Files
TR_SFT2/last-checkpoint/trainer_state.json
ModelHub XC 4730fab543 初始化项目,由ModelHub XC社区提供模型
Model: Ba2han/TR_SFT2
Source: Original Platform
2026-09-12 09:26:17 +08:00

16847 lines
428 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 959,
"global_step": 4793,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0004173187271778821,
"grad_norm": 0.3835424482822418,
"learning_rate": 5.555555555555556e-06,
"loss": 1.9340534210205078,
"step": 2
},
{
"epoch": 0.0008346374543557642,
"grad_norm": 0.40174242854118347,
"learning_rate": 1.6666666666666667e-05,
"loss": 1.8800839185714722,
"step": 4
},
{
"epoch": 0.0012519561815336462,
"grad_norm": 0.2799811363220215,
"learning_rate": 2.777777777777778e-05,
"loss": 2.0257110595703125,
"step": 6
},
{
"epoch": 0.0016692749087115284,
"grad_norm": 0.26403728127479553,
"learning_rate": 3.888888888888889e-05,
"loss": 1.8546348810195923,
"step": 8
},
{
"epoch": 0.0020865936358894104,
"grad_norm": 0.26313158869743347,
"learning_rate": 5e-05,
"loss": 1.7944978475570679,
"step": 10
},
{
"epoch": 0.0025039123630672924,
"grad_norm": 0.27312523126602173,
"learning_rate": 6.111111111111112e-05,
"loss": 1.9034547805786133,
"step": 12
},
{
"epoch": 0.002921231090245175,
"grad_norm": 0.2977233827114105,
"learning_rate": 7.222222222222222e-05,
"loss": 1.997790813446045,
"step": 14
},
{
"epoch": 0.003338549817423057,
"grad_norm": 0.2744559645652771,
"learning_rate": 8.333333333333334e-05,
"loss": 1.7823611497879028,
"step": 16
},
{
"epoch": 0.003755868544600939,
"grad_norm": 0.3069044351577759,
"learning_rate": 9.444444444444444e-05,
"loss": 1.853757619857788,
"step": 18
},
{
"epoch": 0.004173187271778821,
"grad_norm": 0.34834080934524536,
"learning_rate": 0.00010555555555555557,
"loss": 1.9632055759429932,
"step": 20
},
{
"epoch": 0.004590505998956703,
"grad_norm": 0.3554534912109375,
"learning_rate": 0.00011666666666666668,
"loss": 1.8707588911056519,
"step": 22
},
{
"epoch": 0.005007824726134585,
"grad_norm": 0.2353816032409668,
"learning_rate": 0.00012777777777777776,
"loss": 1.8802332878112793,
"step": 24
},
{
"epoch": 0.005425143453312468,
"grad_norm": 0.3149602711200714,
"learning_rate": 0.0001388888888888889,
"loss": 1.9447239637374878,
"step": 26
},
{
"epoch": 0.00584246218049035,
"grad_norm": 0.2753753364086151,
"learning_rate": 0.00015000000000000001,
"loss": 1.90849769115448,
"step": 28
},
{
"epoch": 0.006259780907668232,
"grad_norm": 0.30161598324775696,
"learning_rate": 0.0001611111111111111,
"loss": 1.9604809284210205,
"step": 30
},
{
"epoch": 0.006677099634846114,
"grad_norm": 0.2720056474208832,
"learning_rate": 0.00017222222222222224,
"loss": 2.0461955070495605,
"step": 32
},
{
"epoch": 0.007094418362023996,
"grad_norm": 0.2841447591781616,
"learning_rate": 0.00018333333333333334,
"loss": 1.8429415225982666,
"step": 34
},
{
"epoch": 0.007511737089201878,
"grad_norm": 0.2209928333759308,
"learning_rate": 0.00019444444444444446,
"loss": 1.9665800333023071,
"step": 36
},
{
"epoch": 0.00792905581637976,
"grad_norm": 0.2683916985988617,
"learning_rate": 0.00020555555555555556,
"loss": 2.065295696258545,
"step": 38
},
{
"epoch": 0.008346374543557642,
"grad_norm": 0.23104991018772125,
"learning_rate": 0.00021666666666666666,
"loss": 1.929124355316162,
"step": 40
},
{
"epoch": 0.008763693270735524,
"grad_norm": 0.24987860023975372,
"learning_rate": 0.00022777777777777778,
"loss": 1.9152966737747192,
"step": 42
},
{
"epoch": 0.009181011997913406,
"grad_norm": 0.2709057927131653,
"learning_rate": 0.0002388888888888889,
"loss": 2.0099782943725586,
"step": 44
},
{
"epoch": 0.009598330725091288,
"grad_norm": 0.28906872868537903,
"learning_rate": 0.00025,
"loss": 1.9668352603912354,
"step": 46
},
{
"epoch": 0.01001564945226917,
"grad_norm": 0.3504061996936798,
"learning_rate": 0.0002611111111111111,
"loss": 1.8883315324783325,
"step": 48
},
{
"epoch": 0.010432968179447054,
"grad_norm": 0.252297043800354,
"learning_rate": 0.00027222222222222226,
"loss": 1.9571173191070557,
"step": 50
},
{
"epoch": 0.010850286906624936,
"grad_norm": 0.2441273182630539,
"learning_rate": 0.00028333333333333335,
"loss": 1.7384397983551025,
"step": 52
},
{
"epoch": 0.011267605633802818,
"grad_norm": 0.278609961271286,
"learning_rate": 0.00029444444444444445,
"loss": 1.9565646648406982,
"step": 54
},
{
"epoch": 0.0116849243609807,
"grad_norm": 0.2239801287651062,
"learning_rate": 0.00030555555555555555,
"loss": 1.9094393253326416,
"step": 56
},
{
"epoch": 0.012102243088158582,
"grad_norm": 0.22790458798408508,
"learning_rate": 0.00031666666666666665,
"loss": 1.8878060579299927,
"step": 58
},
{
"epoch": 0.012519561815336464,
"grad_norm": 0.2285269945859909,
"learning_rate": 0.0003277777777777778,
"loss": 1.8861244916915894,
"step": 60
},
{
"epoch": 0.012936880542514346,
"grad_norm": 0.21995387971401215,
"learning_rate": 0.0003388888888888889,
"loss": 2.029430627822876,
"step": 62
},
{
"epoch": 0.013354199269692228,
"grad_norm": 0.2178620547056198,
"learning_rate": 0.00035,
"loss": 1.8410820960998535,
"step": 64
},
{
"epoch": 0.01377151799687011,
"grad_norm": 0.22449888288974762,
"learning_rate": 0.00036111111111111115,
"loss": 2.011507749557495,
"step": 66
},
{
"epoch": 0.014188836724047992,
"grad_norm": 0.22300758957862854,
"learning_rate": 0.00037222222222222225,
"loss": 1.907748818397522,
"step": 68
},
{
"epoch": 0.014606155451225874,
"grad_norm": 0.21932196617126465,
"learning_rate": 0.00038333333333333334,
"loss": 2.0046677589416504,
"step": 70
},
{
"epoch": 0.015023474178403756,
"grad_norm": 0.2256653755903244,
"learning_rate": 0.0003944444444444445,
"loss": 1.7891045808792114,
"step": 72
},
{
"epoch": 0.015440792905581638,
"grad_norm": 0.1998763084411621,
"learning_rate": 0.00040555555555555554,
"loss": 1.9423624277114868,
"step": 74
},
{
"epoch": 0.01585811163275952,
"grad_norm": 0.2960151433944702,
"learning_rate": 0.0004166666666666667,
"loss": 1.8450512886047363,
"step": 76
},
{
"epoch": 0.016275430359937403,
"grad_norm": 0.2305380254983902,
"learning_rate": 0.0004277777777777778,
"loss": 1.8126254081726074,
"step": 78
},
{
"epoch": 0.016692749087115284,
"grad_norm": 0.21758821606636047,
"learning_rate": 0.00043888888888888894,
"loss": 1.9979026317596436,
"step": 80
},
{
"epoch": 0.017110067814293167,
"grad_norm": 0.20961515605449677,
"learning_rate": 0.00045000000000000004,
"loss": 1.8399300575256348,
"step": 82
},
{
"epoch": 0.017527386541471048,
"grad_norm": 0.26532480120658875,
"learning_rate": 0.0004611111111111111,
"loss": 1.9581152200698853,
"step": 84
},
{
"epoch": 0.01794470526864893,
"grad_norm": 0.21992835402488708,
"learning_rate": 0.00047222222222222224,
"loss": 1.8646459579467773,
"step": 86
},
{
"epoch": 0.01836202399582681,
"grad_norm": 0.2172183096408844,
"learning_rate": 0.00048333333333333334,
"loss": 1.9357714653015137,
"step": 88
},
{
"epoch": 0.018779342723004695,
"grad_norm": 0.23725579679012299,
"learning_rate": 0.0004944444444444445,
"loss": 1.965197205543518,
"step": 90
},
{
"epoch": 0.019196661450182576,
"grad_norm": 0.2102695107460022,
"learning_rate": 0.0005055555555555555,
"loss": 1.9785113334655762,
"step": 92
},
{
"epoch": 0.01961398017736046,
"grad_norm": 0.19648884236812592,
"learning_rate": 0.0005166666666666667,
"loss": 1.8086206912994385,
"step": 94
},
{
"epoch": 0.02003129890453834,
"grad_norm": 0.2493210881948471,
"learning_rate": 0.0005277777777777778,
"loss": 1.9426219463348389,
"step": 96
},
{
"epoch": 0.020448617631716223,
"grad_norm": 0.2268703430891037,
"learning_rate": 0.000538888888888889,
"loss": 1.903491497039795,
"step": 98
},
{
"epoch": 0.020865936358894107,
"grad_norm": 0.23800517618656158,
"learning_rate": 0.00055,
"loss": 1.9606008529663086,
"step": 100
},
{
"epoch": 0.021283255086071987,
"grad_norm": 0.18385547399520874,
"learning_rate": 0.0005611111111111111,
"loss": 2.025949716567993,
"step": 102
},
{
"epoch": 0.02170057381324987,
"grad_norm": 0.1959981471300125,
"learning_rate": 0.0005722222222222222,
"loss": 1.9590791463851929,
"step": 104
},
{
"epoch": 0.02211789254042775,
"grad_norm": 0.1910092681646347,
"learning_rate": 0.0005833333333333334,
"loss": 1.9767764806747437,
"step": 106
},
{
"epoch": 0.022535211267605635,
"grad_norm": 0.21451124548912048,
"learning_rate": 0.0005944444444444445,
"loss": 1.8308137655258179,
"step": 108
},
{
"epoch": 0.022952529994783515,
"grad_norm": 0.241910919547081,
"learning_rate": 0.0006055555555555556,
"loss": 1.7464343309402466,
"step": 110
},
{
"epoch": 0.0233698487219614,
"grad_norm": 0.2224787473678589,
"learning_rate": 0.0006166666666666667,
"loss": 1.8866147994995117,
"step": 112
},
{
"epoch": 0.02378716744913928,
"grad_norm": 0.21395538747310638,
"learning_rate": 0.0006277777777777778,
"loss": 1.930424451828003,
"step": 114
},
{
"epoch": 0.024204486176317163,
"grad_norm": 0.23291410505771637,
"learning_rate": 0.0006388888888888889,
"loss": 1.8475985527038574,
"step": 116
},
{
"epoch": 0.024621804903495043,
"grad_norm": 0.21410924196243286,
"learning_rate": 0.0006500000000000001,
"loss": 1.8692370653152466,
"step": 118
},
{
"epoch": 0.025039123630672927,
"grad_norm": 0.18796320259571075,
"learning_rate": 0.0006611111111111111,
"loss": 1.9754891395568848,
"step": 120
},
{
"epoch": 0.025456442357850807,
"grad_norm": 0.19204045832157135,
"learning_rate": 0.0006722222222222223,
"loss": 1.9037480354309082,
"step": 122
},
{
"epoch": 0.02587376108502869,
"grad_norm": 0.19712218642234802,
"learning_rate": 0.0006833333333333333,
"loss": 1.8721704483032227,
"step": 124
},
{
"epoch": 0.02629107981220657,
"grad_norm": 0.2708342969417572,
"learning_rate": 0.0006944444444444445,
"loss": 1.8033452033996582,
"step": 126
},
{
"epoch": 0.026708398539384455,
"grad_norm": 0.29105356335639954,
"learning_rate": 0.0007055555555555556,
"loss": 1.9020501375198364,
"step": 128
},
{
"epoch": 0.027125717266562335,
"grad_norm": 0.16851791739463806,
"learning_rate": 0.0007166666666666668,
"loss": 1.8632913827896118,
"step": 130
},
{
"epoch": 0.02754303599374022,
"grad_norm": 0.1972474306821823,
"learning_rate": 0.0007277777777777778,
"loss": 1.9148986339569092,
"step": 132
},
{
"epoch": 0.027960354720918103,
"grad_norm": 0.18612337112426758,
"learning_rate": 0.000738888888888889,
"loss": 1.966691493988037,
"step": 134
},
{
"epoch": 0.028377673448095983,
"grad_norm": 0.2851947844028473,
"learning_rate": 0.00075,
"loss": 1.8829073905944824,
"step": 136
},
{
"epoch": 0.028794992175273867,
"grad_norm": 0.2141549289226532,
"learning_rate": 0.0007611111111111111,
"loss": 1.8888792991638184,
"step": 138
},
{
"epoch": 0.029212310902451747,
"grad_norm": 0.18543975055217743,
"learning_rate": 0.0007722222222222223,
"loss": 1.8063397407531738,
"step": 140
},
{
"epoch": 0.02962962962962963,
"grad_norm": 0.19061000645160675,
"learning_rate": 0.0007833333333333334,
"loss": 1.8396409749984741,
"step": 142
},
{
"epoch": 0.03004694835680751,
"grad_norm": 0.19235527515411377,
"learning_rate": 0.0007944444444444445,
"loss": 1.9190809726715088,
"step": 144
},
{
"epoch": 0.030464267083985395,
"grad_norm": 0.1778387725353241,
"learning_rate": 0.0008,
"loss": 2.001985788345337,
"step": 146
},
{
"epoch": 0.030881585811163275,
"grad_norm": 0.2677212953567505,
"learning_rate": 0.0008,
"loss": 1.758802890777588,
"step": 148
},
{
"epoch": 0.03129890453834116,
"grad_norm": 0.18686190247535706,
"learning_rate": 0.0008,
"loss": 1.8034826517105103,
"step": 150
},
{
"epoch": 0.03171622326551904,
"grad_norm": 0.23492851853370667,
"learning_rate": 0.0008,
"loss": 1.9053776264190674,
"step": 152
},
{
"epoch": 0.03213354199269692,
"grad_norm": 0.21396663784980774,
"learning_rate": 0.0008,
"loss": 1.8032176494598389,
"step": 154
},
{
"epoch": 0.032550860719874807,
"grad_norm": 0.25143226981163025,
"learning_rate": 0.0008,
"loss": 1.7828192710876465,
"step": 156
},
{
"epoch": 0.03296817944705269,
"grad_norm": 0.20337209105491638,
"learning_rate": 0.0008,
"loss": 1.8441500663757324,
"step": 158
},
{
"epoch": 0.03338549817423057,
"grad_norm": 0.19828999042510986,
"learning_rate": 0.0008,
"loss": 1.8969037532806396,
"step": 160
},
{
"epoch": 0.03380281690140845,
"grad_norm": 0.1736728549003601,
"learning_rate": 0.0008,
"loss": 1.7638834714889526,
"step": 162
},
{
"epoch": 0.034220135628586335,
"grad_norm": 0.22760280966758728,
"learning_rate": 0.0008,
"loss": 1.883575201034546,
"step": 164
},
{
"epoch": 0.034637454355764215,
"grad_norm": 0.1854819804430008,
"learning_rate": 0.0008,
"loss": 1.6732118129730225,
"step": 166
},
{
"epoch": 0.035054773082942095,
"grad_norm": 0.21865177154541016,
"learning_rate": 0.0008,
"loss": 1.91782808303833,
"step": 168
},
{
"epoch": 0.03547209181011998,
"grad_norm": 0.19005519151687622,
"learning_rate": 0.0008,
"loss": 1.7810168266296387,
"step": 170
},
{
"epoch": 0.03588941053729786,
"grad_norm": 0.20469321310520172,
"learning_rate": 0.0008,
"loss": 1.9270379543304443,
"step": 172
},
{
"epoch": 0.03630672926447574,
"grad_norm": 0.2073751538991928,
"learning_rate": 0.0008,
"loss": 1.7687180042266846,
"step": 174
},
{
"epoch": 0.03672404799165362,
"grad_norm": 0.18631917238235474,
"learning_rate": 0.0008,
"loss": 1.9677690267562866,
"step": 176
},
{
"epoch": 0.03714136671883151,
"grad_norm": 0.22901347279548645,
"learning_rate": 0.0008,
"loss": 2.005549669265747,
"step": 178
},
{
"epoch": 0.03755868544600939,
"grad_norm": 0.18291981518268585,
"learning_rate": 0.0008,
"loss": 1.8533040285110474,
"step": 180
},
{
"epoch": 0.03797600417318727,
"grad_norm": 0.20379191637039185,
"learning_rate": 0.0008,
"loss": 1.8690353631973267,
"step": 182
},
{
"epoch": 0.03839332290036515,
"grad_norm": 0.21121641993522644,
"learning_rate": 0.0008,
"loss": 1.8854107856750488,
"step": 184
},
{
"epoch": 0.03881064162754304,
"grad_norm": 0.21457895636558533,
"learning_rate": 0.0008,
"loss": 1.9446922540664673,
"step": 186
},
{
"epoch": 0.03922796035472092,
"grad_norm": 0.1923578381538391,
"learning_rate": 0.0008,
"loss": 1.7910972833633423,
"step": 188
},
{
"epoch": 0.0396452790818988,
"grad_norm": 0.20732815563678741,
"learning_rate": 0.0008,
"loss": 2.0059916973114014,
"step": 190
},
{
"epoch": 0.04006259780907668,
"grad_norm": 0.2116197943687439,
"learning_rate": 0.0008,
"loss": 1.9082348346710205,
"step": 192
},
{
"epoch": 0.040479916536254566,
"grad_norm": 0.18705199658870697,
"learning_rate": 0.0008,
"loss": 1.860048532485962,
"step": 194
},
{
"epoch": 0.040897235263432447,
"grad_norm": 0.25623950362205505,
"learning_rate": 0.0008,
"loss": 1.843970537185669,
"step": 196
},
{
"epoch": 0.04131455399061033,
"grad_norm": 0.20779113471508026,
"learning_rate": 0.0008,
"loss": 1.8779997825622559,
"step": 198
},
{
"epoch": 0.041731872717788214,
"grad_norm": 0.19002337753772736,
"learning_rate": 0.0008,
"loss": 1.9125951528549194,
"step": 200
},
{
"epoch": 0.042149191444966094,
"grad_norm": 0.19544892013072968,
"learning_rate": 0.0008,
"loss": 1.995476484298706,
"step": 202
},
{
"epoch": 0.042566510172143975,
"grad_norm": 0.18213562667369843,
"learning_rate": 0.0008,
"loss": 1.8816639184951782,
"step": 204
},
{
"epoch": 0.042983828899321855,
"grad_norm": 0.22063912451267242,
"learning_rate": 0.0008,
"loss": 1.803364634513855,
"step": 206
},
{
"epoch": 0.04340114762649974,
"grad_norm": 0.20791158080101013,
"learning_rate": 0.0008,
"loss": 1.9003090858459473,
"step": 208
},
{
"epoch": 0.04381846635367762,
"grad_norm": 0.16584512591362,
"learning_rate": 0.0008,
"loss": 1.8435355424880981,
"step": 210
},
{
"epoch": 0.0442357850808555,
"grad_norm": 0.19649940729141235,
"learning_rate": 0.0008,
"loss": 1.80889093875885,
"step": 212
},
{
"epoch": 0.04465310380803338,
"grad_norm": 0.21633127331733704,
"learning_rate": 0.0008,
"loss": 1.6867175102233887,
