555,
"step": 923
},
{
"epoch": 0.7574110007492652,
"grad_norm": 1.2620932487023953,
"learning_rate": 2.9195414743260544e-06,
"loss": 0.6906,
"step": 924
},
{
"epoch": 0.7582307096245349,
"grad_norm": 0.9763862710192635,
"learning_rate": 2.9007975767533714e-06,
"loss": 0.659,
"step": 925
},
{
"epoch": 0.7590504184998047,
"grad_norm": 0.9688574524848766,
"learning_rate": 2.882103829545462e-06,
"loss": 0.6188,
"step": 926
},
{
"epoch": 0.7598701273750744,
"grad_norm": 1.0957759386685164,
"learning_rate": 2.863460364759163e-06,
"loss": 0.6952,
"step": 927
},
{
"epoch": 0.7606898362503443,
"grad_norm": 0.8561482820696182,
"learning_rate": 2.844867314096115e-06,
"loss": 0.6417,
"step": 928
},
{
"epoch": 0.761509545125614,
"grad_norm": 0.8861252769371036,
"learning_rate": 2.8263248089018116e-06,
"loss": 0.6702,
"step": 929
},
{
"epoch": 0.7623292540008837,
"grad_norm": 0.9966333630254995,
"learning_rate": 2.8078329801646876e-06,
"loss": 0.6472,
"step": 930
},
{
"epoch": 0.7631489628761535,
"grad_norm": 0.8282486504921222,
"learning_rate": 2.789391958515183e-06,
"loss": 0.6744,
"step": 931
},
{
"epoch": 0.7639686717514232,
"grad_norm": 1.1103278370889822,
"learning_rate": 2.77100187422483e-06,
"loss": 0.6839,
"step": 932
},
{
"epoch": 0.7647883806266931,
"grad_norm": 0.8806453519237031,
"learning_rate": 2.7526628572053227e-06,
"loss": 0.7128,
"step": 933
},
{
"epoch": 0.7656080895019628,
"grad_norm": 0.8267203278810081,
"learning_rate": 2.7343750370076127e-06,
"loss": 0.6889,
"step": 934
},
{
"epoch": 0.7664277983772326,
"grad_norm": 0.892712892353309,
"learning_rate": 2.7161385428209773e-06,
"loss": 0.7054,
"step": 935
},
{
"epoch": 0.7672475072525023,
"grad_norm": 1.0829122991807687,
"learning_rate": 2.6979535034721195e-06,
"loss": 0.6244,
"step": 936
},
{
"epoch": 0.7680672161277721,
"grad_norm": 0.8525762750924293,
"learning_rate": 2.679820047424253e-06,
"loss": 0.6752,
"step": 937
},
{
"epoch": 0.7688869250030419,
"grad_norm": 1.2260893072367014,
"learning_rate": 2.6617383027761944e-06,
"loss": 0.7256,
"step": 938
},
{
"epoch": 0.7697066338783116,
"grad_norm": 0.9384395609985167,
"learning_rate": 2.6437083972614575e-06,
"loss": 0.6837,
"step": 939
},
{
"epoch": 0.7705263427535815,
"grad_norm": 0.8335126538277536,
"learning_rate": 2.625730458247362e-06,
"loss": 0.6362,
"step": 940
},
{
"epoch": 0.7713460516288512,
"grad_norm": 1.0725359491970305,
"learning_rate": 2.6078046127341138e-06,
"loss": 0.6496,
"step": 941
},
{
"epoch": 0.772165760504121,
"grad_norm": 0.8497460781803522,
"learning_rate": 2.5899309873539245e-06,
"loss": 0.6734,
"step": 942
},
{
"epoch": 0.7729854693793907,
"grad_norm": 1.3102510634357916,
"learning_rate": 2.5721097083701085e-06,
"loss": 0.7318,
"step": 943
},
{
"epoch": 0.7738051782546604,
"grad_norm": 1.2114237913130268,
"learning_rate": 2.5543409016761935e-06,
"loss": 0.6461,
"step": 944
},
{
"epoch": 0.7746248871299303,
"grad_norm": 1.0855926504911957,
"learning_rate": 2.5366246927950287e-06,
"loss": 0.6903,
"step": 945
},
{
"epoch": 0.7754445960052,
"grad_norm": 0.9839067747476218,
"learning_rate": 2.518961206877911e-06,
"loss": 0.6686,
"step": 946
},
{
"epoch": 0.7762643048804698,
"grad_norm": 0.9126411281810786,
"learning_rate": 2.5013505687036787e-06,
"loss": 0.6711,
"step": 947
},
{
"epoch": 0.7770840137557395,
"grad_norm": 0.9447077843596731,
"learning_rate": 2.483792902677847e-06,
"loss": 0.6787,
"step": 948
},
{
"epoch": 0.7779037226310094,
"grad_norm": 0.9226341162115747,
"learning_rate": 2.466288332831722e-06,
"loss": 0.7022,
"step": 949
},
{
"epoch": 0.7787234315062791,
"grad_norm": 0.9851965975926599,
"learning_rate": 2.4488369828215286e-06,
"loss": 0.6807,
"step": 950
},
{
"epoch": 0.7795431403815488,
"grad_norm": 0.8872161413720553,
"learning_rate": 2.4314389759275334e-06,
"loss": 0.6458,
"step": 951
},
{
"epoch": 0.7803628492568186,
"grad_norm": 1.0672803564448874,
"learning_rate": 2.4140944350531714e-06,
"loss": 0.6726,
"step": 952
},
{
"epoch": 0.7811825581320884,
"grad_norm": 0.9294630176621735,
"learning_rate": 2.3968034827241926e-06,
"loss": 0.6731,
"step": 953
},
{
"epoch": 0.7820022670073582,
"grad_norm": 0.9497834192887029,
"learning_rate": 2.379566241087774e-06,
"loss": 0.6735,
"step": 954
},
{
"epoch": 0.7828219758826279,
"grad_norm": 0.964988393322474,
"learning_rate": 2.362382831911675e-06,
"loss": 0.6792,
"step": 955
},
{
"epoch": 0.7836416847578977,
"grad_norm": 0.956024146114012,
"learning_rate": 2.345253376583366e-06,
"loss": 0.7057,
"step": 956
},
{
"epoch": 0.7844613936331675,
"grad_norm": 0.8734126317768305,
"learning_rate": 2.3281779961091777e-06,
"loss": 0.6875,
"step": 957
},
{
"epoch": 0.7852811025084372,
"grad_norm": 0.8547910990409069,
"learning_rate": 2.311156811113444e-06,
"loss": 0.6622,
"step": 958
},
{
"epoch": 0.786100811383707,
"grad_norm": 0.9250699377890648,
"learning_rate": 2.294189941837647e-06,
"loss": 0.6789,
"step": 959
},
{
"epoch": 0.7869205202589767,
"grad_norm": 0.8764004620590585,
"learning_rate": 2.277277508139576e-06,
"loss": 0.639,
"step": 960
},
{
"epoch": 0.7877402291342466,
"grad_norm": 0.8582466335669935,