"step": 214
},
{
"epoch": 0.04507042253521127,
"grad_norm": 0.19792751967906952,
"learning_rate": 0.0008,
"loss": 1.8823671340942383,
"step": 216
},
{
"epoch": 0.04548774126238915,
"grad_norm": 0.1862718164920807,
"learning_rate": 0.0008,
"loss": 1.8919916152954102,
"step": 218
},
{
"epoch": 0.04590505998956703,
"grad_norm": 0.21791967749595642,
"learning_rate": 0.0008,
"loss": 1.9127482175827026,
"step": 220
},
{
"epoch": 0.04632237871674491,
"grad_norm": 0.19531375169754028,
"learning_rate": 0.0008,
"loss": 1.8581024408340454,
"step": 222
},
{
"epoch": 0.0467396974439228,
"grad_norm": 0.20277369022369385,
"learning_rate": 0.0008,
"loss": 1.9121865034103394,
"step": 224
},
{
"epoch": 0.04715701617110068,
"grad_norm": 0.19659613072872162,
"learning_rate": 0.0008,
"loss": 1.789233922958374,
"step": 226
},
{
"epoch": 0.04757433489827856,
"grad_norm": 0.22985541820526123,
"learning_rate": 0.0008,
"loss": 1.9352688789367676,
"step": 228
},
{
"epoch": 0.047991653625456446,
"grad_norm": 0.18611934781074524,
"learning_rate": 0.0008,
"loss": 1.7723150253295898,
"step": 230
},
{
"epoch": 0.048408972352634326,
"grad_norm": 0.19027550518512726,
"learning_rate": 0.0008,
"loss": 1.8290672302246094,
"step": 232
},
{
"epoch": 0.048826291079812206,
"grad_norm": 0.20396903157234192,
"learning_rate": 0.0008,
"loss": 1.832564115524292,
"step": 234
},
{
"epoch": 0.04924360980699009,
"grad_norm": 0.17874160408973694,
"learning_rate": 0.0008,
"loss": 1.814352035522461,
"step": 236
},
{
"epoch": 0.049660928534167974,
"grad_norm": 0.24529126286506653,
"learning_rate": 0.0008,
"loss": 1.922191858291626,
"step": 238
},
{
"epoch": 0.050078247261345854,
"grad_norm": 0.1938621997833252,
"learning_rate": 0.0008,
"loss": 1.8524785041809082,
"step": 240
},
{
"epoch": 0.050495565988523734,
"grad_norm": 0.19732561707496643,
"learning_rate": 0.0008,
"loss": 1.9085856676101685,
"step": 242
},
{
"epoch": 0.050912884715701615,
"grad_norm": 0.18855291604995728,
"learning_rate": 0.0008,
"loss": 1.8275926113128662,
"step": 244
},
{
"epoch": 0.0513302034428795,
"grad_norm": 0.20376136898994446,
"learning_rate": 0.0008,
"loss": 1.7887071371078491,
"step": 246
},
{
"epoch": 0.05174752217005738,
"grad_norm": 0.22787004709243774,
"learning_rate": 0.0008,
"loss": 1.8560370206832886,
"step": 248
},
{
"epoch": 0.05216484089723526,
"grad_norm": 0.17018505930900574,
"learning_rate": 0.0008,
"loss": 1.9317115545272827,
"step": 250
},
{
"epoch": 0.05258215962441314,
"grad_norm": 0.1912796050310135,
"learning_rate": 0.0008,
"loss": 1.895258903503418,
"step": 252
},
{
"epoch": 0.05299947835159103,
"grad_norm": 0.17850643396377563,
"learning_rate": 0.0008,
"loss": 1.9079219102859497,
"step": 254
},
{
"epoch": 0.05341679707876891,
"grad_norm": 0.2072039246559143,
"learning_rate": 0.0008,
"loss": 1.89905846118927,
"step": 256
},
{
"epoch": 0.05383411580594679,
"grad_norm": 0.19988320767879486,
"learning_rate": 0.0008,
"loss": 1.7947583198547363,
"step": 258
},
{
"epoch": 0.05425143453312467,
"grad_norm": 0.18004950881004333,
"learning_rate": 0.0008,
"loss": 1.7849770784378052,
"step": 260
},
{
"epoch": 0.05466875326030256,
"grad_norm": 0.20455290377140045,
"learning_rate": 0.0008,
"loss": 1.853940486907959,
"step": 262
},
{
"epoch": 0.05508607198748044,
"grad_norm": 0.18449728190898895,
"learning_rate": 0.0008,
"loss": 1.7918925285339355,
"step": 264
},
{
"epoch": 0.05550339071465832,
"grad_norm": 0.18341559171676636,
"learning_rate": 0.0008,
"loss": 1.9166698455810547,
"step": 266
},
{
"epoch": 0.055920709441836206,
"grad_norm": 0.18132027983665466,
"learning_rate": 0.0008,
"loss": 1.7890571355819702,
"step": 268
},
{
"epoch": 0.056338028169014086,
"grad_norm": 0.18553175032138824,
"learning_rate": 0.0008,
"loss": 1.8641438484191895,
"step": 270
},
{
"epoch": 0.056755346896191966,
"grad_norm": 0.22429780662059784,
"learning_rate": 0.0008,
"loss": 1.9701159000396729,
"step": 272
},
{
"epoch": 0.057172665623369846,
"grad_norm": 0.1992367058992386,
"learning_rate": 0.0008,
"loss": 2.006492853164673,
"step": 274
},
{
"epoch": 0.057589984350547734,
"grad_norm": 0.19233880937099457,
"learning_rate": 0.0008,
"loss": 1.86127507686615,
"step": 276
},
{
"epoch": 0.058007303077725614,
"grad_norm": 0.20732514560222626,
"learning_rate": 0.0008,
"loss": 1.807320237159729,
"step": 278
},
{
"epoch": 0.058424621804903494,
"grad_norm": 0.1792643517255783,
"learning_rate": 0.0008,
"loss": 1.844096302986145,
"step": 280
},
{
"epoch": 0.058841940532081374,
"grad_norm": 0.192881241440773,
"learning_rate": 0.0008,
"loss": 1.9285340309143066,
"step": 282
},
{
"epoch": 0.05925925925925926,
"grad_norm": 0.2025625854730606,
"learning_rate": 0.0008,
"loss": 1.8632502555847168,
"step": 284
},
{
"epoch": 0.05967657798643714,
"grad_norm": 0.20893600583076477,
"learning_rate": 0.0008,
"loss": 1.81607186794281,
"step": 286
},
{
"epoch": 0.06009389671361502,
"grad_norm": 0.18457506597042084,
"learning_rate": 0.0008,
"loss": 1.7620784044265747,
"step": 288
},
{
"epoch": 0.0605112154407929,
"grad_norm": 0.18458612263202667,
"learning_rate": 0.0008,
"loss": 1.8516631126403809,
"step": 290
},
{
"epoch": 0.06092853416797079,
"grad_norm": 0.20481403172016144,
"learning_rate": 0.0008,
"loss": 1.7973397970199585,
"step": 292
},
{
"epoch": 0.06134585289514867,
"grad_norm": 0.2392035722732544,
"learning_rate": 0.0008,
"loss": 1.6744558811187744,
"step": 294
},
{
"epoch": 0.06176317162232655,
"grad_norm": 0.1839018315076828,
"learning_rate": 0.0008,
"loss": 1.859205722808838,
"step": 296
},
{
"epoch": 0.06218049034950444,
"grad_norm": 0.18390414118766785,
"learning_rate": 0.0008,
"loss": 1.8454444408416748,
"step": 298
},
{
"epoch": 0.06259780907668232,
"grad_norm": 0.23476417362689972,
"learning_rate": 0.0008,
"loss": 1.8064424991607666,
"step": 300
},
{
"epoch": 0.0630151278038602,
"grad_norm": 0.23200395703315735,
"learning_rate": 0.0008,
"loss": 1.8403925895690918,
"step": 302
},
{
"epoch": 0.06343244653103808,
"grad_norm": 0.18426549434661865,
"learning_rate": 0.0008,
"loss": 1.8290090560913086,
"step": 304
},
{
"epoch": 0.06384976525821597,
"grad_norm": 0.21662050485610962,
"learning_rate": 0.0008,
"loss": 1.8032827377319336,
"step": 306
},
{
"epoch": 0.06426708398539384,
"grad_norm": 0.20660407841205597,
"learning_rate": 0.0008,
"loss": 1.9283571243286133,
"step": 308
},
{
"epoch": 0.06468440271257173,
"grad_norm": 0.19536849856376648,
"learning_rate": 0.0008,
"loss": 1.8072845935821533,
"step": 310
},
{
"epoch": 0.06510172143974961,
"grad_norm": 0.2134157121181488,
"learning_rate": 0.0008,
"loss": 1.8819622993469238,
"step": 312
},
{
"epoch": 0.06551904016692749,
"grad_norm": 0.17046234011650085,
"learning_rate": 0.0008,
"loss": 1.7353384494781494,
"step": 314
},
{
"epoch": 0.06593635889410537,
"grad_norm": 0.22714658081531525,
"learning_rate": 0.0008,
"loss": 1.652437686920166,
"step": 316
},
{
"epoch": 0.06635367762128326,
"grad_norm": 0.19747577607631683,
"learning_rate": 0.0008,
"loss": 1.8750182390213013,
"step": 318
},
{
"epoch": 0.06677099634846113,
"grad_norm": 0.18324685096740723,
"learning_rate": 0.0008,
"loss": 1.8665786981582642,
"step": 320
},
{
"epoch": 0.06718831507563902,
"grad_norm": 0.20055170357227325,
"learning_rate": 0.0008,
"loss": 1.8537718057632446,
"step": 322
},
{
"epoch": 0.0676056338028169,
"grad_norm": 0.18057942390441895,
"learning_rate": 0.0008,
"loss": 1.8148211240768433,
"step": 324
},
{
"epoch": 0.06802295252999478,
"grad_norm": 0.19412660598754883,
"learning_rate": 0.0008,
"loss": 1.7722946405410767,
"step": 326
},
{
"epoch": 0.06844027125717267,
"grad_norm": 0.2082413285970688,
"learning_rate": 0.0008,
"loss": 1.8280037641525269,
"step": 328
},
{
"epoch": 0.06885758998435054,
"grad_norm": 0.20311059057712555,
"learning_rate": 0.0008,
"loss": 1.8522393703460693,
"step": 330
},
{
"epoch": 0.06927490871152843,
"grad_norm": 0.1785109043121338,
"learning_rate": 0.0008,
"loss": 1.7460867166519165,
"step": 332
},
{
"epoch": 0.06969222743870632,
"grad_norm": 0.19410011172294617,
"learning_rate": 0.0008,
"loss": 1.940168857574463,
"step": 334
},
{
"epoch": 0.07010954616588419,
"grad_norm": 0.17364338040351868,
"learning_rate": 0.0008,
"loss": 1.7872140407562256,
"step": 336
},
{
"epoch": 0.07052686489306208,
"grad_norm": 0.18756942451000214,
"learning_rate": 0.0008,
"loss": 1.787614345550537,
"step": 338
},
{
"epoch": 0.07094418362023996,
"grad_norm": 0.18927060067653656,
"learning_rate": 0.0008,
"loss": 1.773796558380127,
"step": 340
},
{
"epoch": 0.07136150234741784,
"grad_norm": 0.19137617945671082,
"learning_rate": 0.0008,
"loss": 1.8877743482589722,
"step": 342
},
{
"epoch": 0.07177882107459573,
"grad_norm": 0.2876986861228943,
"learning_rate": 0.0008,
"loss": 1.8008520603179932,
"step": 344
},
{
"epoch": 0.0721961398017736,
"grad_norm": 0.2002546340227127,
"learning_rate": 0.0008,
"loss": 1.8457027673721313,
"step": 346
},
{
"epoch": 0.07261345852895149,
"grad_norm": 0.19405768811702728,
"learning_rate": 0.0008,
"loss": 1.8778666257858276,
"step": 348
},
{
"epoch": 0.07303077725612937,
"grad_norm": 0.17879590392112732,
"learning_rate": 0.0008,
"loss": 1.9125796556472778,
"step": 350
},
{
"epoch": 0.07344809598330725,
"grad_norm": 0.2296588122844696,
"learning_rate": 0.0008,
"loss": 1.84610915184021,
"step": 352
},
{
"epoch": 0.07386541471048513,
"grad_norm": 0.20345892012119293,
"learning_rate": 0.0008,
"loss": 1.859510898590088,
"step": 354
},
{
"epoch": 0.07428273343766302,
"grad_norm": 0.16952411830425262,
"learning_rate": 0.0008,
"loss": 1.9320992231369019,
"step": 356
},
{
"epoch": 0.0747000521648409,
"grad_norm": 0.16383187472820282,
"learning_rate": 0.0008,
"loss": 1.862929344177246,
"step": 358
},
{
"epoch": 0.07511737089201878,
"grad_norm": 0.16575641930103302,
"learning_rate": 0.0008,
"loss": 1.7022333145141602,
"step": 360
},
{
"epoch": 0.07553468961919667,
"grad_norm": 0.17977960407733917,
"learning_rate": 0.0008,
"loss": 1.7043992280960083,
"step": 362
},
{
"epoch": 0.07595200834637454,
"grad_norm": 0.1884085237979889,
"learning_rate": 0.0008,
"loss": 1.729490876197815,
"step": 364
},
{
"epoch": 0.07636932707355243,
"grad_norm": 0.19425225257873535,
"learning_rate": 0.0008,
"loss": 1.9061787128448486,
"step": 366
},
{
"epoch": 0.0767866458007303,
"grad_norm": 0.21979810297489166,
"learning_rate": 0.0008,
"loss": 1.902428388595581,
"step": 368
},
{
"epoch": 0.07720396452790819,
"grad_norm": 0.19381879270076752,
"learning_rate": 0.0008,
"loss": 1.699400544166565,
"step": 370
},
{
"epoch": 0.07762128325508608,
"grad_norm": 0.16549688577651978,
"learning_rate": 0.0008,
"loss": 1.817840814590454,
"step": 372
},
{
"epoch": 0.07803860198226395,
"grad_norm": 0.16603948175907135,
"learning_rate": 0.0008,
"loss": 1.7617093324661255,
"step": 374
},
{
"epoch": 0.07845592070944184,
"grad_norm": 0.17938029766082764,
"learning_rate": 0.0008,
"loss": 1.7205898761749268,
"step": 376
},
{
"epoch": 0.07887323943661972,
"grad_norm": 0.18432219326496124,
"learning_rate": 0.0008,
"loss": 1.9758812189102173,
"step": 378
},
{
"epoch": 0.0792905581637976,
"grad_norm": 0.192180335521698,
"learning_rate": 0.0008,
"loss": 1.7806042432785034,
"step": 380
},
{
"epoch": 0.07970787689097548,
"grad_norm": 0.18882744014263153,
"learning_rate": 0.0008,
"loss": 1.8606830835342407,
"step": 382
},
{
"epoch": 0.08012519561815336,
"grad_norm": 0.20159609615802765,
"learning_rate": 0.0008,
"loss": 1.8759417533874512,
"step": 384
},
{
"epoch": 0.08054251434533125,
"grad_norm": 0.18802420794963837,
"learning_rate": 0.0008,
"loss": 1.7845672369003296,
"step": 386
},
{
"epoch": 0.08095983307250913,
"grad_norm": 0.19080901145935059,
"learning_rate": 0.0008,
"loss": 1.8926739692687988,
"step": 388
},
{
"epoch": 0.081377151799687,
"grad_norm": 0.19579608738422394,
"learning_rate": 0.0008,
"loss": 1.7976164817810059,
"step": 390
},
{
"epoch": 0.08179447052686489,
"grad_norm": 0.18886063992977142,
"learning_rate": 0.0008,
"loss": 1.8368678092956543,
"step": 392
},
{
"epoch": 0.08221178925404278,
"grad_norm": 0.190400168299675,
"learning_rate": 0.0008,
"loss": 1.951639175415039,
"step": 394
},
{
"epoch": 0.08262910798122065,
"grad_norm": 0.2170623242855072,
"learning_rate": 0.0008,
"loss": 1.8079655170440674,
"step": 396
},
{
"epoch": 0.08304642670839854,
"grad_norm": 0.1749616265296936,
"learning_rate": 0.0008,
"loss": 1.8746510744094849,
"step": 398
},
{
"epoch": 0.08346374543557643,
"grad_norm": 0.20247295498847961,
"learning_rate": 0.0008,
"loss": 1.841644525527954,
"step": 400
},
{
"epoch": 0.0838810641627543,
"grad_norm": 0.2169034481048584,
"learning_rate": 0.0008,
"loss": 1.509040355682373,
"step": 402
},
{
"epoch": 0.08429838288993219,
"grad_norm": 0.18721264600753784,
"learning_rate": 0.0008,
"loss": 1.6984519958496094,
"step": 404
},
{
"epoch": 0.08471570161711006,
"grad_norm": 0.20038065314292908,
"learning_rate": 0.0008,
"loss": 1.8178945779800415,
"step": 406
},
{
"epoch": 0.08513302034428795,
"grad_norm": 0.20621061325073242,
"learning_rate": 0.0008,
"loss": 1.8274602890014648,
"step": 408
},
{
"epoch": 0.08555033907146584,
"grad_norm": 0.21223652362823486,
"learning_rate": 0.0008,
"loss": 1.7470018863677979,
"step": 410
},
{
"epoch": 0.08596765779864371,
"grad_norm": 0.17984609305858612,
"learning_rate": 0.0008,
"loss": 1.887202501296997,
"step": 412
},
{
"epoch": 0.0863849765258216,
"grad_norm": 0.18457219004631042,
"learning_rate": 0.0008,
"loss": 1.7406067848205566,
"step": 414
},
{
"epoch": 0.08680229525299948,
"grad_norm": 0.1806093007326126,
"learning_rate": 0.0008,
"loss": 1.8982247114181519,
"step": 416
},
{
"epoch": 0.08721961398017736,
"grad_norm": 0.21698443591594696,
"learning_rate": 0.0008,
"loss": 1.8538949489593506,
"step": 418
},
{
"epoch": 0.08763693270735524,
"grad_norm": 0.18613983690738678,
"learning_rate": 0.0008,
"loss": 1.8653966188430786,
"step": 420
},
{
"epoch": 0.08805425143453312,
"grad_norm": 0.1829523891210556,
"learning_rate": 0.0008,
"loss": 1.7498433589935303,
"step": 422
},
{
"epoch": 0.088471570161711,
"grad_norm": 0.18519410490989685,
"learning_rate": 0.0008,
"loss": 1.792525291442871,
"step": 424
},
{
"epoch": 0.08888888888888889,
"grad_norm": 0.1749446839094162,
"learning_rate": 0.0008,
"loss": 1.9061282873153687,
"step": 426
},
{
"epoch": 0.08930620761606677,
"grad_norm": 0.16339826583862305,
"learning_rate": 0.0008,
"loss": 1.7451319694519043,
"step": 428
},
{
"epoch": 0.08972352634324465,
"grad_norm": 0.17513766884803772,
"learning_rate": 0.0008,
"loss": 1.8854788541793823,
"step": 430
},
{
"epoch": 0.09014084507042254,
"grad_norm": 0.184746652841568,
"learning_rate": 0.0008,
"loss": 1.7705706357955933,
"step": 432
},
{
"epoch": 0.09055816379760041,
"grad_norm": 0.1971534788608551,
"learning_rate": 0.0008,
"loss": 1.7745380401611328,
"step": 434
},
{
"epoch": 0.0909754825247783,
"grad_norm": 0.19071540236473083,
"learning_rate": 0.0008,
"loss": 1.7405657768249512,
"step": 436
},
{
"epoch": 0.09139280125195619,
"grad_norm": 0.17713378369808197,
"learning_rate": 0.0008,
"loss": 1.8192437887191772,
"step": 438
},
{
"epoch": 0.09181011997913406,
"grad_norm": 0.18660472333431244,
"learning_rate": 0.0008,
"loss": 1.5884759426116943,
"step": 440
},
{
"epoch": 0.09222743870631195,
"grad_norm": 0.2162025421857834,
"learning_rate": 0.0008,
"loss": 1.7200651168823242,
"step": 442
},
{
"epoch": 0.09264475743348982,
"grad_norm": 0.17986829578876495,
"learning_rate": 0.0008,
"loss": 1.8388999700546265,
"step": 444
},
{
"epoch": 0.09306207616066771,
"grad_norm": 0.16443675756454468,
"learning_rate": 0.0008,
"loss": 1.8921184539794922,
"step": 446
},
{
"epoch": 0.0934793948878456,
"grad_norm": 0.19607263803482056,
"learning_rate": 0.0008,
"loss": 1.6894744634628296,