"learning_rate": 2.2604196294924696e-06,
"loss": 0.6417,
"step": 961
},
{
"epoch": 0.7885599380095163,
"grad_norm": 0.8668222080213246,
"learning_rate": 2.2436164249841806e-06,
"loss": 0.6349,
"step": 962
},
{
"epoch": 0.7893796468847861,
"grad_norm": 0.8902652843894273,
"learning_rate": 2.226868013316328e-06,
"loss": 0.6635,
"step": 963
},
{
"epoch": 0.7901993557600558,
"grad_norm": 0.9361063946633907,
"learning_rate": 2.2101745128034657e-06,
"loss": 0.69,
"step": 964
},
{
"epoch": 0.7910190646353256,
"grad_norm": 1.1400546909851692,
"learning_rate": 2.1935360413722397e-06,
"loss": 0.6645,
"step": 965
},
{
"epoch": 0.7918387735105954,
"grad_norm": 0.9784546100909558,
"learning_rate": 2.1769527165605563e-06,
"loss": 0.6632,
"step": 966
},
{
"epoch": 0.7926584823858651,
"grad_norm": 1.0519630640070392,
"learning_rate": 2.160424655516764e-06,
"loss": 0.6838,
"step": 967
},
{
"epoch": 0.793478191261135,
"grad_norm": 0.820555237989852,
"learning_rate": 2.1439519749988045e-06,
"loss": 0.6603,
"step": 968
},
{
"epoch": 0.7942979001364047,
"grad_norm": 0.901442143357133,
"learning_rate": 2.1275347913734023e-06,
"loss": 0.6681,
"step": 969
},
{
"epoch": 0.7951176090116745,
"grad_norm": 0.9278369355829535,
"learning_rate": 2.1111732206152424e-06,
"loss": 0.6591,
"step": 970
},
{
"epoch": 0.7959373178869442,
"grad_norm": 0.9082040141710815,
"learning_rate": 2.0948673783061424e-06,
"loss": 0.6546,
"step": 971
},
{
"epoch": 0.7967570267622139,
"grad_norm": 0.8236637234377587,
"learning_rate": 2.0786173796342468e-06,
"loss": 0.6829,
"step": 972
},
{
"epoch": 0.7975767356374838,
"grad_norm": 0.8278950914465038,
"learning_rate": 2.0624233393932024e-06,
"loss": 0.6674,
"step": 973
},
{
"epoch": 0.7983964445127535,
"grad_norm": 0.8277881306353061,
"learning_rate": 2.0462853719813624e-06,
"loss": 0.6549,
"step": 974
},
{
"epoch": 0.7992161533880233,
"grad_norm": 1.0117916579576536,
"learning_rate": 2.03020359140096e-06,
"loss": 0.6983,
"step": 975
},
{
"epoch": 0.800035862263293,
"grad_norm": 1.0031946637109057,
"learning_rate": 2.0141781112573155e-06,
"loss": 0.7053,
"step": 976
},
{
"epoch": 0.8008555711385629,
"grad_norm": 1.0865781514519064,
"learning_rate": 1.9982090447580305e-06,
"loss": 0.6857,
"step": 977
},
{
"epoch": 0.8016752800138326,
"grad_norm": 0.8000496861010923,
"learning_rate": 1.9822965047121854e-06,
"loss": 0.6302,
"step": 978
},
{
"epoch": 0.8024949888891023,
"grad_norm": 0.8618951712994316,
"learning_rate": 1.9664406035295493e-06,
"loss": 0.6989,
"step": 979
},
{
"epoch": 0.8033146977643721,
"grad_norm": 0.8022171421495404,
"learning_rate": 1.950641453219775e-06,
"loss": 0.6887,
"step": 980
},
{
"epoch": 0.8041344066396419,
"grad_norm": 0.9235048546356935,
"learning_rate": 1.934899165391623e-06,
"loss": 0.6269,
"step": 981
},
{
"epoch": 0.8049541155149117,
"grad_norm": 0.9843383531287645,
"learning_rate": 1.9192138512521585e-06,
"loss": 0.6589,
"step": 982
},
{
"epoch": 0.8057738243901814,
"grad_norm": 0.8952486199104005,
"learning_rate": 1.9035856216059724e-06,
"loss": 0.6586,
"step": 983
},
{
"epoch": 0.8065935332654512,
"grad_norm": 0.8304878870826209,
"learning_rate": 1.8880145868543997e-06,
"loss": 0.6655,
"step": 984
},
{
"epoch": 0.807413242140721,
"grad_norm": 0.7784027879651656,
"learning_rate": 1.8725008569947366e-06,
"loss": 0.6735,
"step": 985
},
{
"epoch": 0.8082329510159907,
"grad_norm": 0.9434260894780481,
"learning_rate": 1.8570445416194639e-06,
"loss": 0.6683,
"step": 986
},
{
"epoch": 0.8090526598912605,
"grad_norm": 0.9912694995741538,
"learning_rate": 1.8416457499154727e-06,
"loss": 0.6676,
"step": 987
},
{
"epoch": 0.8098723687665302,
"grad_norm": 0.8386268961105685,
"learning_rate": 1.8263045906633004e-06,
"loss": 0.6616,
"step": 988
},
{
"epoch": 0.8106920776418001,
"grad_norm": 0.8510957122599717,
"learning_rate": 1.811021172236348e-06,
"loss": 0.6668,
"step": 989
},
{
"epoch": 0.8115117865170698,
"grad_norm": 0.9025552221342271,
"learning_rate": 1.7957956026001256e-06,
"loss": 0.6782,
"step": 990
},
{
"epoch": 0.8123314953923396,
"grad_norm": 0.9664891225593596,
"learning_rate": 1.7806279893114874e-06,
"loss": 0.6968,
"step": 991
},
{
"epoch": 0.8131512042676093,
"grad_norm": 0.9287210025886622,
"learning_rate": 1.7655184395178683e-06,
"loss": 0.6634,
"step": 992
},
{
"epoch": 0.8139709131428791,
"grad_norm": 0.8731715791190819,
"learning_rate": 1.750467059956531e-06,
"loss": 0.6563,
"step": 993
},
{
"epoch": 0.8147906220181489,
"grad_norm": 0.8735038937768553,
"learning_rate": 1.7354739569538081e-06,
"loss": 0.6834,
"step": 994
},
{
"epoch": 0.8156103308934186,
"grad_norm": 0.8499903726596456,
"learning_rate": 1.7205392364243623e-06,
"loss": 0.623,
"step": 995
},
{
"epoch": 0.8164300397686884,
"grad_norm": 0.9766262650325924,
"learning_rate": 1.705663003870418e-06,
"loss": 0.6504,
"step": 996
},
{
"epoch": 0.8172497486439582,
"grad_norm": 1.017085956065125,
"learning_rate": 1.6908453643810342e-06,
"loss": 0.6592,
"step": 997
},
{
"epoch": 0.8180694575192279,
"grad_norm": 0.9289727088128958,
"learning_rate": 1.6760864226313534e-06,
"loss": 0.6663,
"step": 998
},
{
"epoch": 0.8188891663944977,
"grad_norm": 0.9951441703648285,
"learning_rate": 1.6613862828818628e-06,
"loss": 0.6639,
"step": 999
},
{
"epoch": 0.8197088752697674,
"grad_norm": 0.9446239827791956,
"learning_rate": 1.6467450489776581e-06,
"loss": 0.6874,
"step": 1000
},
{
"epoch": 0.8205285841450373,
"grad_norm": 1.079838709617824,
"learning_rate": 1.6321628243477194e-06,
"loss": 0.6924,
"step": 1001
},
{
"epoch": 0.821348293020307,
"grad_norm": 0.826828487282096,
"learning_rate": 1.617639712004162e-06,
"loss": 0.6635,
"step": 1002
},
{
"epoch": 0.8221680018955768,
"grad_norm": 0.9088825217559868,
"learning_rate": 1.6031758145415222e-06,
"loss": 0.6955,
"step": 1003
},
{
"epoch": 0.8229877107708465,
"grad_norm": 0.9049161403672648,
"learning_rate": 1.5887712341360294e-06,
"loss": 0.6336,
"step": 1004
},
{
"epoch": 0.8238074196461163,
"grad_norm": 0.9976036028552778,
"learning_rate": 1.5744260725448845e-06,
"loss": 0.6847,
"step": 1005
},
{
"epoch": 0.8246271285213861,
"grad_norm": 0.9287501032058516,
"learning_rate": 1.560140431105539e-06,
"loss": 0.6966,
"step": 1006
},
{
"epoch": 0.8254468373966558,
"grad_norm": 0.7670223010504882,
"learning_rate": 1.5459144107349788e-06,
"loss": 0.6556,
"step": 1007
},
{
"epoch": 0.8262665462719256,
"grad_norm": 0.8773607230727865,
"learning_rate": 1.531748111929021e-06,
"loss": 0.6095,
"step": 1008
},
{
"epoch": 0.8270862551471954,
"grad_norm": 0.8281581453197533,
"learning_rate": 1.5176416347615886e-06,
"loss": 0.634,
"step": 1009
},
{
"epoch": 0.8279059640224652,
"grad_norm": 0.8833993434064099,
"learning_rate": 1.5035950788840125e-06,
"loss": 0.7018,
"step": 1010
},
{
"epoch": 0.8287256728977349,
"grad_norm": 1.0619553805112998,
"learning_rate": 1.4896085435243279e-06,
"loss": 0.712,
"step": 1011
},
{
"epoch": 0.8295453817730046,
"grad_norm": 0.807303499698227,
"learning_rate": 1.4756821274865696e-06,
"loss": 0.6812,
"step": 1012
},
{
"epoch": 0.8303650906482745,
"grad_norm": 0.9987390534039866,
"learning_rate": 1.4618159291500777e-06,
"loss": 0.6761,
"step": 1013
},
{
"epoch": 0.8311847995235442,
"grad_norm": 0.9612082775507431,
"learning_rate": 1.4480100464687973e-06,
"loss": 0.6749,
"step": 1014
},
{
"epoch": 0.832004508398814,
"grad_norm": 1.0170321141390004,
"learning_rate": 1.4342645769705977e-06,
"loss": 0.653,
"step": 1015
},
{
"epoch": 0.8328242172740837,
"grad_norm": 0.9140872644345462,
"learning_rate": 1.4205796177565688e-06,
"loss": 0.65,
"step": 1016
},
{
"epoch": 0.8336439261493536,
"grad_norm": 0.8623418110911005,
"learning_rate": 1.406955265500346e-06,
"loss": 0.6754,
"step": 1017
},
{
"epoch": 0.8344636350246233,
"grad_norm": 0.8654941456068144,
"learning_rate": 1.3933916164474193e-06,
"loss": 0.6854,
"step": 1018
},
{
"epoch": 0.835283343899893,
"grad_norm": 0.7859316136384058,
"learning_rate": 1.3798887664144634e-06,
"loss": 0.6438,
"step": 1019
},
{
"epoch": 0.8361030527751628,
"grad_norm": 0.8278343695885293,
"learning_rate": 1.3664468107886496e-06,
"loss": 0.6458,
"step": 1020
},
{
"epoch": 0.8369227616504326,
"grad_norm": 1.0053271561966406,
"learning_rate": 1.3530658445269784e-06,
"loss": 0.6747,
"step": 1021
},
{
"epoch": 0.8377424705257024,
"grad_norm": 0.8571756315788411,
"learning_rate": 1.339745962155613e-06,
"loss": 0.6327,
"step": 1022
},
{
"epoch": 0.8385621794009721,
"grad_norm": 0.8177173742308131,
"learning_rate": 1.3264872577692022e-06,
"loss": 0.617,
"step": 1023
},
{
"epoch": 0.8393818882762419,
"grad_norm": 0.9488632234767527,
"learning_rate": 1.3132898250302173e-06,
"loss": 0.6827,
"step": 1024
},
{
"epoch": 0.8402015971515117,
"grad_norm": 0.9562228849480632,
"learning_rate": 1.3001537571682965e-06,
"loss": 0.6302,
"step": 1025
},
{
"epoch": 0.8410213060267814,
"grad_norm": 1.1425960755342779,
"learning_rate": 1.2870791469795818e-06,
"loss": 0.6884,
"step": 1026
},
{
"epoch": 0.8418410149020512,
"grad_norm": 0.9853520084534523,
"learning_rate": 1.2740660868260634e-06,
"loss": 0.6806,
"step": 1027
},
{
"epoch": 0.8426607237773209,
"grad_norm": 1.0022219971543274,
"learning_rate": 1.2611146686349284e-06,
"loss": 0.6807,
"step": 1028
},
{
"epoch": 0.8434804326525908,
"grad_norm": 0.893880040976814,
"learning_rate": 1.2482249838979143e-06,
"loss": 0.6786,
"step": 1029
},
{
"epoch": 0.8443001415278605,
"grad_norm": 0.9438293772251487,
"learning_rate": 1.2353971236706564e-06,
"loss": 0.6687,
"step": 1030
},
{
"epoch": 0.8451198504031303,
"grad_norm": 0.850739724195347,
"learning_rate": 1.2226311785720468e-06,
"loss": 0.6814,
"step": 1031
},
{
"epoch": 0.8459395592784,
"grad_norm": 0.9123061511307571,
"learning_rate": 1.209927238783598e-06,
"loss": 0.6377,
"step": 1032
},
{
"epoch": 0.8467592681536698,
"grad_norm": 1.0410952286043216,
"learning_rate": 1.1972853940488017e-06,
"loss": 0.6728,
"step": 1033
},
{
"epoch": 0.8475789770289396,
"grad_norm": 0.8805296139827917,
"learning_rate": 1.1847057336724965e-06,
"loss": 0.6694,
"step": 1034
},
{
"epoch": 0.8483986859042093,
"grad_norm": 1.47221146561222,
"learning_rate": 1.1721883465202333e-06,
"loss": 0.692,
"step": 1035
},
{
"epoch": 0.8492183947794791,