"step": 448
},
{
"epoch": 0.09389671361502347,
"grad_norm": 0.19868682324886322,
"learning_rate": 0.0008,
"loss": 1.8331403732299805,
"step": 450
},
{
"epoch": 0.09431403234220136,
"grad_norm": 0.1826103925704956,
"learning_rate": 0.0008,
"loss": 1.7721960544586182,
"step": 452
},
{
"epoch": 0.09473135106937924,
"grad_norm": 0.20188207924365997,
"learning_rate": 0.0008,
"loss": 1.7395975589752197,
"step": 454
},
{
"epoch": 0.09514866979655712,
"grad_norm": 0.1857413500547409,
"learning_rate": 0.0008,
"loss": 1.5533616542816162,
"step": 456
},
{
"epoch": 0.095565988523735,
"grad_norm": 0.2413434088230133,
"learning_rate": 0.0008,
"loss": 1.8269988298416138,
"step": 458
},
{
"epoch": 0.09598330725091289,
"grad_norm": 0.18487891554832458,
"learning_rate": 0.0008,
"loss": 1.820391297340393,
"step": 460
},
{
"epoch": 0.09640062597809076,
"grad_norm": 0.18426169455051422,
"learning_rate": 0.0008,
"loss": 1.8029463291168213,
"step": 462
},
{
"epoch": 0.09681794470526865,
"grad_norm": 0.20566286146640778,
"learning_rate": 0.0008,
"loss": 1.850946307182312,
"step": 464
},
{
"epoch": 0.09723526343244653,
"grad_norm": 0.18228569626808167,
"learning_rate": 0.0008,
"loss": 1.8585747480392456,
"step": 466
},
{
"epoch": 0.09765258215962441,
"grad_norm": 0.1945704072713852,
"learning_rate": 0.0008,
"loss": 1.7613170146942139,
"step": 468
},
{
"epoch": 0.0980699008868023,
"grad_norm": 0.21596919000148773,
"learning_rate": 0.0008,
"loss": 1.8058515787124634,
"step": 470
},
{
"epoch": 0.09848721961398017,
"grad_norm": 0.20458221435546875,
"learning_rate": 0.0008,
"loss": 1.888073205947876,
"step": 472
},
{
"epoch": 0.09890453834115806,
"grad_norm": 0.16673769056797028,
"learning_rate": 0.0008,
"loss": 1.7520596981048584,
"step": 474
},
{
"epoch": 0.09932185706833595,
"grad_norm": 0.25613194704055786,
"learning_rate": 0.0008,
"loss": 1.7988296747207642,
"step": 476
},
{
"epoch": 0.09973917579551382,
"grad_norm": 0.19962145388126373,
"learning_rate": 0.0008,
"loss": 1.770625352859497,
"step": 478
},
{
"epoch": 0.10015649452269171,
"grad_norm": 0.23697935044765472,
"learning_rate": 0.0008,
"loss": 1.6397219896316528,
"step": 480
},
{
"epoch": 0.10057381324986958,
"grad_norm": 0.19623948633670807,
"learning_rate": 0.0008,
"loss": 1.764510154724121,
"step": 482
},
{
"epoch": 0.10099113197704747,
"grad_norm": 0.16900749504566193,
"learning_rate": 0.0008,
"loss": 1.8732715845108032,
"step": 484
},
{
"epoch": 0.10140845070422536,
"grad_norm": 0.1668040007352829,
"learning_rate": 0.0008,
"loss": 1.8405311107635498,
"step": 486
},
{
"epoch": 0.10182576943140323,
"grad_norm": 0.2098231166601181,
"learning_rate": 0.0008,
"loss": 1.7863925695419312,
"step": 488
},
{
"epoch": 0.10224308815858112,
"grad_norm": 0.19985970854759216,
"learning_rate": 0.0008,
"loss": 1.722977876663208,
"step": 490
},
{
"epoch": 0.102660406885759,
"grad_norm": 0.21394382417201996,
"learning_rate": 0.0008,
"loss": 1.757413387298584,
"step": 492
},
{
"epoch": 0.10307772561293688,
"grad_norm": 0.18714119493961334,
"learning_rate": 0.0008,
"loss": 1.8428552150726318,
"step": 494
},
{
"epoch": 0.10349504434011476,
"grad_norm": 0.19619259238243103,
"learning_rate": 0.0008,
"loss": 1.732613205909729,
"step": 496
},
{
"epoch": 0.10391236306729265,
"grad_norm": 0.1874673068523407,
"learning_rate": 0.0008,
"loss": 1.9147651195526123,
"step": 498
},
{
"epoch": 0.10432968179447052,
"grad_norm": 0.1815664917230606,
"learning_rate": 0.0008,
"loss": 1.887975811958313,
"step": 500
},
{
"epoch": 0.10474700052164841,
"grad_norm": 0.17395725846290588,
"learning_rate": 0.0008,
"loss": 1.7720164060592651,
"step": 502
},
{
"epoch": 0.10516431924882629,
"grad_norm": 0.19323813915252686,
"learning_rate": 0.0008,
"loss": 1.7940523624420166,
"step": 504
},
{
"epoch": 0.10558163797600417,
"grad_norm": 0.21045640110969543,
"learning_rate": 0.0008,
"loss": 1.7197074890136719,
"step": 506
},
{
"epoch": 0.10599895670318206,
"grad_norm": 0.16433264315128326,
"learning_rate": 0.0008,
"loss": 1.834672212600708,
"step": 508
},
{
"epoch": 0.10641627543035993,
"grad_norm": 0.176216721534729,
"learning_rate": 0.0008,
"loss": 1.7884635925292969,
"step": 510
},
{
"epoch": 0.10683359415753782,
"grad_norm": 0.19113944470882416,
"learning_rate": 0.0008,
"loss": 1.7823976278305054,
"step": 512
},
{
"epoch": 0.10725091288471571,
"grad_norm": 0.19091008603572845,
"learning_rate": 0.0008,
"loss": 1.840747356414795,
"step": 514
},
{
"epoch": 0.10766823161189358,
"grad_norm": 0.19281427562236786,
"learning_rate": 0.0008,
"loss": 1.75832200050354,
"step": 516
},
{
"epoch": 0.10808555033907147,
"grad_norm": 0.16847775876522064,
"learning_rate": 0.0008,
"loss": 1.8396658897399902,
"step": 518
},
{
"epoch": 0.10850286906624934,
"grad_norm": 0.2127823531627655,
"learning_rate": 0.0008,
"loss": 1.8640540838241577,
"step": 520
},
{
"epoch": 0.10892018779342723,
"grad_norm": 0.22573991119861603,
"learning_rate": 0.0008,
"loss": 1.7402691841125488,
"step": 522
},
{
"epoch": 0.10933750652060512,
"grad_norm": 0.16429032385349274,
"learning_rate": 0.0008,
"loss": 1.719557285308838,
"step": 524
},
{
"epoch": 0.10975482524778299,
"grad_norm": 0.17622490227222443,
"learning_rate": 0.0008,
"loss": 1.6947720050811768,
"step": 526
},
{
"epoch": 0.11017214397496088,
"grad_norm": 0.19966810941696167,
"learning_rate": 0.0008,
"loss": 1.7054942846298218,
"step": 528
},
{
"epoch": 0.11058946270213876,
"grad_norm": 0.1878957897424698,
"learning_rate": 0.0008,
"loss": 1.7062702178955078,
"step": 530
},
{
"epoch": 0.11100678142931664,
"grad_norm": 0.16309280693531036,
"learning_rate": 0.0008,
"loss": 1.7353664636611938,
"step": 532
},
{
"epoch": 0.11142410015649452,
"grad_norm": 0.17809820175170898,
"learning_rate": 0.0008,
"loss": 1.7127535343170166,
"step": 534
},
{
"epoch": 0.11184141888367241,
"grad_norm": 0.18687750399112701,
"learning_rate": 0.0008,
"loss": 1.8780369758605957,
"step": 536
},
{
"epoch": 0.11225873761085028,
"grad_norm": 0.16795234382152557,
"learning_rate": 0.0008,
"loss": 1.7966887950897217,
"step": 538
},
{
"epoch": 0.11267605633802817,
"grad_norm": 0.18127207458019257,
"learning_rate": 0.0008,
"loss": 1.8800458908081055,
"step": 540
},
{
"epoch": 0.11309337506520604,
"grad_norm": 0.17690226435661316,
"learning_rate": 0.0008,
"loss": 1.812783122062683,
"step": 542
},
{
"epoch": 0.11351069379238393,
"grad_norm": 0.19803214073181152,
"learning_rate": 0.0008,
"loss": 1.7694447040557861,
"step": 544
},
{
"epoch": 0.11392801251956182,
"grad_norm": 0.2248341292142868,
"learning_rate": 0.0008,
"loss": 1.9381588697433472,
"step": 546
},
{
"epoch": 0.11434533124673969,
"grad_norm": 0.21064493060112,
"learning_rate": 0.0008,
"loss": 1.6674803495407104,
"step": 548
},
{
"epoch": 0.11476264997391758,
"grad_norm": 0.1799592822790146,
"learning_rate": 0.0008,
"loss": 1.713584303855896,
"step": 550
},
{
"epoch": 0.11517996870109547,
"grad_norm": 0.17765992879867554,
"learning_rate": 0.0008,
"loss": 1.7880573272705078,
"step": 552
},
{
"epoch": 0.11559728742827334,
"grad_norm": 0.19028837978839874,
"learning_rate": 0.0008,
"loss": 1.764471173286438,
"step": 554
},
{
"epoch": 0.11601460615545123,
"grad_norm": 0.1811382919549942,
"learning_rate": 0.0008,
"loss": 1.7669916152954102,
"step": 556
},
{
"epoch": 0.11643192488262911,
"grad_norm": 0.1973322629928589,
"learning_rate": 0.0008,
"loss": 1.8208307027816772,
"step": 558
},
{
"epoch": 0.11684924360980699,
"grad_norm": 0.24303878843784332,
"learning_rate": 0.0008,
"loss": 1.6340117454528809,
"step": 560
},
{
"epoch": 0.11726656233698488,
"grad_norm": 0.20035336911678314,
"learning_rate": 0.0008,
"loss": 1.6720242500305176,
"step": 562
},
{
"epoch": 0.11768388106416275,
"grad_norm": 0.18954803049564362,
"learning_rate": 0.0008,
"loss": 1.845242977142334,
"step": 564
},
{
"epoch": 0.11810119979134064,
"grad_norm": 0.20096823573112488,
"learning_rate": 0.0008,
"loss": 1.8153274059295654,
"step": 566
},
{
"epoch": 0.11851851851851852,
"grad_norm": 0.19449801743030548,
"learning_rate": 0.0008,
"loss": 1.751156210899353,
"step": 568
},
{
"epoch": 0.1189358372456964,
"grad_norm": 0.17495417594909668,
"learning_rate": 0.0008,
"loss": 1.835421085357666,
"step": 570
},
{
"epoch": 0.11935315597287428,
"grad_norm": 0.1892746388912201,
"learning_rate": 0.0008,
"loss": 1.8085366487503052,
"step": 572
},
{
"epoch": 0.11977047470005217,
"grad_norm": 0.19496633112430573,
"learning_rate": 0.0008,
"loss": 1.7708959579467773,
"step": 574
},
{
"epoch": 0.12018779342723004,
"grad_norm": 0.2227867990732193,
"learning_rate": 0.0008,
"loss": 1.6941637992858887,
"step": 576
},
{
"epoch": 0.12060511215440793,
"grad_norm": 0.18124914169311523,
"learning_rate": 0.0008,
"loss": 1.8590575456619263,
"step": 578
},
{
"epoch": 0.1210224308815858,
"grad_norm": 0.17786574363708496,
"learning_rate": 0.0008,
"loss": 1.8845808506011963,
"step": 580
},
{
"epoch": 0.12143974960876369,
"grad_norm": 0.18539316952228546,
"learning_rate": 0.0008,
"loss": 1.7233104705810547,
"step": 582
},
{
"epoch": 0.12185706833594158,
"grad_norm": 0.22436559200286865,
"learning_rate": 0.0008,
"loss": 1.7101248502731323,
"step": 584
},
{
"epoch": 0.12227438706311945,
"grad_norm": 0.24816755950450897,
"learning_rate": 0.0008,
"loss": 1.8819714784622192,
"step": 586
},
{
"epoch": 0.12269170579029734,
"grad_norm": 0.1978999376296997,
"learning_rate": 0.0008,
"loss": 1.812936782836914,
"step": 588
},
{
"epoch": 0.12310902451747523,
"grad_norm": 0.20076996088027954,
"learning_rate": 0.0008,
"loss": 1.6589692831039429,
"step": 590
},
{
"epoch": 0.1235263432446531,
"grad_norm": 0.18335849046707153,
"learning_rate": 0.0008,
"loss": 1.888481855392456,
"step": 592
},
{
"epoch": 0.12394366197183099,
"grad_norm": 0.22114887833595276,
"learning_rate": 0.0008,
"loss": 1.752468228340149,
"step": 594
},
{
"epoch": 0.12436098069900887,
"grad_norm": 0.1933581680059433,
"learning_rate": 0.0008,
"loss": 1.7702640295028687,
"step": 596
},
{
"epoch": 0.12477829942618675,
"grad_norm": 0.1983618289232254,
"learning_rate": 0.0008,
"loss": 1.8683502674102783,
"step": 598
},
{
"epoch": 0.12519561815336464,
"grad_norm": 0.16474363207817078,
"learning_rate": 0.0008,
"loss": 1.70648193359375,
"step": 600
},
{
"epoch": 0.1256129368805425,
"grad_norm": 0.20615136623382568,
"learning_rate": 0.0008,
"loss": 1.7076618671417236,
"step": 602
},
{
"epoch": 0.1260302556077204,
"grad_norm": 0.21693238615989685,
"learning_rate": 0.0008,
"loss": 1.8148608207702637,
"step": 604
},
{
"epoch": 0.12644757433489828,
"grad_norm": 0.17567431926727295,
"learning_rate": 0.0008,
"loss": 1.837254285812378,
"step": 606
},
{
"epoch": 0.12686489306207616,
"grad_norm": 0.18390585482120514,
"learning_rate": 0.0008,
"loss": 1.8590507507324219,
"step": 608
},
{
"epoch": 0.12728221178925403,
"grad_norm": 0.16418585181236267,
"learning_rate": 0.0008,
"loss": 1.6788272857666016,
"step": 610
},
{
"epoch": 0.12769953051643193,
"grad_norm": 0.16819006204605103,
"learning_rate": 0.0008,
"loss": 1.8437262773513794,
"step": 612
},
{
"epoch": 0.1281168492436098,
"grad_norm": 0.15948879718780518,
"learning_rate": 0.0008,
"loss": 1.7340611219406128,
"step": 614
},
{
"epoch": 0.12853416797078768,
"grad_norm": 0.2011558562517166,
"learning_rate": 0.0008,
"loss": 1.7709379196166992,
"step": 616
},
{
"epoch": 0.12895148669796558,
"grad_norm": 0.19379252195358276,
"learning_rate": 0.0008,
"loss": 1.806614637374878,
"step": 618
},
{
"epoch": 0.12936880542514345,
"grad_norm": 0.22475005686283112,
"learning_rate": 0.0008,
"loss": 1.711434006690979,
"step": 620
},
{
"epoch": 0.12978612415232132,
"grad_norm": 0.1978297084569931,
"learning_rate": 0.0008,
"loss": 1.7318035364151,
"step": 622
},
{
"epoch": 0.13020344287949923,
"grad_norm": 0.18323394656181335,
"learning_rate": 0.0008,
"loss": 1.7498455047607422,
"step": 624
},
{
"epoch": 0.1306207616066771,
"grad_norm": 0.22182956337928772,
"learning_rate": 0.0008,
"loss": 1.7923178672790527,
"step": 626
},
{
"epoch": 0.13103808033385497,
"grad_norm": 0.1770567148923874,
"learning_rate": 0.0008,
"loss": 1.6318275928497314,
"step": 628
},
{
"epoch": 0.13145539906103287,
"grad_norm": 0.20539313554763794,
"learning_rate": 0.0008,
"loss": 1.7883520126342773,
"step": 630
},
{
"epoch": 0.13187271778821075,
"grad_norm": 0.17130990326404572,
"learning_rate": 0.0008,
"loss": 1.7609513998031616,
"step": 632
},
{
"epoch": 0.13229003651538862,
"grad_norm": 0.1961364895105362,
"learning_rate": 0.0008,
"loss": 1.7397228479385376,
"step": 634
},
{
"epoch": 0.13270735524256652,
"grad_norm": 0.16240109503269196,
"learning_rate": 0.0008,
"loss": 1.8233591318130493,
"step": 636
},
{
"epoch": 0.1331246739697444,
"grad_norm": 0.19982294738292694,
"learning_rate": 0.0008,
"loss": 1.806249976158142,
"step": 638
},
{
"epoch": 0.13354199269692227,
"grad_norm": 0.2089509665966034,
"learning_rate": 0.0008,
"loss": 1.8161576986312866,
"step": 640
},
{
"epoch": 0.13395931142410017,
"grad_norm": 0.1830461472272873,
"learning_rate": 0.0008,
"loss": 1.8403010368347168,
"step": 642
},
{
"epoch": 0.13437663015127804,
"grad_norm": 0.20117032527923584,
"learning_rate": 0.0008,
"loss": 1.8304057121276855,
"step": 644
},
{
"epoch": 0.13479394887845592,
"grad_norm": 0.18736901879310608,
"learning_rate": 0.0008,
"loss": 1.773429036140442,
"step": 646
},
{
"epoch": 0.1352112676056338,
"grad_norm": 0.2435261756181717,
"learning_rate": 0.0008,
"loss": 1.6929211616516113,
"step": 648
},
{
"epoch": 0.1356285863328117,
"grad_norm": 0.2115386575460434,
"learning_rate": 0.0008,
"loss": 1.862548589706421,
"step": 650
},
{
"epoch": 0.13604590505998956,
"grad_norm": 0.17527717351913452,
"learning_rate": 0.0008,
"loss": 1.7797799110412598,
"step": 652
},
{
"epoch": 0.13646322378716744,
"grad_norm": 0.1893661767244339,
"learning_rate": 0.0008,
"loss": 1.8103671073913574,
"step": 654
},
{
"epoch": 0.13688054251434534,
"grad_norm": 0.19153577089309692,
"learning_rate": 0.0008,
"loss": 1.6010644435882568,
"step": 656
},
{
"epoch": 0.1372978612415232,
"grad_norm": 0.18063318729400635,
"learning_rate": 0.0008,
"loss": 1.7611302137374878,
"step": 658
},
{
"epoch": 0.13771517996870108,
"grad_norm": 0.19133161008358002,
"learning_rate": 0.0008,
"loss": 1.7387412786483765,
"step": 660
},
{
"epoch": 0.13813249869587899,
"grad_norm": 0.20165090262889862,
"learning_rate": 0.0008,
"loss": 1.7242095470428467,
"step": 662
},
{
"epoch": 0.13854981742305686,
"grad_norm": 0.17320813238620758,
"learning_rate": 0.0008,
"loss": 1.9070261716842651,
"step": 664
},
{
"epoch": 0.13896713615023473,
"grad_norm": 0.19484713673591614,
"learning_rate": 0.0008,
"loss": 1.830524206161499,
"step": 666
},
{
"epoch": 0.13938445487741263,
"grad_norm": 0.17643558979034424,
"learning_rate": 0.0008,
"loss": 1.7296078205108643,
"step": 668
},
{
"epoch": 0.1398017736045905,
"grad_norm": 0.18590174615383148,
"learning_rate": 0.0008,
"loss": 1.8072155714035034,
"step": 670
},
{
"epoch": 0.14021909233176838,
"grad_norm": 0.17337389290332794,
"learning_rate": 0.0008,
"loss": 1.7468055486679077,
"step": 672
},
{
"epoch": 0.14063641105894628,
"grad_norm": 0.20085087418556213,
"learning_rate": 0.0008,
"loss": 1.7447450160980225,
"step": 674
},
{
"epoch": 0.14105372978612415,
"grad_norm": 0.20829616487026215,
"learning_rate": 0.0008,
"loss": 1.8087923526763916,
"step": 676
},
{
"epoch": 0.14147104851330203,
"grad_norm": 0.17774710059165955,
"learning_rate": 0.0008,
"loss": 1.8168830871582031,
"step": 678
},
{
"epoch": 0.14188836724047993,
"grad_norm": 0.1968492716550827,
"learning_rate": 0.0008,
"loss": 1.798553705215454,
"step": 680
},
{
"epoch": 0.1423056859676578,
"grad_norm": 0.17439714074134827,
"learning_rate": 0.0008,
"loss": 1.8232721090316772,
"step": 682