"grad_norm": 0.9067471226862389,
"learning_rate": 1.1597333210176587e-06,
"loss": 0.6574,
"step": 1036
},
{
"epoch": 0.8500381036547489,
"grad_norm": 1.1270618971343118,
"learning_rate": 1.1473407451498752e-06,
"loss": 0.6743,
"step": 1037
},
{
"epoch": 0.8508578125300187,
"grad_norm": 0.9159849495527415,
"learning_rate": 1.1350107064608317e-06,
"loss": 0.6707,
"step": 1038
},
{
"epoch": 0.8516775214052884,
"grad_norm": 0.9233817138782029,
"learning_rate": 1.122743292052697e-06,
"loss": 0.6362,
"step": 1039
},
{
"epoch": 0.8524972302805581,
"grad_norm": 0.9674854700872919,
"learning_rate": 1.1105385885852483e-06,
"loss": 0.6596,
"step": 1040
},
{
"epoch": 0.853316939155828,
"grad_norm": 1.0397637398797948,
"learning_rate": 1.0983966822752624e-06,
"loss": 0.676,
"step": 1041
},
{
"epoch": 0.8541366480310977,
"grad_norm": 0.8718429004826903,
"learning_rate": 1.0863176588958957e-06,
"loss": 0.6991,
"step": 1042
},
{
"epoch": 0.8549563569063675,
"grad_norm": 0.8848179204891644,
"learning_rate": 1.0743016037760946e-06,
"loss": 0.7144,
"step": 1043
},
{
"epoch": 0.8557760657816372,
"grad_norm": 0.9122319534546866,
"learning_rate": 1.0623486017999762e-06,
"loss": 0.634,
"step": 1044
},
{
"epoch": 0.8565957746569071,
"grad_norm": 0.9359504670280605,
"learning_rate": 1.0504587374062392e-06,
"loss": 0.6754,
"step": 1045
},
{
"epoch": 0.8574154835321768,
"grad_norm": 0.8702425244105505,
"learning_rate": 1.0386320945875627e-06,
"loss": 0.6383,
"step": 1046
},
{
"epoch": 0.8582351924074465,
"grad_norm": 0.8737486102788375,
"learning_rate": 1.0268687568900159e-06,
"loss": 0.6489,
"step": 1047
},
{
"epoch": 0.8590549012827163,
"grad_norm": 1.1040377262967278,
"learning_rate": 1.0151688074124645e-06,
"loss": 0.6425,
"step": 1048
},
{
"epoch": 0.859874610157986,
"grad_norm": 0.8518177361283181,
"learning_rate": 1.0035323288059861e-06,
"loss": 0.6656,
"step": 1049
},
{
"epoch": 0.8606943190332559,
"grad_norm": 0.8995051092213161,
"learning_rate": 9.919594032732893e-07,
"loss": 0.6993,
"step": 1050
},
{
"epoch": 0.8615140279085256,
"grad_norm": 0.7937717563799589,
"learning_rate": 9.804501125681243e-07,
"loss": 0.6307,
"step": 1051
},
{
"epoch": 0.8623337367837954,
"grad_norm": 0.8586499893574099,
"learning_rate": 9.690045379947134e-07,
"loss": 0.6334,
"step": 1052
},
{
"epoch": 0.8631534456590652,
"grad_norm": 0.9164736809250863,
"learning_rate": 9.576227604071731e-07,
"loss": 0.6577,
"step": 1053
},
{
"epoch": 0.8639731545343349,
"grad_norm": 0.8103244437689117,
"learning_rate": 9.463048602089431e-07,
"loss": 0.6353,
"step": 1054
},
{
"epoch": 0.8647928634096047,
"grad_norm": 0.8159086830029461,
"learning_rate": 9.350509173522193e-07,
"loss": 0.6391,
"step": 1055
},
{
"epoch": 0.8656125722848744,
"grad_norm": 1.207566574882582,
"learning_rate": 9.23861011337387e-07,
"loss": 0.6902,
"step": 1056
},
{
"epoch": 0.8664322811601443,
"grad_norm": 0.8643053405351546,
"learning_rate": 9.127352212124663e-07,
"loss": 0.6625,
"step": 1057
},
{
"epoch": 0.867251990035414,
"grad_norm": 1.086387273458921,
"learning_rate": 9.016736255725434e-07,
"loss": 0.6848,
"step": 1058
},
{
"epoch": 0.8680716989106837,
"grad_norm": 0.9535182604729994,
"learning_rate": 8.906763025592191e-07,
"loss": 0.6621,
"step": 1059
},
{
"epoch": 0.8688914077859535,
"grad_norm": 0.8200120355091203,
"learning_rate": 8.797433298600622e-07,
"loss": 0.6204,
"step": 1060
},
{
"epoch": 0.8697111166612232,
"grad_norm": 0.8758159915628473,
"learning_rate": 8.688747847080514e-07,
"loss": 0.6821,
"step": 1061
},
{
"epoch": 0.8705308255364931,
"grad_norm": 0.8637198213369558,
"learning_rate": 8.580707438810398e-07,
"loss": 0.6298,
"step": 1062
},
{
"epoch": 0.8713505344117628,
"grad_norm": 1.0925923984170347,
"learning_rate": 8.473312837012027e-07,
"loss": 0.6655,
"step": 1063
},
{
"epoch": 0.8721702432870326,
"grad_norm": 0.9078494764770126,
"learning_rate": 8.366564800345089e-07,
"loss": 0.6705,
"step": 1064
},
{
"epoch": 0.8729899521623024,
"grad_norm": 1.0241245453771914,
"learning_rate": 8.260464082901732e-07,
"loss": 0.6295,
"step": 1065
},
{
"epoch": 0.8738096610375721,
"grad_norm": 0.8380522744507208,
"learning_rate": 8.155011434201332e-07,
"loss": 0.6763,
"step": 1066
},
{
"epoch": 0.8746293699128419,
"grad_norm": 0.8074894103287081,
"learning_rate": 8.050207599185134e-07,
"loss": 0.683,
"step": 1067
},
{
"epoch": 0.8754490787881116,
"grad_norm": 0.8912744939307611,
"learning_rate": 7.946053318211045e-07,
"loss": 0.6312,
"step": 1068
},
{
"epoch": 0.8762687876633815,
"grad_norm": 0.9612186927976536,
"learning_rate": 7.842549327048366e-07,
"loss": 0.7088,
"step": 1069
},
{
"epoch": 0.8770884965386512,
"grad_norm": 1.0008173850430995,
"learning_rate": 7.739696356872562e-07,
"loss": 0.7021,
"step": 1070
},
{
"epoch": 0.877908205413921,
"grad_norm": 0.8849469226035567,
"learning_rate": 7.637495134260242e-07,
"loss": 0.6769,
"step": 1071
},
{
"epoch": 0.8787279142891907,
"grad_norm": 1.0184957374001902,
"learning_rate": 7.535946381183812e-07,
"loss": 0.7037,
"step": 1072
},
{
"epoch": 0.8795476231644604,
"grad_norm": 1.194316427759829,
"learning_rate": 7.435050815006562e-07,
"loss": 0.6789,
"step": 1073
},
{
"epoch": 0.8803673320397303,
"grad_norm": 1.0883084235517622,
"learning_rate": 7.334809148477484e-07,
"loss": 0.7025,
"step": 1074
},
{
"epoch": 0.881187040915,
"grad_norm": 1.0603906546174109,
"learning_rate": 7.23522208972628e-07,
"loss": 0.6578,
"step": 1075
},
{
"epoch": 0.8820067497902698,
"grad_norm": 1.1398300355185007,
"learning_rate": 7.136290342258378e-07,
"loss": 0.6971,
"step": 1076
},
{
"epoch": 0.8828264586655395,
"grad_norm": 0.8774981499188873,
"learning_rate": 7.038014604949883e-07,
"loss": 0.6888,
"step": 1077
},
{
"epoch": 0.8836461675408094,
"grad_norm": 0.8140275698306718,
"learning_rate": 6.940395572042791e-07,
"loss": 0.6858,
"step": 1078
},
{
"epoch": 0.8844658764160791,
"grad_norm": 0.8453390437224966,
"learning_rate": 6.843433933139909e-07,
"loss": 0.6379,
"step": 1079
},
{
"epoch": 0.8852855852913488,
"grad_norm": 1.0411942804380319,
"learning_rate": 6.747130373200095e-07,
"loss": 0.6612,
"step": 1080
},
{
"epoch": 0.8861052941666187,
"grad_norm": 0.8311565105784688,
"learning_rate": 6.651485572533379e-07,
"loss": 0.6691,
"step": 1081
},
{
"epoch": 0.8869250030418884,
"grad_norm": 0.912035929715239,
"learning_rate": 6.556500206796179e-07,
"loss": 0.6868,
"step": 1082
},
{
"epoch": 0.8877447119171582,
"grad_norm": 0.876072423271328,
"learning_rate": 6.462174946986511e-07,
"loss": 0.6767,
"step": 1083
},
{
"epoch": 0.8885644207924279,
"grad_norm": 1.094558648301763,
"learning_rate": 6.368510459439248e-07,
"loss": 0.6551,
"step": 1084
},
{
"epoch": 0.8893841296676978,
"grad_norm": 1.0058106680496375,
"learning_rate": 6.275507405821435e-07,
"loss": 0.659,
"step": 1085
},
{
"epoch": 0.8902038385429675,
"grad_norm": 1.0281716615641452,
"learning_rate": 6.183166443127587e-07,
"loss": 0.6195,
"step": 1086
},
{
"epoch": 0.8910235474182372,
"grad_norm": 0.8900490729372945,
"learning_rate": 6.091488223675058e-07,
"loss": 0.6431,
"step": 1087
},
{
"epoch": 0.891843256293507,
"grad_norm": 0.89377999658744,
"learning_rate": 6.000473395099438e-07,
"loss": 0.6933,
"step": 1088
},
{
"epoch": 0.8926629651687767,
"grad_norm": 0.8798450324869875,
"learning_rate": 5.910122600349966e-07,
"loss": 0.6074,
"step": 1089
},
{
"epoch": 0.8934826740440466,
"grad_norm": 0.9568494676429672,
"learning_rate": 5.820436477685021e-07,
"loss": 0.6213,
"step": 1090
},
{
"epoch": 0.8943023829193163,
"grad_norm": 0.8433964541751362,
"learning_rate": 5.73141566066755e-07,
"loss": 0.6458,
"step": 1091
},
{
"epoch": 0.8951220917945861,
"grad_norm": 0.9750678694695359,
"learning_rate": 5.64306077816068e-07,
"loss": 0.6787,
"step": 1092
},
{
"epoch": 0.8959418006698558,
"grad_norm": 0.8936310375649036,
"learning_rate": 5.555372454323182e-07,
"loss": 0.6221,
"step": 1093
},
{
"epoch": 0.8967615095451256,
"grad_norm": 0.8979672389708601,
"learning_rate": 5.468351308605135e-07,
"loss": 0.6712,
"step": 1094
},
{
"epoch": 0.8975812184203954,
"grad_norm": 0.914244757493469,
"learning_rate": 5.3819979557435e-07,
"loss": 0.6353,
"step": 1095
},
{
"epoch": 0.8984009272956651,
"grad_norm": 0.8932300642857027,
"learning_rate": 5.296313005757814e-07,
"loss": 0.6469,
"step": 1096
},
{
"epoch": 0.899220636170935,
"grad_norm": 0.865137475671906,
"learning_rate": 5.211297063945875e-07,
"loss": 0.6762,
"step": 1097
},
{
"epoch": 0.9000403450462047,
"grad_norm": 1.098552252129474,
"learning_rate": 5.126950730879399e-07,
"loss": 0.6685,
"step": 1098
},
{
"epoch": 0.9008600539214745,
"grad_norm": 0.8708434134433578,
"learning_rate": 5.043274602399939e-07,
"loss": 0.6523,
"step": 1099
},
{
"epoch": 0.9016797627967442,
"grad_norm": 0.9645676611418972,
"learning_rate": 4.96026926961447e-07,
"loss": 0.6556,
"step": 1100
},
{
"epoch": 0.9024994716720139,
"grad_norm": 1.0830024107973766,
"learning_rate": 4.877935318891381e-07,
"loss": 0.6746,
"step": 1101
},
{
"epoch": 0.9033191805472838,
"grad_norm": 0.8242951838444681,
"learning_rate": 4.796273331856227e-07,
"loss": 0.6753,
"step": 1102
},
{
"epoch": 0.9041388894225535,
"grad_norm": 0.8975910957649728,
"learning_rate": 4.715283885387678e-07,
"loss": 0.6277,
"step": 1103
},
{
"epoch": 0.9049585982978233,
"grad_norm": 0.9292669529538012,
"learning_rate": 4.634967551613423e-07,
"loss": 0.6818,
"step": 1104
},
{
"epoch": 0.905778307173093,
"grad_norm": 0.9312011634702003,
"learning_rate": 4.555324897906133e-07,
"loss": 0.6392,
"step": 1105
},
{
"epoch": 0.9065980160483629,
"grad_norm": 0.8494587954637014,
"learning_rate": 4.476356486879474e-07,
"loss": 0.6465,
"step": 1106
},
{
"epoch": 0.9074177249236326,
"grad_norm": 0.8155434404735615,
"learning_rate": 4.398062876384046e-07,
"loss": 0.6289,
"step": 1107
},
{
"epoch": 0.9082374337989023,
"grad_norm": 1.2227705613074369,
"learning_rate": 4.320444619503583e-07,
"loss": 0.7109,
"step": 1108
},
{
"epoch": 0.9090571426741721,
"grad_norm": 0.9180089891055171,
"learning_rate": 4.243502264550903e-07,
"loss": 0.6954,
"step": 1109
},
{
"epoch": 0.9098768515494419,
"grad_norm": 0.9133146645632352,
"learning_rate": 4.1672363550641415e-07,
"loss": 0.6839,
"step": 1110
},
{