},
{
"epoch": 0.14272300469483568,
"grad_norm": 0.16724388301372528,
"learning_rate": 0.0008,
"loss": 1.596436619758606,
"step": 684
},
{
"epoch": 0.14314032342201358,
"grad_norm": 0.2098611444234848,
"learning_rate": 0.0008,
"loss": 1.7849904298782349,
"step": 686
},
{
"epoch": 0.14355764214919145,
"grad_norm": 0.18629077076911926,
"learning_rate": 0.0008,
"loss": 1.7334709167480469,
"step": 688
},
{
"epoch": 0.14397496087636932,
"grad_norm": 0.18207907676696777,
"learning_rate": 0.0008,
"loss": 1.973663330078125,
"step": 690
},
{
"epoch": 0.1443922796035472,
"grad_norm": 0.16263729333877563,
"learning_rate": 0.0008,
"loss": 1.8095067739486694,
"step": 692
},
{
"epoch": 0.1448095983307251,
"grad_norm": 0.18567156791687012,
"learning_rate": 0.0008,
"loss": 1.765048623085022,
"step": 694
},
{
"epoch": 0.14522691705790297,
"grad_norm": 0.18123917281627655,
"learning_rate": 0.0008,
"loss": 1.8273537158966064,
"step": 696
},
{
"epoch": 0.14564423578508084,
"grad_norm": 0.20769929885864258,
"learning_rate": 0.0008,
"loss": 1.8118430376052856,
"step": 698
},
{
"epoch": 0.14606155451225875,
"grad_norm": 0.1864776760339737,
"learning_rate": 0.0008,
"loss": 1.6669319868087769,
"step": 700
},
{
"epoch": 0.14647887323943662,
"grad_norm": 0.18755193054676056,
"learning_rate": 0.0008,
"loss": 1.6617186069488525,
"step": 702
},
{
"epoch": 0.1468961919666145,
"grad_norm": 0.18912330269813538,
"learning_rate": 0.0008,
"loss": 1.7503671646118164,
"step": 704
},
{
"epoch": 0.1473135106937924,
"grad_norm": 0.1890300065279007,
"learning_rate": 0.0008,
"loss": 1.695054531097412,
"step": 706
},
{
"epoch": 0.14773082942097027,
"grad_norm": 0.17214125394821167,
"learning_rate": 0.0008,
"loss": 1.8216289281845093,
"step": 708
},
{
"epoch": 0.14814814814814814,
"grad_norm": 0.20777983963489532,
"learning_rate": 0.0008,
"loss": 1.8245275020599365,
"step": 710
},
{
"epoch": 0.14856546687532604,
"grad_norm": 0.20084725320339203,
"learning_rate": 0.0008,
"loss": 1.7887156009674072,
"step": 712
},
{
"epoch": 0.14898278560250391,
"grad_norm": 0.15811371803283691,
"learning_rate": 0.0008,
"loss": 1.7509719133377075,
"step": 714
},
{
"epoch": 0.1494001043296818,
"grad_norm": 0.16521748900413513,
"learning_rate": 0.0008,
"loss": 1.7708344459533691,
"step": 716
},
{
"epoch": 0.1498174230568597,
"grad_norm": 0.19016806781291962,
"learning_rate": 0.0008,
"loss": 1.7238750457763672,
"step": 718
},
{
"epoch": 0.15023474178403756,
"grad_norm": 0.20112618803977966,
"learning_rate": 0.0008,
"loss": 1.7684839963912964,
"step": 720
},
{
"epoch": 0.15065206051121544,
"grad_norm": 0.21002675592899323,
"learning_rate": 0.0008,
"loss": 1.668319821357727,
"step": 722
},
{
"epoch": 0.15106937923839334,
"grad_norm": 0.18244609236717224,
"learning_rate": 0.0008,
"loss": 1.8438975811004639,
"step": 724
},
{
"epoch": 0.1514866979655712,
"grad_norm": 0.17502567172050476,
"learning_rate": 0.0008,
"loss": 1.7859959602355957,
"step": 726
},
{
"epoch": 0.15190401669274908,
"grad_norm": 0.200490802526474,
"learning_rate": 0.0008,
"loss": 1.6951396465301514,
"step": 728
},
{
"epoch": 0.15232133541992696,
"grad_norm": 0.19410744309425354,
"learning_rate": 0.0008,
"loss": 1.7291679382324219,
"step": 730
},
{
"epoch": 0.15273865414710486,
"grad_norm": 0.17764674127101898,
"learning_rate": 0.0008,
"loss": 1.872592806816101,
"step": 732
},
{
"epoch": 0.15315597287428273,
"grad_norm": 0.19124142825603485,
"learning_rate": 0.0008,
"loss": 1.7468235492706299,
"step": 734
},
{
"epoch": 0.1535732916014606,
"grad_norm": 0.202019602060318,
"learning_rate": 0.0008,
"loss": 1.6559799909591675,
"step": 736
},
{
"epoch": 0.1539906103286385,
"grad_norm": 0.193822979927063,
"learning_rate": 0.0008,
"loss": 1.6957603693008423,
"step": 738
},
{
"epoch": 0.15440792905581638,
"grad_norm": 0.25882846117019653,
"learning_rate": 0.0008,
"loss": 1.6634539365768433,
"step": 740
},
{
"epoch": 0.15482524778299425,
"grad_norm": 0.2067813277244568,
"learning_rate": 0.0008,
"loss": 1.93544340133667,
"step": 742
},
{
"epoch": 0.15524256651017215,
"grad_norm": 0.1865290254354477,
"learning_rate": 0.0008,
"loss": 1.8601853847503662,
"step": 744
},
{
"epoch": 0.15565988523735003,
"grad_norm": 0.1881532222032547,
"learning_rate": 0.0008,
"loss": 1.7317306995391846,
"step": 746
},
{
"epoch": 0.1560772039645279,
"grad_norm": 0.22773560881614685,
"learning_rate": 0.0008,
"loss": 1.680111050605774,
"step": 748
},
{
"epoch": 0.1564945226917058,
"grad_norm": 0.18615806102752686,
"learning_rate": 0.0008,
"loss": 1.7144302129745483,
"step": 750
},
{
"epoch": 0.15691184141888367,
"grad_norm": 0.16930066049098969,
"learning_rate": 0.0008,
"loss": 1.8201231956481934,
"step": 752
},
{
"epoch": 0.15732916014606155,
"grad_norm": 0.19441792368888855,
"learning_rate": 0.0008,
"loss": 1.7160589694976807,
"step": 754
},
{
"epoch": 0.15774647887323945,
"grad_norm": 0.15927526354789734,
"learning_rate": 0.0008,
"loss": 1.8689998388290405,
"step": 756
},
{
"epoch": 0.15816379760041732,
"grad_norm": 0.19691766798496246,
"learning_rate": 0.0008,
"loss": 1.7673816680908203,
"step": 758
},
{
"epoch": 0.1585811163275952,
"grad_norm": 0.1787274181842804,
"learning_rate": 0.0008,
"loss": 1.7166606187820435,
"step": 760
},
{
"epoch": 0.1589984350547731,
"grad_norm": 0.1795300543308258,
"learning_rate": 0.0008,
"loss": 1.7627328634262085,
"step": 762
},
{
"epoch": 0.15941575378195097,
"grad_norm": 0.17841686308383942,
"learning_rate": 0.0008,
"loss": 1.6936697959899902,
"step": 764
},
{
"epoch": 0.15983307250912884,
"grad_norm": 0.1797875016927719,
"learning_rate": 0.0008,
"loss": 1.799784541130066,
"step": 766
},
{
"epoch": 0.16025039123630672,
"grad_norm": 0.21431873738765717,
"learning_rate": 0.0008,
"loss": 1.7679109573364258,
"step": 768
},
{
"epoch": 0.16066770996348462,
"grad_norm": 0.1949765682220459,
"learning_rate": 0.0008,
"loss": 1.719260334968567,
"step": 770
},
{
"epoch": 0.1610850286906625,
"grad_norm": 0.19613662362098694,
"learning_rate": 0.0008,
"loss": 1.7998367547988892,
"step": 772
},
{
"epoch": 0.16150234741784036,
"grad_norm": 0.17787288129329681,
"learning_rate": 0.0008,
"loss": 1.8460516929626465,
"step": 774
},
{
"epoch": 0.16191966614501827,
"grad_norm": 0.22785474359989166,
"learning_rate": 0.0008,
"loss": 1.6813404560089111,
"step": 776
},
{
"epoch": 0.16233698487219614,
"grad_norm": 0.1806975156068802,
"learning_rate": 0.0008,
"loss": 1.7418760061264038,
"step": 778
},
{
"epoch": 0.162754303599374,
"grad_norm": 0.1856619268655777,
"learning_rate": 0.0008,
"loss": 1.7497477531433105,
"step": 780
},
{
"epoch": 0.1631716223265519,
"grad_norm": 0.1854953020811081,
"learning_rate": 0.0008,
"loss": 1.7518516778945923,
"step": 782
},
{
"epoch": 0.16358894105372979,
"grad_norm": 0.22963276505470276,
"learning_rate": 0.0008,
"loss": 1.595357060432434,
"step": 784
},
{
"epoch": 0.16400625978090766,
"grad_norm": 0.17343032360076904,
"learning_rate": 0.0008,
"loss": 1.7026077508926392,
"step": 786
},
{
"epoch": 0.16442357850808556,
"grad_norm": 0.20681700110435486,
"learning_rate": 0.0008,
"loss": 1.686074137687683,
"step": 788
},
{
"epoch": 0.16484089723526343,
"grad_norm": 0.17650754749774933,
"learning_rate": 0.0008,
"loss": 1.6811248064041138,
"step": 790
},
{
"epoch": 0.1652582159624413,
"grad_norm": 0.1636354923248291,
"learning_rate": 0.0008,
"loss": 1.7299340963363647,
"step": 792
},
{
"epoch": 0.1656755346896192,
"grad_norm": 0.1806851327419281,
"learning_rate": 0.0008,
"loss": 1.8136711120605469,
"step": 794
},
{
"epoch": 0.16609285341679708,
"grad_norm": 0.19005830585956573,
"learning_rate": 0.0008,
"loss": 1.8194869756698608,
"step": 796
},
{
"epoch": 0.16651017214397495,
"grad_norm": 0.18151606619358063,
"learning_rate": 0.0008,
"loss": 1.8666720390319824,
"step": 798
},
{
"epoch": 0.16692749087115286,
"grad_norm": 0.1761510819196701,
"learning_rate": 0.0008,
"loss": 1.7758731842041016,
"step": 800
},
{
"epoch": 0.16734480959833073,
"grad_norm": 0.19962024688720703,
"learning_rate": 0.0008,
"loss": 1.736701250076294,
"step": 802
},
{
"epoch": 0.1677621283255086,
"grad_norm": 0.17402872443199158,
"learning_rate": 0.0008,
"loss": 1.6931071281433105,
"step": 804
},
{
"epoch": 0.16817944705268648,
"grad_norm": 0.18706336617469788,
"learning_rate": 0.0008,
"loss": 1.7926737070083618,
"step": 806
},
{
"epoch": 0.16859676577986438,
"grad_norm": 0.16924500465393066,
"learning_rate": 0.0008,
"loss": 1.7937047481536865,
"step": 808
},
{
"epoch": 0.16901408450704225,
"grad_norm": 0.21804003417491913,
"learning_rate": 0.0008,
"loss": 1.6593379974365234,
"step": 810
},
{
"epoch": 0.16943140323422012,
"grad_norm": 0.17710910737514496,
"learning_rate": 0.0008,
"loss": 1.8356101512908936,
"step": 812
},
{
"epoch": 0.16984872196139802,
"grad_norm": 0.1932608038187027,
"learning_rate": 0.0008,
"loss": 1.6077665090560913,
"step": 814
},
{
"epoch": 0.1702660406885759,
"grad_norm": 0.17858533561229706,
"learning_rate": 0.0008,
"loss": 1.8420521020889282,
"step": 816
},
{
"epoch": 0.17068335941575377,
"grad_norm": 0.1795278787612915,
"learning_rate": 0.0008,
"loss": 1.7781267166137695,
"step": 818
},
{
"epoch": 0.17110067814293167,
"grad_norm": 0.19953025877475739,
"learning_rate": 0.0008,
"loss": 1.8236104249954224,
"step": 820
},
{
"epoch": 0.17151799687010955,
"grad_norm": 0.21639135479927063,
"learning_rate": 0.0008,
"loss": 1.7180458307266235,
"step": 822
},
{
"epoch": 0.17193531559728742,
"grad_norm": 0.21022126078605652,
"learning_rate": 0.0008,
"loss": 1.679661750793457,
"step": 824
},
{
"epoch": 0.17235263432446532,
"grad_norm": 0.20463056862354279,
"learning_rate": 0.0008,
"loss": 1.6936852931976318,
"step": 826
},
{
"epoch": 0.1727699530516432,
"grad_norm": 0.1905946135520935,
"learning_rate": 0.0008,
"loss": 1.8837283849716187,
"step": 828
},
{
"epoch": 0.17318727177882107,
"grad_norm": 0.18571698665618896,
"learning_rate": 0.0008,
"loss": 1.8333275318145752,
"step": 830
},
{
"epoch": 0.17360459050599897,
"grad_norm": 0.1744394600391388,
"learning_rate": 0.0008,
"loss": 1.7868090867996216,
"step": 832
},
{
"epoch": 0.17402190923317684,
"grad_norm": 0.20897187292575836,
"learning_rate": 0.0008,
"loss": 1.5670733451843262,
"step": 834
},
{
"epoch": 0.17443922796035471,
"grad_norm": 0.2123873084783554,
"learning_rate": 0.0008,
"loss": 1.6381040811538696,
"step": 836
},
{
"epoch": 0.17485654668753262,
"grad_norm": 0.18907363712787628,
"learning_rate": 0.0008,
"loss": 1.7446024417877197,
"step": 838
},
{
"epoch": 0.1752738654147105,
"grad_norm": 0.20559073984622955,
"learning_rate": 0.0008,
"loss": 1.6502224206924438,
"step": 840
},
{
"epoch": 0.17569118414188836,
"grad_norm": 0.19022029638290405,
"learning_rate": 0.0008,
"loss": 1.8096051216125488,
"step": 842
},
{
"epoch": 0.17610850286906624,
"grad_norm": 0.16331380605697632,
"learning_rate": 0.0008,
"loss": 1.818407416343689,
"step": 844
},
{
"epoch": 0.17652582159624414,
"grad_norm": 0.2162676900625229,
"learning_rate": 0.0008,
"loss": 1.7258923053741455,
"step": 846
},
{
"epoch": 0.176943140323422,
"grad_norm": 0.18963304162025452,
"learning_rate": 0.0008,
"loss": 1.729779839515686,
"step": 848
},
{
"epoch": 0.17736045905059988,
"grad_norm": 0.19266974925994873,
"learning_rate": 0.0008,
"loss": 1.797463059425354,
"step": 850
},
{
"epoch": 0.17777777777777778,
"grad_norm": 0.1722777932882309,
"learning_rate": 0.0008,
"loss": 1.7928173542022705,
"step": 852
},
{
"epoch": 0.17819509650495566,
"grad_norm": 0.1919032633304596,
"learning_rate": 0.0008,
"loss": 1.8268866539001465,
"step": 854
},
{
"epoch": 0.17861241523213353,
"grad_norm": 0.22310693562030792,
"learning_rate": 0.0008,
"loss": 1.7955056428909302,
"step": 856
},
{
"epoch": 0.17902973395931143,
"grad_norm": 0.19400399923324585,
"learning_rate": 0.0008,
"loss": 1.7923102378845215,
"step": 858
},
{
"epoch": 0.1794470526864893,
"grad_norm": 0.21149571239948273,
"learning_rate": 0.0008,
"loss": 1.7186392545700073,
"step": 860
},
{
"epoch": 0.17986437141366718,
"grad_norm": 0.17790478467941284,
"learning_rate": 0.0008,
"loss": 1.7033177614212036,
"step": 862
},
{
"epoch": 0.18028169014084508,
"grad_norm": 0.16812187433242798,
"learning_rate": 0.0008,
"loss": 1.8667421340942383,
"step": 864
},
{
"epoch": 0.18069900886802295,
"grad_norm": 0.1852063685655594,
"learning_rate": 0.0008,
"loss": 1.7483296394348145,
"step": 866
},
{
"epoch": 0.18111632759520083,
"grad_norm": 0.18790170550346375,
"learning_rate": 0.0008,
"loss": 1.7478761672973633,
"step": 868
},
{
"epoch": 0.18153364632237873,
"grad_norm": 0.1689843088388443,
"learning_rate": 0.0008,
"loss": 1.9035534858703613,
"step": 870
},
{
"epoch": 0.1819509650495566,
"grad_norm": 0.19190579652786255,
"learning_rate": 0.0008,
"loss": 1.67540442943573,
"step": 872
},
{
"epoch": 0.18236828377673447,
"grad_norm": 0.17893852293491364,
"learning_rate": 0.0008,
"loss": 1.7380026578903198,
"step": 874
},
{
"epoch": 0.18278560250391238,
"grad_norm": 0.20149573683738708,
"learning_rate": 0.0008,
"loss": 1.6873263120651245,
"step": 876
},
{
"epoch": 0.18320292123109025,
"grad_norm": 0.17393316328525543,
"learning_rate": 0.0008,
"loss": 1.7170566320419312,
"step": 878
},
{
"epoch": 0.18362023995826812,
"grad_norm": 0.21316109597682953,
"learning_rate": 0.0008,
"loss": 1.6952152252197266,
"step": 880
},
{
"epoch": 0.18403755868544602,
"grad_norm": 0.20032194256782532,
"learning_rate": 0.0008,
"loss": 1.754309892654419,
"step": 882
},
{
"epoch": 0.1844548774126239,
"grad_norm": 0.20200282335281372,
"learning_rate": 0.0008,
"loss": 1.7858588695526123,
"step": 884
},
{
"epoch": 0.18487219613980177,
"grad_norm": 0.18926279246807098,
"learning_rate": 0.0008,
"loss": 1.7106082439422607,
"step": 886
},
{
"epoch": 0.18528951486697964,
"grad_norm": 0.16581322252750397,
"learning_rate": 0.0008,
"loss": 1.8427886962890625,
"step": 888
},
{
"epoch": 0.18570683359415754,
"grad_norm": 0.20399515330791473,
"learning_rate": 0.0008,
"loss": 1.7196152210235596,
"step": 890
},
{
"epoch": 0.18612415232133542,
"grad_norm": 0.19511224329471588,
"learning_rate": 0.0008,
"loss": 1.8195468187332153,
"step": 892
},
{
"epoch": 0.1865414710485133,
"grad_norm": 0.19643206894397736,
"learning_rate": 0.0008,
"loss": 1.6806929111480713,
"step": 894
},
{
"epoch": 0.1869587897756912,
"grad_norm": 0.1884148269891739,
"learning_rate": 0.0008,
"loss": 1.6993000507354736,
"step": 896
},
{
"epoch": 0.18737610850286907,
"grad_norm": 0.17956124246120453,
"learning_rate": 0.0008,
"loss": 1.8369438648223877,
"step": 898
},
{
"epoch": 0.18779342723004694,
"grad_norm": 0.1978001445531845,
"learning_rate": 0.0008,
"loss": 1.80470609664917,
"step": 900
},
{
"epoch": 0.18821074595722484,
"grad_norm": 0.14546366035938263,
"learning_rate": 0.0008,
"loss": 1.6158053874969482,
"step": 902
},
{
"epoch": 0.1886280646844027,
"grad_norm": 0.1694905012845993,
"learning_rate": 0.0008,
"loss": 1.7942674160003662,
"step": 904
},
{
"epoch": 0.1890453834115806,
"grad_norm": 0.1844390481710434,
"learning_rate": 0.0008,
"loss": 1.7278395891189575,
"step": 906
},
{
"epoch": 0.1894627021387585,
"grad_norm": 0.18887804448604584,
"learning_rate": 0.0008,
"loss": 1.7810070514678955,
"step": 908
},
{
"epoch": 0.18988002086593636,
"grad_norm": 0.17793181538581848,
"learning_rate": 0.0008,
"loss": 1.715239405632019,
"step": 910
},
{
"epoch": 0.19029733959311423,
"grad_norm": 0.1882665455341339,
"learning_rate": 0.0008,
"loss": 1.8410365581512451,
"step": 912
},
{
"epoch": 0.19071465832029214,
"grad_norm": 0.1665799915790558,
"learning_rate": 0.0008,
"loss": 1.8286895751953125,
"step": 914
},
{
"epoch": 0.19113197704747,
"grad_norm": 0.1985129565000534,
"learning_rate": 0.0008,
"loss": 1.7565686702728271,
"step": 916
},
{