"epoch": 0.9106965604247117,
"grad_norm": 0.8529117809566804,
"learning_rate": 4.0916474298028694e-07,
"loss": 0.6074,
"step": 1111
},
{
"epoch": 0.9115162692999814,
"grad_norm": 0.8455126799753356,
"learning_rate": 4.016736022744272e-07,
"loss": 0.6156,
"step": 1112
},
{
"epoch": 0.9123359781752512,
"grad_norm": 0.9282035962270619,
"learning_rate": 3.942502663079395e-07,
"loss": 0.6654,
"step": 1113
},
{
"epoch": 0.913155687050521,
"grad_norm": 0.9543527686686821,
"learning_rate": 3.8689478752094167e-07,
"loss": 0.6537,
"step": 1114
},
{
"epoch": 0.9139753959257907,
"grad_norm": 0.8637297139852697,
"learning_rate": 3.7960721787419164e-07,
"loss": 0.6474,
"step": 1115
},
{
"epoch": 0.9147951048010605,
"grad_norm": 0.926433061110871,
"learning_rate": 3.7238760884872216e-07,
"loss": 0.6457,
"step": 1116
},
{
"epoch": 0.9156148136763302,
"grad_norm": 0.8768086985477676,
"learning_rate": 3.6523601144548003e-07,
"loss": 0.6189,
"step": 1117
},
{
"epoch": 0.9164345225516001,
"grad_norm": 0.8850033598348138,
"learning_rate": 3.5815247618495753e-07,
"loss": 0.6636,
"step": 1118
},
{
"epoch": 0.9172542314268698,
"grad_norm": 0.8505582672316075,
"learning_rate": 3.5113705310684363e-07,
"loss": 0.6551,
"step": 1119
},
{
"epoch": 0.9180739403021395,
"grad_norm": 0.9087474484429193,
"learning_rate": 3.441897917696679e-07,
"loss": 0.6502,
"step": 1120
},
{
"epoch": 0.9188936491774093,
"grad_norm": 0.8445372614257574,
"learning_rate": 3.373107412504473e-07,
"loss": 0.6821,
"step": 1121
},
{
"epoch": 0.9197133580526791,
"grad_norm": 0.8912731411271559,
"learning_rate": 3.30499950144344e-07,
"loss": 0.6549,
"step": 1122
},
{
"epoch": 0.9205330669279489,
"grad_norm": 0.9415857195110771,
"learning_rate": 3.2375746656432284e-07,
"loss": 0.6606,
"step": 1123
},
{
"epoch": 0.9213527758032186,
"grad_norm": 0.7727180500458501,
"learning_rate": 3.170833381408045e-07,
"loss": 0.639,
"step": 1124
},
{
"epoch": 0.9221724846784884,
"grad_norm": 0.9154816443757596,
"learning_rate": 3.10477612021336e-07,
"loss": 0.6512,
"step": 1125
},
{
"epoch": 0.9229921935537582,
"grad_norm": 0.9440876864198443,
"learning_rate": 3.03940334870253e-07,
"loss": 0.6368,
"step": 1126
},
{
"epoch": 0.9238119024290279,
"grad_norm": 1.0560060401565659,
"learning_rate": 2.974715528683547e-07,
"loss": 0.636,
"step": 1127
},
{
"epoch": 0.9246316113042977,
"grad_norm": 0.8139717887838253,
"learning_rate": 2.910713117125719e-07,
"loss": 0.6302,
"step": 1128
},
{
"epoch": 0.9254513201795674,
"grad_norm": 1.045414314928848,
"learning_rate": 2.8473965661565353e-07,
"loss": 0.663,
"step": 1129
},
{
"epoch": 0.9262710290548373,
"grad_norm": 0.8417438374139112,
"learning_rate": 2.784766323058352e-07,
"loss": 0.6633,
"step": 1130
},
{
"epoch": 0.927090737930107,
"grad_norm": 0.8951342780580024,
"learning_rate": 2.7228228302653037e-07,
"loss": 0.6604,
"step": 1131
},
{
"epoch": 0.9279104468053768,
"grad_norm": 0.8089248766972518,
"learning_rate": 2.661566525360193e-07,
"loss": 0.6462,
"step": 1132
},
{
"epoch": 0.9287301556806465,
"grad_norm": 0.8468240944436343,
"learning_rate": 2.600997841071329e-07,
"loss": 0.6285,
"step": 1133
},
{
"epoch": 0.9295498645559163,
"grad_norm": 0.9333797154879018,
"learning_rate": 2.54111720526955e-07,
"loss": 0.6656,
"step": 1134
},
{
"epoch": 0.9303695734311861,
"grad_norm": 0.8329444797621915,
"learning_rate": 2.4819250409651605e-07,
"loss": 0.6831,
"step": 1135
},
{
"epoch": 0.9311892823064558,
"grad_norm": 1.2136965442756251,
"learning_rate": 2.423421766304934e-07,
"loss": 0.6543,
"step": 1136
},
{
"epoch": 0.9320089911817256,
"grad_norm": 0.8816308441932743,
"learning_rate": 2.3656077945691802e-07,
"loss": 0.6543,
"step": 1137
},
{
"epoch": 0.9328287000569954,
"grad_norm": 0.9392060239540891,
"learning_rate": 2.308483534168815e-07,
"loss": 0.6733,
"step": 1138
},
{
"epoch": 0.9336484089322652,
"grad_norm": 0.8497801568980519,
"learning_rate": 2.2520493886424743e-07,
"loss": 0.6325,
"step": 1139
},
{
"epoch": 0.9344681178075349,
"grad_norm": 0.9362822791260688,
"learning_rate": 2.1963057566536715e-07,
"loss": 0.6936,
"step": 1140
},
{
"epoch": 0.9352878266828046,
"grad_norm": 0.8327642996107147,
"learning_rate": 2.1412530319879888e-07,
"loss": 0.6679,
"step": 1141
},
{
"epoch": 0.9361075355580745,
"grad_norm": 0.9460014740494975,
"learning_rate": 2.0868916035502785e-07,
"loss": 0.6843,
"step": 1142
},
{
"epoch": 0.9369272444333442,
"grad_norm": 0.7481253771773546,
"learning_rate": 2.0332218553618888e-07,
"loss": 0.6417,
"step": 1143
},
{
"epoch": 0.937746953308614,
"grad_norm": 0.8811727341291541,
"learning_rate": 1.9802441665580208e-07,
"loss": 0.6493,
"step": 1144
},
{
"epoch": 0.9385666621838837,
"grad_norm": 0.9481590444527167,
"learning_rate": 1.9279589113850082e-07,
"loss": 0.7065,
"step": 1145
},
{
"epoch": 0.9393863710591536,
"grad_norm": 0.8560722176171286,
"learning_rate": 1.8763664591976426e-07,
"loss": 0.6344,
"step": 1146
},
{
"epoch": 0.9402060799344233,
"grad_norm": 0.9370144149327383,
"learning_rate": 1.825467174456652e-07,
"loss": 0.6197,
"step": 1147
},
{
"epoch": 0.941025788809693,
"grad_norm": 0.8969052651632967,
"learning_rate": 1.7752614167260484e-07,
"loss": 0.6497,
"step": 1148
},
{
"epoch": 0.9418454976849628,
"grad_norm": 0.8869090802777161,
"learning_rate": 1.7257495406705959e-07,
"loss": 0.6609,
"step": 1149
},
{
"epoch": 0.9426652065602326,
"grad_norm": 0.9539412238499507,
"learning_rate": 1.6769318960533465e-07,
"loss": 0.6796,
"step": 1150
},
{
"epoch": 0.9434849154355024,
"grad_norm": 0.7521417320985249,
"learning_rate": 1.6288088277331303e-07,
"loss": 0.6121,
"step": 1151
},
{
"epoch": 0.9443046243107721,
"grad_norm": 1.0216434527329838,
"learning_rate": 1.5813806756621475e-07,
"loss": 0.6834,
"step": 1152
},
{
"epoch": 0.9451243331860419,
"grad_norm": 0.8289968916518831,
"learning_rate": 1.5346477748835353e-07,
"loss": 0.6639,
"step": 1153
},
{
"epoch": 0.9459440420613117,
"grad_norm": 0.9267471219405004,
"learning_rate": 1.488610455529038e-07,
"loss": 0.6564,
"step": 1154
},
{
"epoch": 0.9467637509365814,
"grad_norm": 0.96416966239322,
"learning_rate": 1.4432690428166528e-07,
"loss": 0.666,
"step": 1155
},
{
"epoch": 0.9475834598118512,
"grad_norm": 0.930348875793358,
"learning_rate": 1.39862385704832e-07,
"loss": 0.6035,
"step": 1156
},
{
"epoch": 0.9484031686871209,
"grad_norm": 0.9275098597443717,
"learning_rate": 1.3546752136076924e-07,
"loss": 0.6654,
"step": 1157
},
{
"epoch": 0.9492228775623908,
"grad_norm": 0.808206298742425,
"learning_rate": 1.3114234229578803e-07,
"loss": 0.6572,
"step": 1158
},
{
"epoch": 0.9500425864376605,
"grad_norm": 0.9072462331016414,
"learning_rate": 1.2688687906392772e-07,
"loss": 0.6523,
"step": 1159
},
{
"epoch": 0.9508622953129303,
"grad_norm": 0.8199613565505572,
"learning_rate": 1.2270116172673818e-07,
"loss": 0.6383,
"step": 1160
},
{
"epoch": 0.9516820041882,
"grad_norm": 1.1689008518797992,
"learning_rate": 1.1858521985307126e-07,
"loss": 0.6434,
"step": 1161
},
{
"epoch": 0.9525017130634698,
"grad_norm": 0.8559227052772139,
"learning_rate": 1.1453908251886636e-07,
"loss": 0.69,
"step": 1162
},
{
"epoch": 0.9533214219387396,
"grad_norm": 1.0858928746075616,
"learning_rate": 1.105627783069485e-07,
"loss": 0.6668,
"step": 1163
},
{
"epoch": 0.9541411308140093,
"grad_norm": 1.0404087369266204,
"learning_rate": 1.0665633530682618e-07,
"loss": 0.6338,
"step": 1164
},
{
"epoch": 0.9549608396892791,
"grad_norm": 0.9323201277074105,
"learning_rate": 1.0281978111449375e-07,
"loss": 0.6843,
"step": 1165
},
{
"epoch": 0.9557805485645489,
"grad_norm": 0.8860786574248212,
"learning_rate": 9.905314283223166e-08,
"loss": 0.6883,
"step": 1166
},
{
"epoch": 0.9566002574398187,
"grad_norm": 0.9267917293471627,
"learning_rate": 9.535644706842318e-08,
"loss": 0.662,
"step": 1167
},
{
"epoch": 0.9574199663150884,
"grad_norm": 0.9511008740738184,
"learning_rate": 9.172971993735902e-08,
"loss": 0.681,
"step": 1168
},
{
"epoch": 0.9582396751903581,
"grad_norm": 0.8712624840997685,
"learning_rate": 8.817298705905642e-08,
"loss": 0.6504,
"step": 1169
},
{
"epoch": 0.959059384065628,
"grad_norm": 1.1239770942637772,
"learning_rate": 8.468627355907811e-08,
"loss": 0.6455,
"step": 1170
},
{
"epoch": 0.9598790929408977,
"grad_norm": 0.8854046562132684,
"learning_rate": 8.12696040683525e-08,
"loss": 0.6633,
"step": 1171
},
{
"epoch": 0.9606988018161675,
"grad_norm": 0.8830016826566768,
"learning_rate": 7.792300272300268e-08,
"loss": 0.6753,
"step": 1172
},
{
"epoch": 0.9615185106914372,
"grad_norm": 0.884040350153176,
"learning_rate": 7.464649316417438e-08,
"loss": 0.6677,
"step": 1173
},
{
"epoch": 0.9623382195667071,
"grad_norm": 1.0779691457914105,
"learning_rate": 7.144009853786826e-08,
"loss": 0.6591,
"step": 1174
},
{
"epoch": 0.9631579284419768,
"grad_norm": 1.0755682068772612,
"learning_rate": 6.830384149478009e-08,
"loss": 0.6517,
"step": 1175
},
{
"epoch": 0.9639776373172465,
"grad_norm": 0.8139973115734406,
"learning_rate": 6.523774419013418e-08,
"loss": 0.6124,
"step": 1176
},
{
"epoch": 0.9647973461925163,
"grad_norm": 1.1415285071293786,
"learning_rate": 6.224182828353243e-08,
"loss": 0.7064,
"step": 1177
},
{
"epoch": 0.9656170550677861,
"grad_norm": 0.9120707624115388,
"learning_rate": 5.9316114938801076e-08,
"loss": 0.6801,
"step": 1178
},
{
"epoch": 0.9664367639430559,
"grad_norm": 0.7944323352663182,
"learning_rate": 5.6460624823836405e-08,
"loss": 0.659,
"step": 1179
},
{
"epoch": 0.9672564728183256,
"grad_norm": 1.0033435493336276,
"learning_rate": 5.367537811046486e-08,
"loss": 0.7032,
"step": 1180
},
{
"epoch": 0.9680761816935954,
"grad_norm": 0.9741049522935634,
"learning_rate": 5.096039447429535e-08,
"loss": 0.6953,
"step": 1181
},
{
"epoch": 0.9688958905688652,
"grad_norm": 0.9586762263507326,
"learning_rate": 4.8315693094584945e-08,
"loss": 0.636,
"step": 1182
},
{
"epoch": 0.9697155994441349,
"grad_norm": 0.8332538279492082,
"learning_rate": 4.57412926541001e-08,
"loss": 0.6561,
"step": 1183
},
{
"epoch": 0.9705353083194047,
"grad_norm": 0.8659183413630247,
"learning_rate": 4.3237211338983396e-08,
"loss": 0.6886,
"step": 1184
},
{
"epoch": 0.9713550171946744,
"grad_norm": 0.8502833327273421,
"learning_rate": 4.0803466838631454e-08,