"epoch": 0.19154929577464788,
"grad_norm": 0.16701538860797882,
"learning_rate": 0.0008,
"loss": 1.7466310262680054,
"step": 918
},
{
"epoch": 0.19196661450182578,
"grad_norm": 0.16946537792682648,
"learning_rate": 0.0008,
"loss": 1.6163197755813599,
"step": 920
},
{
"epoch": 0.19238393322900366,
"grad_norm": 0.1770622581243515,
"learning_rate": 0.0008,
"loss": 1.7802177667617798,
"step": 922
},
{
"epoch": 0.19280125195618153,
"grad_norm": 0.1672552227973938,
"learning_rate": 0.0008,
"loss": 1.6146728992462158,
"step": 924
},
{
"epoch": 0.1932185706833594,
"grad_norm": 0.16920343041419983,
"learning_rate": 0.0008,
"loss": 1.829084873199463,
"step": 926
},
{
"epoch": 0.1936358894105373,
"grad_norm": 0.19028623402118683,
"learning_rate": 0.0008,
"loss": 1.794987440109253,
"step": 928
},
{
"epoch": 0.19405320813771518,
"grad_norm": 0.17238767445087433,
"learning_rate": 0.0008,
"loss": 1.7409614324569702,
"step": 930
},
{
"epoch": 0.19447052686489305,
"grad_norm": 0.2275260090827942,
"learning_rate": 0.0008,
"loss": 1.6707667112350464,
"step": 932
},
{
"epoch": 0.19488784559207095,
"grad_norm": 0.216999351978302,
"learning_rate": 0.0008,
"loss": 1.7779223918914795,
"step": 934
},
{
"epoch": 0.19530516431924883,
"grad_norm": 0.16746197640895844,
"learning_rate": 0.0008,
"loss": 1.7970597743988037,
"step": 936
},
{
"epoch": 0.1957224830464267,
"grad_norm": 0.18787866830825806,
"learning_rate": 0.0008,
"loss": 1.804955005645752,
"step": 938
},
{
"epoch": 0.1961398017736046,
"grad_norm": 0.18515397608280182,
"learning_rate": 0.0008,
"loss": 1.8307769298553467,
"step": 940
},
{
"epoch": 0.19655712050078247,
"grad_norm": 0.20320598781108856,
"learning_rate": 0.0008,
"loss": 1.7676329612731934,
"step": 942
},
{
"epoch": 0.19697443922796035,
"grad_norm": 0.17765004932880402,
"learning_rate": 0.0008,
"loss": 1.674498438835144,
"step": 944
},
{
"epoch": 0.19739175795513825,
"grad_norm": 0.19797641038894653,
"learning_rate": 0.0008,
"loss": 1.8213812112808228,
"step": 946
},
{
"epoch": 0.19780907668231612,
"grad_norm": 0.19149918854236603,
"learning_rate": 0.0008,
"loss": 1.8154566287994385,
"step": 948
},
{
"epoch": 0.198226395409494,
"grad_norm": 0.1972745805978775,
"learning_rate": 0.0008,
"loss": 1.761203408241272,
"step": 950
},
{
"epoch": 0.1986437141366719,
"grad_norm": 0.16813834011554718,
"learning_rate": 0.0008,
"loss": 1.769463300704956,
"step": 952
},
{
"epoch": 0.19906103286384977,
"grad_norm": 0.17149922251701355,
"learning_rate": 0.0008,
"loss": 1.7954530715942383,
"step": 954
},
{
"epoch": 0.19947835159102764,
"grad_norm": 0.17822957038879395,
"learning_rate": 0.0008,
"loss": 1.719478964805603,
"step": 956
},
{
"epoch": 0.19989567031820554,
"grad_norm": 0.16247020661830902,
"learning_rate": 0.0008,
"loss": 1.5951590538024902,
"step": 958
},
{
"epoch": 0.20010432968179448,
"eval_loss": 1.7186682224273682,
"eval_runtime": 26.6933,
"eval_samples_per_second": 58.067,
"eval_steps_per_second": 7.268,
"step": 959
},
{
"epoch": 0.20031298904538342,
"grad_norm": 0.19236379861831665,
"learning_rate": 0.0008,
"loss": 1.7191988229751587,
"step": 960
},
{
"epoch": 0.2007303077725613,
"grad_norm": 0.18168877065181732,
"learning_rate": 0.0008,
"loss": 1.7212427854537964,
"step": 962
},
{
"epoch": 0.20114762649973916,
"grad_norm": 0.1752994805574417,
"learning_rate": 0.0008,
"loss": 1.822859525680542,
"step": 964
},
{
"epoch": 0.20156494522691706,
"grad_norm": 0.2033938467502594,
"learning_rate": 0.0008,
"loss": 1.7316899299621582,
"step": 966
},
{
"epoch": 0.20198226395409494,
"grad_norm": 0.17674045264720917,
"learning_rate": 0.0008,
"loss": 1.6789060831069946,
"step": 968
},
{
"epoch": 0.2023995826812728,
"grad_norm": 0.17077626287937164,
"learning_rate": 0.0008,
"loss": 1.712573528289795,
"step": 970
},
{
"epoch": 0.2028169014084507,
"grad_norm": 0.17276069521903992,
"learning_rate": 0.0008,
"loss": 1.842163324356079,
"step": 972
},
{
"epoch": 0.20323422013562858,
"grad_norm": 0.174738347530365,
"learning_rate": 0.0008,
"loss": 1.637934684753418,
"step": 974
},
{
"epoch": 0.20365153886280646,
"grad_norm": 0.18814903497695923,
"learning_rate": 0.0008,
"loss": 1.7716403007507324,
"step": 976
},
{
"epoch": 0.20406885758998436,
"grad_norm": 0.19485580921173096,
"learning_rate": 0.0008,
"loss": 1.7023565769195557,
"step": 978
},
{
"epoch": 0.20448617631716223,
"grad_norm": 0.19737398624420166,
"learning_rate": 0.0008,
"loss": 1.7805975675582886,
"step": 980
},
{
"epoch": 0.2049034950443401,
"grad_norm": 0.1848728358745575,
"learning_rate": 0.0008,
"loss": 1.7100493907928467,
"step": 982
},
{
"epoch": 0.205320813771518,
"grad_norm": 0.2276126742362976,
"learning_rate": 0.0008,
"loss": 1.6969510316848755,
"step": 984
},
{
"epoch": 0.20573813249869588,
"grad_norm": 0.16853907704353333,
"learning_rate": 0.0008,
"loss": 1.7329500913619995,
"step": 986
},
{
"epoch": 0.20615545122587375,
"grad_norm": 0.19605912268161774,
"learning_rate": 0.0008,
"loss": 1.783444881439209,
"step": 988
},
{
"epoch": 0.20657276995305165,
"grad_norm": 0.20781081914901733,
"learning_rate": 0.0008,
"loss": 1.7230795621871948,
"step": 990
},
{
"epoch": 0.20699008868022953,
"grad_norm": 0.20055805146694183,
"learning_rate": 0.0008,
"loss": 1.8591951131820679,
"step": 992
},
{
"epoch": 0.2074074074074074,
"grad_norm": 0.1904158890247345,
"learning_rate": 0.0008,
"loss": 1.7955559492111206,
"step": 994
},
{
"epoch": 0.2078247261345853,
"grad_norm": 0.19634753465652466,
"learning_rate": 0.0008,
"loss": 1.6735583543777466,
"step": 996
},
{
"epoch": 0.20824204486176318,
"grad_norm": 0.20250703394412994,
"learning_rate": 0.0008,
"loss": 1.802290439605713,
"step": 998
},
{
"epoch": 0.20865936358894105,
"grad_norm": 0.23513184487819672,
"learning_rate": 0.0008,
"loss": 1.5549508333206177,
"step": 1000
},
{
"epoch": 0.20907668231611892,
"grad_norm": 0.19205506145954132,
"learning_rate": 0.0008,
"loss": 1.6269066333770752,
"step": 1002
},
{
"epoch": 0.20949400104329682,
"grad_norm": 0.19439759850502014,
"learning_rate": 0.0008,
"loss": 1.7897770404815674,
"step": 1004
},
{
"epoch": 0.2099113197704747,
"grad_norm": 0.1730080395936966,
"learning_rate": 0.0008,
"loss": 1.6401969194412231,
"step": 1006
},
{
"epoch": 0.21032863849765257,
"grad_norm": 0.18513290584087372,
"learning_rate": 0.0008,
"loss": 1.7203575372695923,
"step": 1008
},
{
"epoch": 0.21074595722483047,
"grad_norm": 0.22435349225997925,
"learning_rate": 0.0008,
"loss": 1.7020659446716309,
"step": 1010
},
{
"epoch": 0.21116327595200834,
"grad_norm": 0.2310158610343933,
"learning_rate": 0.0008,
"loss": 1.7132530212402344,
"step": 1012
},
{
"epoch": 0.21158059467918622,
"grad_norm": 0.1832805722951889,
"learning_rate": 0.0008,
"loss": 1.785457968711853,
"step": 1014
},
{
"epoch": 0.21199791340636412,
"grad_norm": 0.19769711792469025,
"learning_rate": 0.0008,
"loss": 1.711798906326294,
"step": 1016
},
{
"epoch": 0.212415232133542,
"grad_norm": 0.18251265585422516,
"learning_rate": 0.0008,
"loss": 1.651078462600708,
"step": 1018
},
{
"epoch": 0.21283255086071987,
"grad_norm": 0.166689932346344,
"learning_rate": 0.0008,
"loss": 1.8827983140945435,
"step": 1020
},
{
"epoch": 0.21324986958789777,
"grad_norm": 0.19592298567295074,
"learning_rate": 0.0008,
"loss": 1.7221782207489014,
"step": 1022
},
{
"epoch": 0.21366718831507564,
"grad_norm": 0.18124715983867645,
"learning_rate": 0.0008,
"loss": 1.661760926246643,
"step": 1024
},
{
"epoch": 0.2140845070422535,
"grad_norm": 0.20759986340999603,
"learning_rate": 0.0008,
"loss": 1.643500804901123,
"step": 1026
},
{
"epoch": 0.21450182576943141,
"grad_norm": 0.191911518573761,
"learning_rate": 0.0008,
"loss": 1.696365475654602,
"step": 1028
},
{
"epoch": 0.2149191444966093,
"grad_norm": 0.19675235450267792,
"learning_rate": 0.0008,
"loss": 1.7411773204803467,
"step": 1030
},
{
"epoch": 0.21533646322378716,
"grad_norm": 0.16785453259944916,
"learning_rate": 0.0008,
"loss": 1.8436439037322998,
"step": 1032
},
{
"epoch": 0.21575378195096506,
"grad_norm": 0.17273321747779846,
"learning_rate": 0.0008,
"loss": 1.6798033714294434,
"step": 1034
},
{
"epoch": 0.21617110067814294,
"grad_norm": 0.1928907185792923,
"learning_rate": 0.0008,
"loss": 1.7037917375564575,
"step": 1036
},
{
"epoch": 0.2165884194053208,
"grad_norm": 0.2089846283197403,
"learning_rate": 0.0008,
"loss": 1.5725737810134888,
"step": 1038
},
{
"epoch": 0.21700573813249868,
"grad_norm": 0.18144100904464722,
"learning_rate": 0.0008,
"loss": 1.615740418434143,
"step": 1040
},
{
"epoch": 0.21742305685967658,
"grad_norm": 0.18197332322597504,
"learning_rate": 0.0008,
"loss": 1.5837657451629639,
"step": 1042
},
{
"epoch": 0.21784037558685446,
"grad_norm": 0.18892040848731995,
"learning_rate": 0.0008,
"loss": 1.7325462102890015,
"step": 1044
},
{
"epoch": 0.21825769431403233,
"grad_norm": 0.19149884581565857,
"learning_rate": 0.0008,
"loss": 1.6922088861465454,
"step": 1046
},
{
"epoch": 0.21867501304121023,
"grad_norm": 0.21839947998523712,
"learning_rate": 0.0008,
"loss": 1.545508861541748,
"step": 1048
},
{
"epoch": 0.2190923317683881,
"grad_norm": 0.20385684072971344,
"learning_rate": 0.0008,
"loss": 1.6834397315979004,
"step": 1050
},
{
"epoch": 0.21950965049556598,
"grad_norm": 0.22142915427684784,
"learning_rate": 0.0008,
"loss": 1.7758318185806274,
"step": 1052
},
{
"epoch": 0.21992696922274388,
"grad_norm": 0.2012168914079666,
"learning_rate": 0.0008,
"loss": 1.6646374464035034,
"step": 1054
},
{
"epoch": 0.22034428794992175,
"grad_norm": 0.24923035502433777,
"learning_rate": 0.0008,
"loss": 1.824601173400879,
"step": 1056
},
{
"epoch": 0.22076160667709963,
"grad_norm": 0.18179042637348175,
"learning_rate": 0.0008,
"loss": 1.7717976570129395,
"step": 1058
},
{
"epoch": 0.22117892540427753,
"grad_norm": 0.2071542739868164,
"learning_rate": 0.0008,
"loss": 1.8223323822021484,
"step": 1060
},
{
"epoch": 0.2215962441314554,
"grad_norm": 0.18816639482975006,
"learning_rate": 0.0008,
"loss": 1.7220277786254883,
"step": 1062
},
{
"epoch": 0.22201356285863327,
"grad_norm": 0.1935083419084549,
"learning_rate": 0.0008,
"loss": 1.7538706064224243,
"step": 1064
},
{
"epoch": 0.22243088158581117,
"grad_norm": 0.2136363685131073,
"learning_rate": 0.0008,
"loss": 1.6352862119674683,
"step": 1066
},
{
"epoch": 0.22284820031298905,
"grad_norm": 0.20609316229820251,
"learning_rate": 0.0008,
"loss": 1.7321614027023315,
"step": 1068
},
{
"epoch": 0.22326551904016692,
"grad_norm": 0.16606248915195465,
"learning_rate": 0.0008,
"loss": 1.7904727458953857,
"step": 1070
},
{
"epoch": 0.22368283776734482,
"grad_norm": 0.16676722466945648,
"learning_rate": 0.0008,
"loss": 1.8159418106079102,
"step": 1072
},
{
"epoch": 0.2241001564945227,
"grad_norm": 0.16745080053806305,
"learning_rate": 0.0008,
"loss": 1.7739876508712769,
"step": 1074
},
{
"epoch": 0.22451747522170057,
"grad_norm": 0.15250159800052643,
"learning_rate": 0.0008,
"loss": 1.786299705505371,
"step": 1076
},
{
"epoch": 0.22493479394887847,
"grad_norm": 0.210290789604187,
"learning_rate": 0.0008,
"loss": 1.6917192935943604,
"step": 1078
},
{
"epoch": 0.22535211267605634,
"grad_norm": 0.2053413689136505,
"learning_rate": 0.0008,
"loss": 1.5421149730682373,
"step": 1080
},
{
"epoch": 0.22576943140323422,
"grad_norm": 0.18984360992908478,
"learning_rate": 0.0008,
"loss": 1.7947566509246826,
"step": 1082
},
{
"epoch": 0.2261867501304121,
"grad_norm": 0.1953098177909851,
"learning_rate": 0.0008,
"loss": 1.762155294418335,
"step": 1084
},
{
"epoch": 0.22660406885759,
"grad_norm": 0.18576960265636444,
"learning_rate": 0.0008,
"loss": 1.6483376026153564,
"step": 1086
},
{
"epoch": 0.22702138758476786,
"grad_norm": 0.17228055000305176,
"learning_rate": 0.0008,
"loss": 1.625029444694519,
"step": 1088
},
{
"epoch": 0.22743870631194574,
"grad_norm": 0.20845091342926025,
"learning_rate": 0.0008,
"loss": 1.8286375999450684,
"step": 1090
},
{
"epoch": 0.22785602503912364,
"grad_norm": 0.20916111767292023,
"learning_rate": 0.0008,
"loss": 1.703885793685913,
"step": 1092
},
{
"epoch": 0.2282733437663015,
"grad_norm": 0.19794872403144836,
"learning_rate": 0.0008,
"loss": 1.7797656059265137,
"step": 1094
},
{
"epoch": 0.22869066249347939,
"grad_norm": 0.1863049566745758,
"learning_rate": 0.0008,
"loss": 1.7136446237564087,
"step": 1096
},
{
"epoch": 0.2291079812206573,
"grad_norm": 0.17499655485153198,
"learning_rate": 0.0008,
"loss": 1.7133187055587769,
"step": 1098
},
{
"epoch": 0.22952529994783516,
"grad_norm": 0.22497563064098358,
"learning_rate": 0.0008,
"loss": 1.573231816291809,
"step": 1100
},
{
"epoch": 0.22994261867501303,
"grad_norm": 0.21501412987709045,
"learning_rate": 0.0008,
"loss": 1.6877411603927612,
"step": 1102
},
{
"epoch": 0.23035993740219093,
"grad_norm": 0.18949785828590393,
"learning_rate": 0.0008,
"loss": 1.723860502243042,
"step": 1104
},
{
"epoch": 0.2307772561293688,
"grad_norm": 0.15616081655025482,
"learning_rate": 0.0008,
"loss": 1.6869328022003174,
"step": 1106
},
{
"epoch": 0.23119457485654668,
"grad_norm": 0.1717623770236969,
"learning_rate": 0.0008,
"loss": 1.6576188802719116,
"step": 1108
},
{
"epoch": 0.23161189358372458,
"grad_norm": 0.19391900300979614,
"learning_rate": 0.0008,
"loss": 1.6833415031433105,
"step": 1110
},
{
"epoch": 0.23202921231090246,
"grad_norm": 0.17734292149543762,
"learning_rate": 0.0008,
"loss": 1.8352059125900269,
"step": 1112
},
{
"epoch": 0.23244653103808033,
"grad_norm": 0.18125556409358978,
"learning_rate": 0.0008,
"loss": 1.6645547151565552,
"step": 1114
},
{
"epoch": 0.23286384976525823,
"grad_norm": 0.1773635894060135,
"learning_rate": 0.0008,
"loss": 1.698193073272705,
"step": 1116
},
{
"epoch": 0.2332811684924361,
"grad_norm": 0.18629439175128937,
"learning_rate": 0.0008,
"loss": 1.632277250289917,
"step": 1118
},
{
"epoch": 0.23369848721961398,
"grad_norm": 0.1882934719324112,
"learning_rate": 0.0008,
"loss": 1.6378971338272095,
"step": 1120
},
{
"epoch": 0.23411580594679185,
"grad_norm": 0.21717925369739532,
"learning_rate": 0.0008,
"loss": 1.7084206342697144,
"step": 1122
},
{
"epoch": 0.23453312467396975,
"grad_norm": 0.173837810754776,
"learning_rate": 0.0008,
"loss": 1.7402987480163574,
"step": 1124
},
{
"epoch": 0.23495044340114762,
"grad_norm": 0.1985453963279724,
"learning_rate": 0.0008,
"loss": 1.6419814825057983,
"step": 1126
},
{
"epoch": 0.2353677621283255,
"grad_norm": 0.2158537209033966,
"learning_rate": 0.0008,
"loss": 1.7257808446884155,
"step": 1128
},
{
"epoch": 0.2357850808555034,
"grad_norm": 0.18577049672603607,
"learning_rate": 0.0008,
"loss": 1.8305696249008179,
"step": 1130
},
{
"epoch": 0.23620239958268127,
"grad_norm": 0.20353929698467255,
"learning_rate": 0.0008,
"loss": 1.682106852531433,
"step": 1132
},
{
"epoch": 0.23661971830985915,
"grad_norm": 0.1592322438955307,
"learning_rate": 0.0008,
"loss": 1.8037075996398926,
"step": 1134
},
{
"epoch": 0.23703703703703705,
"grad_norm": 0.2094297856092453,
"learning_rate": 0.0008,
"loss": 1.7751059532165527,
"step": 1136
},
{
"epoch": 0.23745435576421492,
"grad_norm": 0.16734875738620758,
"learning_rate": 0.0008,
"loss": 1.730952501296997,
"step": 1138
},
{
"epoch": 0.2378716744913928,
"grad_norm": 0.15643708407878876,
"learning_rate": 0.0008,
"loss": 1.622879147529602,
"step": 1140
},
{
"epoch": 0.2382889932185707,
"grad_norm": 0.18474481999874115,
"learning_rate": 0.0008,
"loss": 1.833251953125,
"step": 1142
},
{
"epoch": 0.23870631194574857,
"grad_norm": 0.15061011910438538,
"learning_rate": 0.0008,
"loss": 1.6784833669662476,
"step": 1144
},
{
"epoch": 0.23912363067292644,
"grad_norm": 0.23532772064208984,
"learning_rate": 0.0008,
"loss": 1.6097378730773926,
"step": 1146
},
{
"epoch": 0.23954094940010434,
"grad_norm": 0.21782603859901428,
"learning_rate": 0.0008,
"loss": 1.810443639755249,
"step": 1148
},
{
"epoch": 0.23995826812728221,
"grad_norm": 0.19933411478996277,
"learning_rate": 0.0008,
"loss": 1.7105321884155273,
"step": 1150
},
{
"epoch": 0.2403755868544601,
"grad_norm": 0.17027534544467926,
"learning_rate": 0.0008,
"loss": 1.7290586233139038,
"step": 1152
},
{
"epoch": 0.240792905581638,
"grad_norm": 0.16234391927719116,
"learning_rate": 0.0008,
"loss": 1.756253719329834,
"step": 1154
},
{
"epoch": 0.24121022430881586,
"grad_norm": 0.16337856650352478,
"learning_rate": 0.0008,
"loss": 1.7716472148895264,
"step": 1156
},
{
"epoch": 0.24162754303599374,
"grad_norm": 0.1832415759563446,
"learning_rate": 0.0008,
"loss": 1.7059361934661865,
"step": 1158
},
{
"epoch": 0.2420448617631716,
"grad_norm": 0.17876803874969482,
"learning_rate": 0.0008,
"loss": 1.7458018064498901,
"step": 1160
},
{
"epoch": 0.2424621804903495,
"grad_norm": 0.1841387152671814,
"learning_rate": 0.0008,
"loss": 1.6689091920852661,
"step": 1162
},
{
"epoch": 0.24287949921752738,
"grad_norm": 0.19834557175636292,
"learning_rate": 0.0008,
"loss": 1.6324760913848877,
"step": 1164
},
{
"epoch": 0.24329681794470526,
"grad_norm": 0.1947917491197586,
"learning_rate": 0.0008,
"loss": 1.6829209327697754,
"step": 1166
},
{
"epoch": 0.24371413667188316,
"grad_norm": 0.17691686749458313,
"learning_rate": 0.0008,
"loss": 1.790049433708191,
"step": 1168
},
{
"epoch": 0.24413145539906103,
"grad_norm": 0.16536809504032135,
"learning_rate": 0.0008,
"loss": 1.8083261251449585,
"step": 1170
},
{
"epoch": 0.2445487741262389,
"grad_norm": 0.21526339650154114,
"learning_rate": 0.0008,
"loss": 1.712167501449585,
"step": 1172
},
{
"epoch": 0.2449660928534168,
"grad_norm": 0.20896495878696442,
"learning_rate": 0.0008,
"loss": 1.696351170539856,
"step": 1174
},
{
"epoch": 0.24538341158059468,
"grad_norm": 0.20524796843528748,
"learning_rate": 0.0008,
"loss": 1.790766954421997,
"step": 1176
},
{
"epoch": 0.24580073030777255,
"grad_norm": 0.1848706603050232,
"learning_rate": 0.0008,
"loss": 1.654806137084961,
"step": 1178
},
{
"epoch": 0.24621804903495045,
"grad_norm": 0.15635696053504944,
"learning_rate": 0.0008,
"loss": 1.7692698240280151,
"step": 1180
},
{
"epoch": 0.24663536776212833,
"grad_norm": 0.18491347134113312,
"learning_rate": 0.0008,
"loss": 1.7738521099090576,
"step": 1182
},
{
"epoch": 0.2470526864893062,
"grad_norm": 0.17178966104984283,
"learning_rate": 0.0008,
"loss": 1.685498595237732,
"step": 1184
},
{
"epoch": 0.2474700052164841,
"grad_norm": 0.18646575510501862,
"learning_rate": 0.0008,
"loss": 1.6837756633758545,
"step": 1186
},
{
"epoch": 0.24788732394366197,
"grad_norm": 0.20478443801403046,
"learning_rate": 0.0008,
"loss": 1.7419281005859375,
"step": 1188
},
{
"epoch": 0.24830464267083985,
"grad_norm": 0.1943219006061554,
"learning_rate": 0.0008,
"loss": 1.6575567722320557,
"step": 1190
},
{
"epoch": 0.24872196139801775,
"grad_norm": 0.17080877721309662,
"learning_rate": 0.0008,
"loss": 1.66692316532135,
"step": 1192
},
{
"epoch": 0.24913928012519562,
"grad_norm": 0.19454547762870789,
"learning_rate": 0.0008,
"loss": 1.7892298698425293,
"step": 1194
},
{
"epoch": 0.2495565988523735,
"grad_norm": 0.15055574476718903,
"learning_rate": 0.0008,
"loss": 1.7876267433166504,
"step": 1196
},
{
"epoch": 0.24997391757955137,
"grad_norm": 0.2156514823436737,
"learning_rate": 0.0008,
"loss": 1.5266536474227905,
"step": 1198
},
{
"epoch": 0.25039123630672927,
"grad_norm": 0.20630790293216705,
"learning_rate": 0.0008,
"loss": 1.866032361984253,
"step": 1200
},
{
"epoch": 0.25080855503390714,
"grad_norm": 0.20811523497104645,
"learning_rate": 0.0008,
"loss": 1.7445225715637207,
"step": 1202
},
{
"epoch": 0.251225873761085,
"grad_norm": 0.17733268439769745,
"learning_rate": 0.0008,
"loss": 1.6907681226730347,
"step": 1204
},
{
"epoch": 0.2516431924882629,
"grad_norm": 0.18805359303951263,
"learning_rate": 0.0008,
"loss": 1.6510014533996582,
"step": 1206
},
{
"epoch": 0.2520605112154408,
"grad_norm": 0.178513303399086,
"learning_rate": 0.0008,
"loss": 1.54683518409729,
"step": 1208
},
{
"epoch": 0.2524778299426187,
"grad_norm": 0.20541603863239288,
"learning_rate": 0.0008,
"loss": 1.6216344833374023,
"step": 1210
},
{
"epoch": 0.25289514866979657,
"grad_norm": 0.18689969182014465,
"learning_rate": 0.0008,
"loss": 1.662935495376587,
"step": 1212
},
{
"epoch": 0.25331246739697444,
"grad_norm": 0.17605124413967133,
"learning_rate": 0.0008,
"loss": 1.7557682991027832,
"step": 1214
},
{
"epoch": 0.2537297861241523,
"grad_norm": 0.22767335176467896,
"learning_rate": 0.0008,
"loss": 1.6123594045639038,
"step": 1216
},
{
"epoch": 0.2541471048513302,
"grad_norm": 0.1697077602148056,
"learning_rate": 0.0008,
"loss": 1.7527412176132202,
"step": 1218
},
{
"epoch": 0.25456442357850806,
"grad_norm": 0.2071792483329773,
"learning_rate": 0.0008,
"loss": 1.710493803024292,
"step": 1220
},
{
"epoch": 0.254981742305686,
"grad_norm": 0.18517006933689117,
"learning_rate": 0.0008,
"loss": 1.7851550579071045,
"step": 1222
},
{
"epoch": 0.25539906103286386,
"grad_norm": 0.1847924441099167,
"learning_rate": 0.0008,
"loss": 1.6022294759750366,
"step": 1224
},
{
"epoch": 0.25581637976004173,
"grad_norm": 0.18159456551074982,
"learning_rate": 0.0008,
"loss": 1.6699817180633545,
"step": 1226
},
{
"epoch": 0.2562336984872196,
"grad_norm": 0.21658466756343842,
"learning_rate": 0.0008,
"loss": 1.8245761394500732,
"step": 1228
},
{
"epoch": 0.2566510172143975,
"grad_norm": 0.19825735688209534,
"learning_rate": 0.0008,
"loss": 1.7192044258117676,
"step": 1230
},
{
"epoch": 0.25706833594157535,
"grad_norm": 0.19651353359222412,
"learning_rate": 0.0008,
"loss": 1.6715978384017944,
"step": 1232
},
{
"epoch": 0.2574856546687533,
"grad_norm": 0.17961759865283966,
"learning_rate": 0.0008,
"loss": 1.6323541402816772,
"step": 1234
},
{
"epoch": 0.25790297339593116,
"grad_norm": 0.22247351706027985,
"learning_rate": 0.0008,
"loss": 1.673211932182312,
"step": 1236
},
{
"epoch": 0.25832029212310903,
"grad_norm": 0.15685665607452393,
"learning_rate": 0.0008,
"loss": 1.639365553855896,
"step": 1238
},
{
"epoch": 0.2587376108502869,
"grad_norm": 0.20231615006923676,
"learning_rate": 0.0008,
"loss": 1.6564083099365234,
"step": 1240
},
{
"epoch": 0.2591549295774648,
"grad_norm": 0.21895039081573486,
"learning_rate": 0.0008,
"loss": 1.7256592512130737,
"step": 1242
},
{
"epoch": 0.25957224830464265,
"grad_norm": 0.1818832904100418,
"learning_rate": 0.0008,
"loss": 1.5334216356277466,
"step": 1244
},
{
"epoch": 0.2599895670318206,
"grad_norm": 0.1985233575105667,
"learning_rate": 0.0008,
"loss": 1.635616660118103,
"step": 1246
},
{
"epoch": 0.26040688575899845,
"grad_norm": 0.17793521285057068,
"learning_rate": 0.0008,
"loss": 1.7438145875930786,
"step": 1248
},
{
"epoch": 0.2608242044861763,
"grad_norm": 0.15812844038009644,
"learning_rate": 0.0008,
"loss": 1.7429944276809692,
"step": 1250
},
{
"epoch": 0.2612415232133542,
"grad_norm": 0.1952613890171051,
"learning_rate": 0.0008,
"loss": 1.7084968090057373,
"step": 1252
},
{
"epoch": 0.26165884194053207,
"grad_norm": 0.18432091176509857,
"learning_rate": 0.0008,
"loss": 1.7054474353790283,
"step": 1254
},
{
"epoch": 0.26207616066770995,
"grad_norm": 0.178706094622612,
"learning_rate": 0.0008,
"loss": 1.644567608833313,
"step": 1256
},
{
"epoch": 0.2624934793948878,
"grad_norm": 0.18715743720531464,
"learning_rate": 0.0008,
"loss": 1.6602011919021606,
"step": 1258
},
{
"epoch": 0.26291079812206575,
"grad_norm": 0.19909054040908813,
"learning_rate": 0.0008,
"loss": 1.7505033016204834,
"step": 1260
},
{
"epoch": 0.2633281168492436,
"grad_norm": 0.1657245010137558,
"learning_rate": 0.0008,
"loss": 1.660007119178772,
"step": 1262
},
{
"epoch": 0.2637454355764215,
"grad_norm": 0.16391392052173615,
"learning_rate": 0.0008,
"loss": 1.6795694828033447,
"step": 1264
},
{
"epoch": 0.26416275430359937,
"grad_norm": 0.2146173119544983,
"learning_rate": 0.0008,
"loss": 1.7402732372283936,
"step": 1266
},
{
"epoch": 0.26458007303077724,
"grad_norm": 0.1988341510295868,
"learning_rate": 0.0008,
"loss": 1.755741834640503,
"step": 1268
},
{
"epoch": 0.2649973917579551,
"grad_norm": 0.17654849588871002,
"learning_rate": 0.0008,
"loss": 1.6526844501495361,
"step": 1270
},
{
"epoch": 0.26541471048513304,
"grad_norm": 0.17885872721672058,
"learning_rate": 0.0008,
"loss": 1.6377933025360107,
"step": 1272
},
{
"epoch": 0.2658320292123109,
"grad_norm": 0.17087270319461823,
"learning_rate": 0.0008,
"loss": 1.7254188060760498,
"step": 1274
},
{
"epoch": 0.2662493479394888,
"grad_norm": 0.18458805978298187,
"learning_rate": 0.0008,
"loss": 1.6448115110397339,
"step": 1276
},
{
"epoch": 0.26666666666666666,
"grad_norm": 0.16309542953968048,
"learning_rate": 0.0008,
"loss": 1.7670483589172363,
"step": 1278
},
{
"epoch": 0.26708398539384454,
"grad_norm": 0.2134781926870346,
"learning_rate": 0.0008,
"loss": 1.6897218227386475,
"step": 1280
},
{
"epoch": 0.2675013041210224,
"grad_norm": 0.17648102343082428,
"learning_rate": 0.0008,
"loss": 1.7560442686080933,
"step": 1282
},
{
"epoch": 0.26791862284820034,
"grad_norm": 0.21210359036922455,
"learning_rate": 0.0008,
"loss": 1.7754197120666504,
"step": 1284
},
{
"epoch": 0.2683359415753782,
"grad_norm": 0.1943667083978653,
"learning_rate": 0.0008,
"loss": 1.7837111949920654,
"step": 1286
},
{
"epoch": 0.2687532603025561,
"grad_norm": 0.1632717102766037,
"learning_rate": 0.0008,
"loss": 1.662658929824829,
"step": 1288
},
{
"epoch": 0.26917057902973396,
"grad_norm": 0.18424485623836517,
"learning_rate": 0.0008,
"loss": 1.731174349784851,
"step": 1290
},
{
"epoch": 0.26958789775691183,
"grad_norm": 0.19336508214473724,
"learning_rate": 0.0008,
"loss": 1.7278025150299072,
"step": 1292
},
{
"epoch": 0.2700052164840897,
"grad_norm": 0.21472306549549103,
"learning_rate": 0.0008,
"loss": 1.5542975664138794,
"step": 1294
},
{
"epoch": 0.2704225352112676,
"grad_norm": 0.1842101365327835,
"learning_rate": 0.0008,
"loss": 1.6445286273956299,
"step": 1296
},
{
"epoch": 0.2708398539384455,
"grad_norm": 0.19083024561405182,
"learning_rate": 0.0008,
"loss": 1.783835768699646,
"step": 1298
},
{
"epoch": 0.2712571726656234,
"grad_norm": 0.20647959411144257,
"learning_rate": 0.0008,
"loss": 1.732171893119812,
"step": 1300
},
{
"epoch": 0.27167449139280125,
"grad_norm": 0.1935419738292694,
"learning_rate": 0.0008,
"loss": 1.6313058137893677,
"step": 1302
},
{
"epoch": 0.2720918101199791,
"grad_norm": 0.1897238790988922,
"learning_rate": 0.0008,
"loss": 1.7380390167236328,
"step": 1304
},
{
"epoch": 0.272509128847157,
"grad_norm": 0.17351466417312622,
"learning_rate": 0.0008,
"loss": 1.6031750440597534,
"step": 1306
},
{
"epoch": 0.2729264475743349,
"grad_norm": 0.19890092313289642,
"learning_rate": 0.0008,
"loss": 1.7074499130249023,
"step": 1308
},
{
"epoch": 0.2733437663015128,
"grad_norm": 0.19245579838752747,
"learning_rate": 0.0008,
"loss": 1.5905141830444336,
"step": 1310
},
{
"epoch": 0.2737610850286907,
"grad_norm": 0.1777419000864029,
"learning_rate": 0.0008,
"loss": 1.710379719734192,
"step": 1312
},
{
"epoch": 0.27417840375586855,
"grad_norm": 0.19055373966693878,
"learning_rate": 0.0008,
"loss": 1.6207036972045898,
"step": 1314
},
{
"epoch": 0.2745957224830464,
"grad_norm": 0.2008567452430725,
"learning_rate": 0.0008,
"loss": 1.9208149909973145,
"step": 1316
},
{
"epoch": 0.2750130412102243,
"grad_norm": 0.18052977323532104,
"learning_rate": 0.0008,
"loss": 1.6596641540527344,
"step": 1318
},
{
"epoch": 0.27543035993740217,
"grad_norm": 0.18079730868339539,
"learning_rate": 0.0008,
"loss": 1.710309624671936,
"step": 1320
},
{
"epoch": 0.2758476786645801,
"grad_norm": 0.26781997084617615,
"learning_rate": 0.0008,
"loss": 1.548280119895935,
"step": 1322
},
{
"epoch": 0.27626499739175797,
"grad_norm": 0.1694943606853485,
"learning_rate": 0.0008,
"loss": 1.683698058128357,
"step": 1324
},
{
"epoch": 0.27668231611893584,
"grad_norm": 0.16780099272727966,
"learning_rate": 0.0008,
"loss": 1.7376729249954224,
"step": 1326
},
{
"epoch": 0.2770996348461137,
"grad_norm": 0.17869985103607178,
"learning_rate": 0.0008,
"loss": 1.763671636581421,
"step": 1328
},
{
"epoch": 0.2775169535732916,
"grad_norm": 0.1948542296886444,
"learning_rate": 0.0008,
"loss": 1.720281720161438,
"step": 1330
},
{
"epoch": 0.27793427230046946,
"grad_norm": 0.18689291179180145,
"learning_rate": 0.0008,
"loss": 1.726885437965393,
"step": 1332
},
{
"epoch": 0.27835159102764734,
"grad_norm": 0.22598035633563995,
"learning_rate": 0.0008,
"loss": 1.5073474645614624,
"step": 1334
},
{
"epoch": 0.27876890975482527,
"grad_norm": 0.18119733035564423,
"learning_rate": 0.0008,
"loss": 1.693723201751709,
"step": 1336
},
{
"epoch": 0.27918622848200314,
"grad_norm": 0.2110997885465622,
"learning_rate": 0.0008,
"loss": 1.7115660905838013,
"step": 1338
},
{
"epoch": 0.279603547209181,
"grad_norm": 0.16438189148902893,
"learning_rate": 0.0008,
"loss": 1.618546724319458,
"step": 1340
},
{
"epoch": 0.2800208659363589,
"grad_norm": 0.19227030873298645,
"learning_rate": 0.0008,
"loss": 1.786084771156311,
"step": 1342
},
{
"epoch": 0.28043818466353676,
"grad_norm": 0.1998574584722519,
"learning_rate": 0.0008,
"loss": 1.7135698795318604,
"step": 1344
},
{
"epoch": 0.28085550339071463,
"grad_norm": 0.20090679824352264,
"learning_rate": 0.0008,
"loss": 1.6880048513412476,
"step": 1346
},
{
"epoch": 0.28127282211789256,
"grad_norm": 0.1762552410364151,
"learning_rate": 0.0008,
"loss": 1.8207820653915405,
"step": 1348
},
{
"epoch": 0.28169014084507044,
"grad_norm": 0.17200258374214172,
"learning_rate": 0.0008,
"loss": 1.7109915018081665,
"step": 1350
},
{
"epoch": 0.2821074595722483,
"grad_norm": 0.26064231991767883,
"learning_rate": 0.0008,
"loss": 1.6104341745376587,
"step": 1352
},
{
"epoch": 0.2825247782994262,
"grad_norm": 0.19276520609855652,
"learning_rate": 0.0008,
"loss": 1.7332199811935425,
"step": 1354
},
{
"epoch": 0.28294209702660406,
"grad_norm": 0.18370100855827332,
"learning_rate": 0.0008,
"loss": 1.6367721557617188,
"step": 1356
},
{
"epoch": 0.28335941575378193,
"grad_norm": 0.2341209053993225,
"learning_rate": 0.0008,
"loss": 1.6277238130569458,
"step": 1358
},
{
"epoch": 0.28377673448095986,
"grad_norm": 0.2038310021162033,
"learning_rate": 0.0008,
"loss": 1.6990879774093628,
"step": 1360
},
{
"epoch": 0.28419405320813773,
"grad_norm": 0.15962599217891693,
"learning_rate": 0.0008,
"loss": 1.6954760551452637,
"step": 1362
},
{
"epoch": 0.2846113719353156,
"grad_norm": 0.19616512954235077,
"learning_rate": 0.0008,
"loss": 1.8433548212051392,
"step": 1364
},
{
"epoch": 0.2850286906624935,
"grad_norm": 0.16708795726299286,
"learning_rate": 0.0008,
"loss": 1.6375163793563843,
"step": 1366
},
{
"epoch": 0.28544600938967135,
"grad_norm": 0.17578016221523285,
"learning_rate": 0.0008,
"loss": 1.7065175771713257,
"step": 1368
},
{
"epoch": 0.2858633281168492,
"grad_norm": 0.25199878215789795,
"learning_rate": 0.0008,
"loss": 1.6407272815704346,
"step": 1370
},
{
"epoch": 0.28628064684402715,
"grad_norm": 0.18789313733577728,
"learning_rate": 0.0008,
"loss": 1.794905662536621,
"step": 1372
},
{
"epoch": 0.286697965571205,
"grad_norm": 0.18339446187019348,
"learning_rate": 0.0008,
"loss": 1.6440658569335938,
"step": 1374
},
{
"epoch": 0.2871152842983829,
"grad_norm": 0.17305827140808105,
"learning_rate": 0.0008,
"loss": 1.7367901802062988,
"step": 1376
},
{
"epoch": 0.2875326030255608,
"grad_norm": 0.21036109328269958,
"learning_rate": 0.0008,
"loss": 1.7592393159866333,
"step": 1378
},
{
"epoch": 0.28794992175273865,
"grad_norm": 0.2018461972475052,
"learning_rate": 0.0008,
"loss": 1.6359031200408936,
"step": 1380
},
{