"loss": 0.6514,
"step": 1185
},
{
"epoch": 0.9721747260699443,
"grad_norm": 0.8801268259936795,
"learning_rate": 3.8440076345561685e-08,
"loss": 0.6989,
"step": 1186
},
{
"epoch": 0.972994434945214,
"grad_norm": 0.9815266465485792,
"learning_rate": 3.6147056555296825e-08,
"loss": 0.6312,
"step": 1187
},
{
"epoch": 0.9738141438204837,
"grad_norm": 0.8195201150649793,
"learning_rate": 3.392442366624615e-08,
"loss": 0.6643,
"step": 1188
},
{
"epoch": 0.9746338526957535,
"grad_norm": 0.9884157193267219,
"learning_rate": 3.177219337958892e-08,
"loss": 0.6655,
"step": 1189
},
{
"epoch": 0.9754535615710233,
"grad_norm": 0.9615340054851734,
"learning_rate": 2.969038089916443e-08,
"loss": 0.6555,
"step": 1190
},
{
"epoch": 0.9762732704462931,
"grad_norm": 0.8830624624788647,
"learning_rate": 2.767900093136544e-08,
"loss": 0.678,
"step": 1191
},
{
"epoch": 0.9770929793215628,
"grad_norm": 0.9162151154244101,
"learning_rate": 2.573806768503162e-08,
"loss": 0.6502,
"step": 1192
},
{
"epoch": 0.9779126881968326,
"grad_norm": 0.9115645921281903,
"learning_rate": 2.386759487135293e-08,
"loss": 0.678,
"step": 1193
},
{
"epoch": 0.9787323970721024,
"grad_norm": 0.8965745626021584,
"learning_rate": 2.206759570377193e-08,
"loss": 0.6952,
"step": 1194
},
{
"epoch": 0.9795521059473721,
"grad_norm": 1.0136408679708961,
"learning_rate": 2.033808289788608e-08,
"loss": 0.6901,
"step": 1195
},
{
"epoch": 0.9803718148226419,
"grad_norm": 0.90284874193244,
"learning_rate": 1.8679068671364488e-08,
"loss": 0.6645,
"step": 1196
},
{
"epoch": 0.9811915236979116,
"grad_norm": 0.8809877723003451,
"learning_rate": 1.7090564743857952e-08,
"loss": 0.6952,
"step": 1197
},
{
"epoch": 0.9820112325731815,
"grad_norm": 1.075113564772916,
"learning_rate": 1.557258233691572e-08,
"loss": 0.6406,
"step": 1198
},
{
"epoch": 0.9828309414484512,
"grad_norm": 1.0505057573509975,
"learning_rate": 1.4125132173905542e-08,
"loss": 0.6584,
"step": 1199
},
{
"epoch": 0.983650650323721,
"grad_norm": 0.9583531732138156,
"learning_rate": 1.2748224479943727e-08,
"loss": 0.698,
"step": 1200
},
{
"epoch": 0.9844703591989907,
"grad_norm": 0.9538313348927708,
"learning_rate": 1.1441868981815207e-08,
"loss": 0.6647,
"step": 1201
},
{
"epoch": 0.9852900680742605,
"grad_norm": 0.9359924453677338,
"learning_rate": 1.0206074907909147e-08,
"loss": 0.669,
"step": 1202
},
{
"epoch": 0.9861097769495303,
"grad_norm": 0.9384542126846807,
"learning_rate": 9.040850988153437e-09,
"loss": 0.6736,
"step": 1203
},
{
"epoch": 0.9869294858248,
"grad_norm": 0.8361701993893207,
"learning_rate": 7.946205453953638e-09,
"loss": 0.6179,
"step": 1204
},
{
"epoch": 0.9877491947000698,
"grad_norm": 0.8020568763140312,
"learning_rate": 6.9221460381296845e-09,
"loss": 0.6434,
"step": 1205
},
{
"epoch": 0.9885689035753396,
"grad_norm": 0.8800542256020114,
"learning_rate": 5.968679974870384e-09,
"loss": 0.6559,
"step": 1206
},
{
"epoch": 0.9893886124506094,
"grad_norm": 0.7570994326751808,
"learning_rate": 5.08581399967345e-09,
"loss": 0.6639,
"step": 1207
},
{
"epoch": 0.9902083213258791,
"grad_norm": 0.8163023726380116,
"learning_rate": 4.273554349303321e-09,
"loss": 0.6327,
"step": 1208
},
{
"epoch": 0.9910280302011488,
"grad_norm": 1.070967296195311,
"learning_rate": 3.5319067617445302e-09,
"loss": 0.6434,
"step": 1209
},
{
"epoch": 0.9918477390764187,
"grad_norm": 0.8642716256165196,
"learning_rate": 2.8608764761639542e-09,
"loss": 0.6579,
"step": 1210
},
{
"epoch": 0.9926674479516884,
"grad_norm": 0.8446828834072195,
"learning_rate": 2.2604682328697393e-09,
"loss": 0.6354,
"step": 1211
},
{
"epoch": 0.9934871568269582,
"grad_norm": 1.304113829527474,
"learning_rate": 1.7306862732813234e-09,
"loss": 0.6888,
"step": 1212
},
{
"epoch": 0.9943068657022279,
"grad_norm": 0.9922023852076565,
"learning_rate": 1.2715343398972401e-09,
"loss": 0.6394,
"step": 1213
},
{
"epoch": 0.9951265745774978,
"grad_norm": 0.8883112145392539,
"learning_rate": 8.830156762706932e-10,
"loss": 0.6842,
"step": 1214
},
{
"epoch": 0.9959462834527675,
"grad_norm": 1.0592344407944017,
"learning_rate": 5.651330269840216e-10,
"loss": 0.7027,
"step": 1215
},
{
"epoch": 0.9967659923280372,
"grad_norm": 0.9725664148306387,
"learning_rate": 3.178886376331569e-10,
"loss": 0.6742,
"step": 1216
},
{
"epoch": 0.997585701203307,
"grad_norm": 1.0104687985558842,
"learning_rate": 1.4128425480763874e-10,
"loss": 0.6561,
"step": 1217
},
{
"epoch": 0.9984054100785767,
"grad_norm": 0.8839687606153,
"learning_rate": 3.532112608062299e-11,
"loss": 0.6851,
"step": 1218
},
{
"epoch": 0.9992251189538466,
"grad_norm": 0.8545858765534571,
"learning_rate": 0.0,
"loss": 0.6603,
"step": 1219
},
{
"epoch": 0.9992251189538466,
"step": 1219,
"total_flos": 4672501865250816.0,
"train_loss": 0.7232581392187286,
"train_runtime": 108618.5367,
"train_samples_per_second": 5.75,
"train_steps_per_second": 0.011
}
],
"logging_steps": 1.0,
"max_steps": 1219,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 15,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 4672501865250816.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}