"epoch": 0.2883672404799165,
"grad_norm": 0.21195334196090698,
"learning_rate": 0.0008,
"loss": 1.6179341077804565,
"step": 1382
},
{
"epoch": 0.2887845592070944,
"grad_norm": 0.19959783554077148,
"learning_rate": 0.0008,
"loss": 1.7731112241744995,
"step": 1384
},
{
"epoch": 0.2892018779342723,
"grad_norm": 0.19282294809818268,
"learning_rate": 0.0008,
"loss": 1.7109217643737793,
"step": 1386
},
{
"epoch": 0.2896191966614502,
"grad_norm": 0.17679518461227417,
"learning_rate": 0.0008,
"loss": 1.8279452323913574,
"step": 1388
},
{
"epoch": 0.29003651538862807,
"grad_norm": 0.1713707596063614,
"learning_rate": 0.0008,
"loss": 1.519986629486084,
"step": 1390
},
{
"epoch": 0.29045383411580594,
"grad_norm": 0.19550099968910217,
"learning_rate": 0.0008,
"loss": 1.6809767484664917,
"step": 1392
},
{
"epoch": 0.2908711528429838,
"grad_norm": 0.19049036502838135,
"learning_rate": 0.0008,
"loss": 1.6240314245224,
"step": 1394
},
{
"epoch": 0.2912884715701617,
"grad_norm": 0.18985776603221893,
"learning_rate": 0.0008,
"loss": 1.6918792724609375,
"step": 1396
},
{
"epoch": 0.2917057902973396,
"grad_norm": 0.20161029696464539,
"learning_rate": 0.0008,
"loss": 1.673765778541565,
"step": 1398
},
{
"epoch": 0.2921231090245175,
"grad_norm": 0.16617177426815033,
"learning_rate": 0.0008,
"loss": 1.7973628044128418,
"step": 1400
},
{
"epoch": 0.29254042775169536,
"grad_norm": 0.19058004021644592,
"learning_rate": 0.0008,
"loss": 1.7038174867630005,
"step": 1402
},
{
"epoch": 0.29295774647887324,
"grad_norm": 0.24385130405426025,
"learning_rate": 0.0008,
"loss": 1.6953752040863037,
"step": 1404
},
{
"epoch": 0.2933750652060511,
"grad_norm": 0.19995364546775818,
"learning_rate": 0.0008,
"loss": 1.7429099082946777,
"step": 1406
},
{
"epoch": 0.293792383933229,
"grad_norm": 0.20031686127185822,
"learning_rate": 0.0008,
"loss": 1.5738037824630737,
"step": 1408
},
{
"epoch": 0.2942097026604069,
"grad_norm": 0.22707857191562653,
"learning_rate": 0.0008,
"loss": 1.5364428758621216,
"step": 1410
},
{
"epoch": 0.2946270213875848,
"grad_norm": 0.19964051246643066,
"learning_rate": 0.0008,
"loss": 1.7416555881500244,
"step": 1412
},
{
"epoch": 0.29504434011476266,
"grad_norm": 0.19109734892845154,
"learning_rate": 0.0008,
"loss": 1.755820393562317,
"step": 1414
},
{
"epoch": 0.29546165884194053,
"grad_norm": 0.16195525228977203,
"learning_rate": 0.0008,
"loss": 1.7414896488189697,
"step": 1416
},
{
"epoch": 0.2958789775691184,
"grad_norm": 0.20762008428573608,
"learning_rate": 0.0008,
"loss": 1.641227126121521,
"step": 1418
},
{
"epoch": 0.2962962962962963,
"grad_norm": 0.2282640039920807,
"learning_rate": 0.0008,
"loss": 1.6108753681182861,
"step": 1420
},
{
"epoch": 0.29671361502347415,
"grad_norm": 0.18947291374206543,
"learning_rate": 0.0008,
"loss": 1.6883621215820312,
"step": 1422
},
{
"epoch": 0.2971309337506521,
"grad_norm": 0.19278880953788757,
"learning_rate": 0.0008,
"loss": 1.5904206037521362,
"step": 1424
},
{
"epoch": 0.29754825247782996,
"grad_norm": 0.19088414311408997,
"learning_rate": 0.0008,
"loss": 1.608932375907898,
"step": 1426
},
{
"epoch": 0.29796557120500783,
"grad_norm": 0.1806740164756775,
"learning_rate": 0.0008,
"loss": 1.4758405685424805,
"step": 1428
},
{
"epoch": 0.2983828899321857,
"grad_norm": 0.17433306574821472,
"learning_rate": 0.0008,
"loss": 1.7317631244659424,
"step": 1430
},
{
"epoch": 0.2988002086593636,
"grad_norm": 0.20986224710941315,
"learning_rate": 0.0008,
"loss": 1.6544620990753174,
"step": 1432
},
{
"epoch": 0.29921752738654145,
"grad_norm": 0.17693914473056793,
"learning_rate": 0.0008,
"loss": 1.7037383317947388,
"step": 1434
},
{
"epoch": 0.2996348461137194,
"grad_norm": 0.2005043476819992,
"learning_rate": 0.0008,
"loss": 1.6176735162734985,
"step": 1436
},
{
"epoch": 0.30005216484089725,
"grad_norm": 0.1743563711643219,
"learning_rate": 0.0008,
"loss": 1.661389708518982,
"step": 1438
},
{
"epoch": 0.3004694835680751,
"grad_norm": 0.18514449894428253,
"learning_rate": 0.0008,
"loss": 1.7333942651748657,
"step": 1440
},
{
"epoch": 0.300886802295253,
"grad_norm": 0.22628803551197052,
"learning_rate": 0.0008,
"loss": 1.6381453275680542,
"step": 1442
},
{
"epoch": 0.30130412102243087,
"grad_norm": 0.1918284296989441,
"learning_rate": 0.0008,
"loss": 1.6575736999511719,
"step": 1444
},
{
"epoch": 0.30172143974960874,
"grad_norm": 0.17359037697315216,
"learning_rate": 0.0008,
"loss": 1.655394196510315,
"step": 1446
},
{
"epoch": 0.3021387584767867,
"grad_norm": 0.17278282344341278,
"learning_rate": 0.0008,
"loss": 1.7414089441299438,
"step": 1448
},
{
"epoch": 0.30255607720396455,
"grad_norm": 0.19828394055366516,
"learning_rate": 0.0008,
"loss": 1.5962445735931396,
"step": 1450
},
{
"epoch": 0.3029733959311424,
"grad_norm": 0.17131860554218292,
"learning_rate": 0.0008,
"loss": 1.7640455961227417,
"step": 1452
},
{
"epoch": 0.3033907146583203,
"grad_norm": 0.16310322284698486,
"learning_rate": 0.0008,
"loss": 1.753563642501831,
"step": 1454
},
{
"epoch": 0.30380803338549817,
"grad_norm": 0.18980418145656586,
"learning_rate": 0.0008,
"loss": 1.671080470085144,
"step": 1456
},
{
"epoch": 0.30422535211267604,
"grad_norm": 0.20802569389343262,
"learning_rate": 0.0008,
"loss": 1.4919934272766113,
"step": 1458
},
{
"epoch": 0.3046426708398539,
"grad_norm": 0.18515172600746155,
"learning_rate": 0.0008,
"loss": 1.716540813446045,
"step": 1460
},
{
"epoch": 0.30505998956703184,
"grad_norm": 0.18607166409492493,
"learning_rate": 0.0008,
"loss": 1.6162197589874268,
"step": 1462
},
{
"epoch": 0.3054773082942097,
"grad_norm": 0.22768165171146393,
"learning_rate": 0.0008,
"loss": 1.7302019596099854,
"step": 1464
},
{
"epoch": 0.3058946270213876,
"grad_norm": 0.17985276877880096,
"learning_rate": 0.0008,
"loss": 1.6336026191711426,
"step": 1466
},
{
"epoch": 0.30631194574856546,
"grad_norm": 0.22725899517536163,
"learning_rate": 0.0008,
"loss": 1.5572943687438965,
"step": 1468
},
{
"epoch": 0.30672926447574334,
"grad_norm": 0.20522980391979218,
"learning_rate": 0.0008,
"loss": 1.642440676689148,
"step": 1470
},
{
"epoch": 0.3071465832029212,
"grad_norm": 0.18183737993240356,
"learning_rate": 0.0008,
"loss": 1.6298116445541382,
"step": 1472
},
{
"epoch": 0.30756390193009914,
"grad_norm": 0.22883403301239014,
"learning_rate": 0.0008,
"loss": 1.6762281656265259,
"step": 1474
},
{
"epoch": 0.307981220657277,
"grad_norm": 0.17760437726974487,
"learning_rate": 0.0008,
"loss": 1.6878924369812012,
"step": 1476
},
{
"epoch": 0.3083985393844549,
"grad_norm": 0.19906498491764069,
"learning_rate": 0.0008,
"loss": 1.6999027729034424,
"step": 1478
},
{
"epoch": 0.30881585811163276,
"grad_norm": 0.21793927252292633,
"learning_rate": 0.0008,
"loss": 1.7294009923934937,
"step": 1480
},
{
"epoch": 0.30923317683881063,
"grad_norm": 0.1849290132522583,
"learning_rate": 0.0008,
"loss": 1.7409818172454834,
"step": 1482
},
{
"epoch": 0.3096504955659885,
"grad_norm": 0.18612521886825562,
"learning_rate": 0.0008,
"loss": 1.6981582641601562,
"step": 1484
},
{
"epoch": 0.31006781429316643,
"grad_norm": 0.1877032220363617,
"learning_rate": 0.0008,
"loss": 1.6937371492385864,
"step": 1486
},
{
"epoch": 0.3104851330203443,
"grad_norm": 0.1802709549665451,
"learning_rate": 0.0008,
"loss": 1.604067087173462,
"step": 1488
},
{
"epoch": 0.3109024517475222,
"grad_norm": 0.21858984231948853,
"learning_rate": 0.0008,
"loss": 1.6908204555511475,
"step": 1490
},
{
"epoch": 0.31131977047470005,
"grad_norm": 0.19200074672698975,
"learning_rate": 0.0008,
"loss": 1.694463849067688,
"step": 1492
},
{
"epoch": 0.3117370892018779,
"grad_norm": 0.16088292002677917,
"learning_rate": 0.0008,
"loss": 1.6398215293884277,
"step": 1494
},
{
"epoch": 0.3121544079290558,
"grad_norm": 0.18364711105823517,
"learning_rate": 0.0008,
"loss": 1.6783623695373535,
"step": 1496
},
{
"epoch": 0.3125717266562337,
"grad_norm": 0.19473858177661896,
"learning_rate": 0.0008,
"loss": 1.6297110319137573,
"step": 1498
},
{
"epoch": 0.3129890453834116,
"grad_norm": 0.1612580418586731,
"learning_rate": 0.0008,
"loss": 1.7179768085479736,
"step": 1500
},
{
"epoch": 0.3134063641105895,
"grad_norm": 0.23626890778541565,
"learning_rate": 0.0008,
"loss": 1.6406357288360596,
"step": 1502
},
{
"epoch": 0.31382368283776735,
"grad_norm": 0.21286119520664215,
"learning_rate": 0.0008,
"loss": 1.663693904876709,
"step": 1504
},
{
"epoch": 0.3142410015649452,
"grad_norm": 0.16071361303329468,
"learning_rate": 0.0008,
"loss": 1.7518222332000732,
"step": 1506
},
{
"epoch": 0.3146583202921231,
"grad_norm": 0.194560244679451,
"learning_rate": 0.0008,
"loss": 1.7725356817245483,
"step": 1508
},
{
"epoch": 0.31507563901930097,
"grad_norm": 0.19977954030036926,
"learning_rate": 0.0008,
"loss": 1.699697494506836,
"step": 1510
},
{
"epoch": 0.3154929577464789,
"grad_norm": 0.1823112517595291,
"learning_rate": 0.0008,
"loss": 1.7693278789520264,
"step": 1512
},
{
"epoch": 0.31591027647365677,
"grad_norm": 0.2436540573835373,
"learning_rate": 0.0008,
"loss": 1.507657766342163,
"step": 1514
},
{
"epoch": 0.31632759520083464,
"grad_norm": 0.18154767155647278,
"learning_rate": 0.0008,
"loss": 1.708625078201294,
"step": 1516
},
{
"epoch": 0.3167449139280125,
"grad_norm": 0.18343544006347656,
"learning_rate": 0.0008,
"loss": 1.6040462255477905,
"step": 1518
},
{
"epoch": 0.3171622326551904,
"grad_norm": 0.18127363920211792,
"learning_rate": 0.0008,
"loss": 1.6993781328201294,
"step": 1520
},
{
"epoch": 0.31757955138236826,
"grad_norm": 0.20277267694473267,
"learning_rate": 0.0008,
"loss": 1.7215282917022705,
"step": 1522
},
{
"epoch": 0.3179968701095462,
"grad_norm": 0.1993628740310669,
"learning_rate": 0.0008,
"loss": 1.5789881944656372,
"step": 1524
},
{
"epoch": 0.31841418883672407,
"grad_norm": 0.18073564767837524,
"learning_rate": 0.0008,
"loss": 1.6811549663543701,
"step": 1526
},
{
"epoch": 0.31883150756390194,
"grad_norm": 0.19242148101329803,
"learning_rate": 0.0008,
"loss": 1.6774734258651733,
"step": 1528
},
{
"epoch": 0.3192488262910798,
"grad_norm": 0.15729664266109467,
"learning_rate": 0.0008,
"loss": 1.6646878719329834,
"step": 1530
},
{
"epoch": 0.3196661450182577,
"grad_norm": 0.18678613007068634,
"learning_rate": 0.0008,
"loss": 1.6796011924743652,
"step": 1532
},
{
"epoch": 0.32008346374543556,
"grad_norm": 0.17345048487186432,
"learning_rate": 0.0008,
"loss": 1.6147273778915405,
"step": 1534
},
{
"epoch": 0.32050078247261343,
"grad_norm": 0.20197319984436035,
"learning_rate": 0.0008,
"loss": 1.6092185974121094,
"step": 1536
},
{
"epoch": 0.32091810119979136,
"grad_norm": 0.17410239577293396,
"learning_rate": 0.0008,
"loss": 1.815969705581665,
"step": 1538
},
{
"epoch": 0.32133541992696923,
"grad_norm": 0.18178392946720123,
"learning_rate": 0.0008,
"loss": 1.7562639713287354,
"step": 1540
},
{
"epoch": 0.3217527386541471,
"grad_norm": 0.18078863620758057,
"learning_rate": 0.0008,
"loss": 1.5952532291412354,
"step": 1542
},
{
"epoch": 0.322170057381325,
"grad_norm": 0.17517058551311493,
"learning_rate": 0.0008,
"loss": 1.6517339944839478,
"step": 1544
},
{
"epoch": 0.32258737610850285,
"grad_norm": 0.17974406480789185,
"learning_rate": 0.0008,
"loss": 1.6895688772201538,
"step": 1546
},
{
"epoch": 0.32300469483568073,
"grad_norm": 0.22083960473537445,
"learning_rate": 0.0008,
"loss": 1.7174173593521118,
"step": 1548
},
{
"epoch": 0.32342201356285866,
"grad_norm": 0.19608071446418762,
"learning_rate": 0.0008,
"loss": 1.4775753021240234,
"step": 1550
},
{
"epoch": 0.32383933229003653,
"grad_norm": 0.1948099136352539,
"learning_rate": 0.0008,
"loss": 1.6510119438171387,
"step": 1552
},
{
"epoch": 0.3242566510172144,
"grad_norm": 0.1592712551355362,
"learning_rate": 0.0008,
"loss": 1.7928534746170044,
"step": 1554
},
{
"epoch": 0.3246739697443923,
"grad_norm": 0.17988751828670502,
"learning_rate": 0.0008,
"loss": 1.490371584892273,
"step": 1556
},
{
"epoch": 0.32509128847157015,
"grad_norm": 0.15834231674671173,
"learning_rate": 0.0008,
"loss": 1.6932299137115479,
"step": 1558
},
{
"epoch": 0.325508607198748,
"grad_norm": 0.20159009099006653,
"learning_rate": 0.0008,
"loss": 1.6639045476913452,
"step": 1560
},
{
"epoch": 0.32592592592592595,
"grad_norm": 0.20743347704410553,
"learning_rate": 0.0008,
"loss": 1.7608636617660522,
"step": 1562
},
{
"epoch": 0.3263432446531038,
"grad_norm": 0.19364625215530396,
"learning_rate": 0.0008,
"loss": 1.6866730451583862,
"step": 1564
},
{
"epoch": 0.3267605633802817,
"grad_norm": 0.16306082904338837,
"learning_rate": 0.0008,
"loss": 1.6245598793029785,
"step": 1566
},
{
"epoch": 0.32717788210745957,
"grad_norm": 0.21652166545391083,
"learning_rate": 0.0008,
"loss": 1.711512565612793,
"step": 1568
},
{
"epoch": 0.32759520083463745,
"grad_norm": 0.20035091042518616,
"learning_rate": 0.0008,
"loss": 1.6806864738464355,
"step": 1570
},
{
"epoch": 0.3280125195618153,
"grad_norm": 0.1903257668018341,
"learning_rate": 0.0008,
"loss": 1.6966125965118408,
"step": 1572
},
{
"epoch": 0.3284298382889932,
"grad_norm": 0.19555501639842987,
"learning_rate": 0.0008,
"loss": 1.7105661630630493,
"step": 1574
},
{
"epoch": 0.3288471570161711,
"grad_norm": 0.20062613487243652,
"learning_rate": 0.0008,
"loss": 1.6041440963745117,
"step": 1576
},
{
"epoch": 0.329264475743349,
"grad_norm": 0.18546223640441895,
"learning_rate": 0.0008,
"loss": 1.7582355737686157,
"step": 1578
},
{
"epoch": 0.32968179447052687,
"grad_norm": 0.18299883604049683,
"learning_rate": 0.0008,
"loss": 1.7387192249298096,
"step": 1580
},
{
"epoch": 0.33009911319770474,
"grad_norm": 0.17446210980415344,
"learning_rate": 0.0008,
"loss": 1.7612531185150146,
"step": 1582
},
{
"epoch": 0.3305164319248826,
"grad_norm": 0.16974785923957825,
"learning_rate": 0.0007997508564310183,
"loss": 1.7209926843643188,
"step": 1584
},
{
"epoch": 0.3309337506520605,
"grad_norm": 0.21876934170722961,
"learning_rate": 0.0007992525692930552,
"loss": 1.6248265504837036,
"step": 1586
},
{
"epoch": 0.3313510693792384,
"grad_norm": 0.20953874289989471,
"learning_rate": 0.0007987542821550919,
"loss": 1.6179856061935425,
"step": 1588
},
{
"epoch": 0.3317683881064163,
"grad_norm": 0.18122386932373047,
"learning_rate": 0.0007982559950171286,
"loss": 1.6957544088363647,
"step": 1590
},
{
"epoch": 0.33218570683359416,
"grad_norm": 0.19939370453357697,
"learning_rate": 0.0007977577078791654,
"loss": 1.6774882078170776,
"step": 1592
},
{
"epoch": 0.33260302556077204,
"grad_norm": 0.20314426720142365,
"learning_rate": 0.0007972594207412021,
"loss": 1.7296922206878662,
"step": 1594
},
{
"epoch": 0.3330203442879499,
"grad_norm": 0.1788739711046219,
"learning_rate": 0.0007967611336032388,
"loss": 1.5888195037841797,
"step": 1596
},
{
"epoch": 0.3334376630151278,
"grad_norm": 0.18745437264442444,
"learning_rate": 0.0007962628464652756,
"loss": 1.7956527471542358,
"step": 1598
},
{
"epoch": 0.3338549817423057,
"grad_norm": 0.20792031288146973,
"learning_rate": 0.0007957645593273124,
"loss": 1.7094107866287231,
"step": 1600
},
{
"epoch": 0.3342723004694836,
"grad_norm": 0.19007836282253265,
"learning_rate": 0.0007952662721893492,
"loss": 1.6883479356765747,
"step": 1602
},
{
"epoch": 0.33468961919666146,
"grad_norm": 0.19692915678024292,
"learning_rate": 0.0007947679850513859,
"loss": 1.6604087352752686,
"step": 1604
},
{
"epoch": 0.33510693792383933,
"grad_norm": 0.1727243959903717,
"learning_rate": 0.0007942696979134226,
"loss": 1.717360258102417,
"step": 1606
},
{
"epoch": 0.3355242566510172,
"grad_norm": 0.17717720568180084,
"learning_rate": 0.0007937714107754594,
"loss": 1.7308374643325806,
"step": 1608
},
{
"epoch": 0.3359415753781951,
"grad_norm": 0.1923617273569107,
"learning_rate": 0.0007932731236374961,
"loss": 1.714017629623413,
"step": 1610
},
{
"epoch": 0.33635889410537295,
"grad_norm": 0.18227413296699524,
"learning_rate": 0.0007927748364995328,
"loss": 1.7420536279678345,
"step": 1612
},
{
"epoch": 0.3367762128325509,
"grad_norm": 0.1786462366580963,
"learning_rate": 0.0007922765493615697,
"loss": 1.7462730407714844,
"step": 1614
},
{
"epoch": 0.33719353155972875,
"grad_norm": 0.27649492025375366,
"learning_rate": 0.0007917782622236064,
"loss": 1.5927529335021973,
"step": 1616
},
{
"epoch": 0.3376108502869066,
"grad_norm": 0.18073488771915436,
"learning_rate": 0.0007912799750856432,
"loss": 1.655815601348877,
"step": 1618
},
{
"epoch": 0.3380281690140845,
"grad_norm": 0.18013472855091095,
"learning_rate": 0.0007907816879476799,
"loss": 1.750601053237915,
"step": 1620
},
{
"epoch": 0.3384454877412624,
"grad_norm": 0.1647185981273651,
"learning_rate": 0.0007902834008097166,
"loss": 1.7260658740997314,
"step": 1622
},
{
"epoch": 0.33886280646844025,
"grad_norm": 0.19831311702728271,
"learning_rate": 0.0007897851136717534,
"loss": 1.6708985567092896,
"step": 1624
},
{
"epoch": 0.3392801251956182,
"grad_norm": 0.17761173844337463,
"learning_rate": 0.0007892868265337901,
"loss": 1.6845767498016357,
"step": 1626
},
{
"epoch": 0.33969744392279605,
"grad_norm": 0.21154843270778656,
"learning_rate": 0.0007887885393958268,
"loss": 1.6907172203063965,
"step": 1628
},
{
"epoch": 0.3401147626499739,
"grad_norm": 0.18469585478305817,
"learning_rate": 0.0007882902522578637,
"loss": 1.635585904121399,
"step": 1630
},
{
"epoch": 0.3405320813771518,
"grad_norm": 0.17891524732112885,
"learning_rate": 0.0007877919651199004,
"loss": 1.8234612941741943,
"step": 1632
},
{
"epoch": 0.34094940010432967,
"grad_norm": 0.203752800822258,
"learning_rate": 0.0007872936779819372,
"loss": 1.8275575637817383,
"step": 1634
},
{
"epoch": 0.34136671883150754,
"grad_norm": 0.19048547744750977,
"learning_rate": 0.0007867953908439739,
"loss": 1.685380220413208,
"step": 1636
},
{
"epoch": 0.34178403755868547,
"grad_norm": 0.19416184723377228,
"learning_rate": 0.0007862971037060106,
"loss": 1.661330223083496,
"step": 1638
},
{
"epoch": 0.34220135628586335,
"grad_norm": 0.15870237350463867,
"learning_rate": 0.0007857988165680474,
"loss": 1.6118049621582031,
"step": 1640
},
{
"epoch": 0.3426186750130412,
"grad_norm": 0.163543239235878,
"learning_rate": 0.0007853005294300841,
"loss": 1.747663140296936,
"step": 1642
},
{
"epoch": 0.3430359937402191,
"grad_norm": 0.17139224708080292,
"learning_rate": 0.0007848022422921209,
"loss": 1.6567996740341187,
"step": 1644
},
{
"epoch": 0.34345331246739697,
"grad_norm": 0.18835851550102234,
"learning_rate": 0.0007843039551541577,
"loss": 1.539109468460083,
"step": 1646
},
{
"epoch": 0.34387063119457484,
"grad_norm": 0.23088303208351135,
"learning_rate": 0.0007838056680161944,
"loss": 1.7184960842132568,
"step": 1648
},
{
"epoch": 0.3442879499217527,
"grad_norm": 0.16237686574459076,
"learning_rate": 0.0007833073808782311,
"loss": 1.5384950637817383,
"step": 1650
},
{
"epoch": 0.34470526864893064,
"grad_norm": 0.2083551287651062,
"learning_rate": 0.0007828090937402679,
"loss": 1.7102184295654297,
"step": 1652
},
{
"epoch": 0.3451225873761085,
"grad_norm": 0.16917714476585388,
"learning_rate": 0.0007823108066023046,
"loss": 1.6713474988937378,
"step": 1654
},
{
"epoch": 0.3455399061032864,
"grad_norm": 0.15906453132629395,
"learning_rate": 0.0007818125194643414,
"loss": 1.6190345287322998,
"step": 1656
},
{
"epoch": 0.34595722483046426,
"grad_norm": 0.24151946604251862,
"learning_rate": 0.0007813142323263782,
"loss": 1.679801344871521,
"step": 1658
},
{
"epoch": 0.34637454355764213,
"grad_norm": 0.2049770951271057,
"learning_rate": 0.0007808159451884149,
"loss": 1.554489016532898,
"step": 1660
},
{
"epoch": 0.34679186228482,
"grad_norm": 0.1780240833759308,
"learning_rate": 0.0007803176580504517,
"loss": 1.7561157941818237,
"step": 1662
},
{
"epoch": 0.34720918101199794,
"grad_norm": 0.1903304159641266,
"learning_rate": 0.0007798193709124884,
"loss": 1.621431827545166,
"step": 1664
},
{
"epoch": 0.3476264997391758,
"grad_norm": 0.17454200983047485,
"learning_rate": 0.0007793210837745251,
"loss": 1.6879749298095703,
"step": 1666
},
{
"epoch": 0.3480438184663537,
"grad_norm": 0.16695593297481537,
"learning_rate": 0.0007788227966365618,
"loss": 1.6259939670562744,
"step": 1668
},
{
"epoch": 0.34846113719353156,
"grad_norm": 0.1823844164609909,
"learning_rate": 0.0007783245094985985,
"loss": 1.7128453254699707,
"step": 1670
},
{
"epoch": 0.34887845592070943,
"grad_norm": 0.16730143129825592,
"learning_rate": 0.0007778262223606353,
"loss": 1.6664397716522217,
"step": 1672
},
{
"epoch": 0.3492957746478873,
"grad_norm": 0.21896038949489594,
"learning_rate": 0.0007773279352226721,
"loss": 1.6306133270263672,
"step": 1674
},
{
"epoch": 0.34971309337506523,
"grad_norm": 0.17780514061450958,
"learning_rate": 0.0007768296480847088,
"loss": 1.6109522581100464,
"step": 1676
},
{
"epoch": 0.3501304121022431,
"grad_norm": 0.2353464514017105,
"learning_rate": 0.0007763313609467456,
"loss": 1.675118327140808,
"step": 1678
},
{
"epoch": 0.350547730829421,
"grad_norm": 0.1789163202047348,
"learning_rate": 0.0007758330738087823,
"loss": 1.6949915885925293,
"step": 1680
},
{
"epoch": 0.35096504955659885,
"grad_norm": 0.21965782344341278,
"learning_rate": 0.0007753347866708191,
"loss": 1.831260085105896,
"step": 1682
},
{
"epoch": 0.3513823682837767,
"grad_norm": 0.1997511088848114,
"learning_rate": 0.0007748364995328558,
"loss": 1.6694724559783936,
"step": 1684
},
{
"epoch": 0.3517996870109546,
"grad_norm": 0.20894435048103333,
"learning_rate": 0.0007743382123948925,
"loss": 1.6326568126678467,
"step": 1686
},
{
"epoch": 0.35221700573813247,
"grad_norm": 0.19318808615207672,
"learning_rate": 0.0007738399252569294,
"loss": 1.8593822717666626,
"step": 1688
},
{
"epoch": 0.3526343244653104,
"grad_norm": 0.1976274996995926,
"learning_rate": 0.0007733416381189661,
"loss": 1.6631479263305664,
"step": 1690
},
{
"epoch": 0.3530516431924883,
"grad_norm": 0.177598774433136,
"learning_rate": 0.0007728433509810028,
"loss": 1.7659837007522583,
"step": 1692
},
{
"epoch": 0.35346896191966615,
"grad_norm": 0.21398799121379852,
"learning_rate": 0.0007723450638430396,
"loss": 1.6660094261169434,
"step": 1694
},
{
"epoch": 0.353886280646844,
"grad_norm": 0.1718406230211258,
"learning_rate": 0.0007718467767050763,
"loss": 1.7130343914031982,
"step": 1696
},
{
"epoch": 0.3543035993740219,
"grad_norm": 0.18095459043979645,
"learning_rate": 0.0007713484895671131,
"loss": 1.5778597593307495,
"step": 1698
},
{
"epoch": 0.35472091810119977,
"grad_norm": 0.17821018397808075,
"learning_rate": 0.0007708502024291498,
"loss": 1.60727858543396,
"step": 1700
},
{
"epoch": 0.3551382368283777,
"grad_norm": 0.17765848338603973,
"learning_rate": 0.0007703519152911865,
"loss": 1.7595034837722778,
"step": 1702
},
{
"epoch": 0.35555555555555557,
"grad_norm": 0.2136840969324112,
"learning_rate": 0.0007698536281532234,
"loss": 1.7453587055206299,
"step": 1704
},
{
"epoch": 0.35597287428273344,
"grad_norm": 0.19191712141036987,
"learning_rate": 0.0007693553410152601,
"loss": 1.6884596347808838,
"step": 1706
},
{
"epoch": 0.3563901930099113,
"grad_norm": 0.199738547205925,
"learning_rate": 0.0007688570538772968,
"loss": 1.6309882402420044,
"step": 1708
},
{
"epoch": 0.3568075117370892,
"grad_norm": 0.16454599797725677,
"learning_rate": 0.0007683587667393336,
"loss": 1.6823391914367676,
"step": 1710
},
{
"epoch": 0.35722483046426706,
"grad_norm": 0.20531423389911652,
"learning_rate": 0.0007678604796013703,
"loss": 1.6628015041351318,
"step": 1712
},
{
"epoch": 0.357642149191445,
"grad_norm": 0.1935228407382965,
"learning_rate": 0.000767362192463407,
"loss": 1.7234675884246826,
"step": 1714
},
{
"epoch": 0.35805946791862286,
"grad_norm": 0.18299990892410278,
"learning_rate": 0.0007668639053254439,
"loss": 1.8582141399383545,
"step": 1716
},
{
"epoch": 0.35847678664580074,
"grad_norm": 0.18308095633983612,
"learning_rate": 0.0007663656181874806,
"loss": 1.7067391872406006,
"step": 1718
},
{
"epoch": 0.3588941053729786,
"grad_norm": 0.21962682902812958,
"learning_rate": 0.0007658673310495174,
"loss": 1.7675175666809082,
"step": 1720
},
{
"epoch": 0.3593114241001565,
"grad_norm": 0.18697305023670197,
"learning_rate": 0.0007653690439115541,
"loss": 1.7001218795776367,
"step": 1722
},
{
"epoch": 0.35972874282733436,
"grad_norm": 0.20427927374839783,
"learning_rate": 0.0007648707567735908,
"loss": 1.5238335132598877,
"step": 1724
},
{
"epoch": 0.36014606155451223,
"grad_norm": 0.17253802716732025,
"learning_rate": 0.0007643724696356276,
"loss": 1.6772478818893433,
"step": 1726
},
{
"epoch": 0.36056338028169016,
"grad_norm": 0.1966213881969452,
"learning_rate": 0.0007638741824976643,
"loss": 1.5761340856552124,
"step": 1728
},
{
"epoch": 0.36098069900886803,
"grad_norm": 0.1600823998451233,
"learning_rate": 0.000763375895359701,
"loss": 1.7040597200393677,
"step": 1730
},
{
"epoch": 0.3613980177360459,
"grad_norm": 0.2740441560745239,
"learning_rate": 0.0007628776082217379,
"loss": 1.5225825309753418,
"step": 1732
},
{
"epoch": 0.3618153364632238,
"grad_norm": 0.210254967212677,
"learning_rate": 0.0007623793210837746,
"loss": 1.5670322179794312,
"step": 1734
},
{
"epoch": 0.36223265519040165,
"grad_norm": 0.1629776507616043,
"learning_rate": 0.0007618810339458114,
"loss": 1.5454801321029663,
"step": 1736
},
{
"epoch": 0.3626499739175795,
"grad_norm": 0.1893051415681839,
"learning_rate": 0.0007613827468078481,
"loss": 1.6624454259872437,
"step": 1738
},
{
"epoch": 0.36306729264475746,
"grad_norm": 0.2080356478691101,
"learning_rate": 0.0007608844596698848,
"loss": 1.6686328649520874,
"step": 1740
},
{
"epoch": 0.36348461137193533,
"grad_norm": 0.2037697583436966,
"learning_rate": 0.0007603861725319216,
"loss": 1.710297703742981,
"step": 1742
},
{
"epoch": 0.3639019300991132,
"grad_norm": 0.19695091247558594,
"learning_rate": 0.0007598878853939583,
"loss": 1.743046760559082,
"step": 1744
},
{
"epoch": 0.3643192488262911,
"grad_norm": 0.19401082396507263,
"learning_rate": 0.000759389598255995,
"loss": 1.6581921577453613,
"step": 1746
},
{
"epoch": 0.36473656755346895,
"grad_norm": 0.1991203874349594,
"learning_rate": 0.0007588913111180319,
"loss": 1.6929634809494019,
"step": 1748
},
{
"epoch": 0.3651538862806468,
"grad_norm": 0.1913372427225113,
"learning_rate": 0.0007583930239800685,
"loss": 1.6746728420257568,
"step": 1750
},
{
"epoch": 0.36557120500782475,
"grad_norm": 0.17644254863262177,
"learning_rate": 0.0007578947368421053,
"loss": 1.7083065509796143,
"step": 1752
},
{
"epoch": 0.3659885237350026,
"grad_norm": 0.16979239881038666,
"learning_rate": 0.000757396449704142,
"loss": 1.7234247922897339,
"step": 1754
},
{
"epoch": 0.3664058424621805,
"grad_norm": 0.17721712589263916,
"learning_rate": 0.0007568981625661787,
"loss": 1.6517670154571533,
"step": 1756
},
{
"epoch": 0.36682316118935837,
"grad_norm": 0.17361366748809814,
"learning_rate": 0.0007563998754282155,
"loss": 1.7030823230743408,
"step": 1758
},
{
"epoch": 0.36724047991653624,
"grad_norm": 0.18775466084480286,
"learning_rate": 0.0007559015882902522,
"loss": 1.7514017820358276,
"step": 1760
},
{
"epoch": 0.3676577986437141,
"grad_norm": 0.16860514879226685,
"learning_rate": 0.0007554033011522891,
"loss": 1.77925705909729,
"step": 1762
},
{
"epoch": 0.36807511737089205,
"grad_norm": 0.22899678349494934,
"learning_rate": 0.0007549050140143258,
"loss": 1.6921634674072266,
"step": 1764
},
{
"epoch": 0.3684924360980699,
"grad_norm": 0.1936270296573639,
"learning_rate": 0.0007544067268763625,
"loss": 1.657052755355835,
"step": 1766
},
{
"epoch": 0.3689097548252478,
"grad_norm": 0.1795278787612915,
"learning_rate": 0.0007539084397383993,
"loss": 1.5817052125930786,
"step": 1768
},
{
"epoch": 0.36932707355242567,
"grad_norm": 0.17500486969947815,
"learning_rate": 0.000753410152600436,
"loss": 1.676497220993042,
"step": 1770
},
{
"epoch": 0.36974439227960354,
"grad_norm": 0.1667417585849762,
"learning_rate": 0.0007529118654624727,
"loss": 1.6697306632995605,
"step": 1772
},
{
"epoch": 0.3701617110067814,
"grad_norm": 0.15290115773677826,
"learning_rate": 0.0007524135783245095,
"loss": 1.6589303016662598,
"step": 1774
},
{
"epoch": 0.3705790297339593,
"grad_norm": 0.19947949051856995,
"learning_rate": 0.0007519152911865463,
"loss": 1.6404383182525635,
"step": 1776
},
{
"epoch": 0.3709963484611372,
"grad_norm": 0.1735069751739502,
"learning_rate": 0.000751417004048583,
"loss": 1.777587652206421,
"step": 1778
},
{
"epoch": 0.3714136671883151,
"grad_norm": 0.20354051887989044,
"learning_rate": 0.0007509187169106198,
"loss": 1.6576002836227417,
"step": 1780
},
{
"epoch": 0.37183098591549296,
"grad_norm": 0.21735714375972748,
"learning_rate": 0.0007504204297726565,
"loss": 1.6496496200561523,
"step": 1782
},
{
"epoch": 0.37224830464267084,
"grad_norm": 0.16551290452480316,
"learning_rate": 0.0007499221426346933,
"loss": 1.6653001308441162,
"step": 1784
},
{
"epoch": 0.3726656233698487,
"grad_norm": 0.20047254860401154,
"learning_rate": 0.00074942385549673,
"loss": 1.6044127941131592,
"step": 1786
},
{
"epoch": 0.3730829420970266,
"grad_norm": 0.18200041353702545,
"learning_rate": 0.0007489255683587667,
"loss": 1.7514146566390991,
"step": 1788
},
{
"epoch": 0.3735002608242045,
"grad_norm": 0.17506684362888336,
"learning_rate": 0.0007484272812208036,
"loss": 1.637345790863037,
"step": 1790
},
{
"epoch": 0.3739175795513824,
"grad_norm": 0.17149655520915985,
"learning_rate": 0.0007479289940828403,
"loss": 1.7213884592056274,
"step": 1792
},
{
"epoch": 0.37433489827856026,
"grad_norm": 0.1874895542860031,
"learning_rate": 0.000747430706944877,
"loss": 1.7007603645324707,
"step": 1794
},
{
"epoch": 0.37475221700573813,
"grad_norm": 0.20078080892562866,
"learning_rate": 0.0007469324198069138,
"loss": 1.7445753812789917,
"step": 1796
},
{
"epoch": 0.375169535732916,
"grad_norm": 0.18600624799728394,
"learning_rate": 0.0007464341326689505,
"loss": 1.6251606941223145,
"step": 1798
},
{
"epoch": 0.3755868544600939,
"grad_norm": 0.18509207665920258,
"learning_rate": 0.0007459358455309873,
"loss": 1.814162254333496,
"step": 1800
},
{
"epoch": 0.3760041731872718,
"grad_norm": 0.23867130279541016,
"learning_rate": 0.000745437558393024,
"loss": 1.7590008974075317,
"step": 1802
},
{
"epoch": 0.3764214919144497,
"grad_norm": 0.18108072876930237,
"learning_rate": 0.0007449392712550607,
"loss": 1.6219512224197388,
"step": 1804
},
{
"epoch": 0.37683881064162755,
"grad_norm": 0.2047312706708908,
"learning_rate": 0.0007444409841170976,
"loss": 1.6824214458465576,
"step": 1806
},
{
"epoch": 0.3772561293688054,
"grad_norm": 0.1728755086660385,
"learning_rate": 0.0007439426969791343,
"loss": 1.6953681707382202,
"step": 1808
},
{
"epoch": 0.3776734480959833,
"grad_norm": 0.17071813344955444,
"learning_rate": 0.000743444409841171,
"loss": 1.6445077657699585,
"step": 1810
},
{
"epoch": 0.3780907668231612,
"grad_norm": 0.1784115582704544,
"learning_rate": 0.0007429461227032078,
"loss": 1.7983906269073486,
"step": 1812
},
{
"epoch": 0.37850808555033905,
"grad_norm": 0.21386082470417023,
"learning_rate": 0.0007424478355652445,
"loss": 1.6372995376586914,
"step": 1814
},
{
"epoch": 0.378925404277517,
"grad_norm": 0.1896844208240509,
"learning_rate": 0.0007419495484272812,
"loss": 1.734636664390564,
"step": 1816
},
{
"epoch": 0.37934272300469485,
"grad_norm": 0.21693526208400726,
"learning_rate": 0.000741451261289318,
"loss": 1.7244417667388916,
"step": 1818
},
{
"epoch": 0.3797600417318727,
"grad_norm": 0.20639991760253906,
"learning_rate": 0.0007409529741513548,
"loss": 1.6903407573699951,
"step": 1820
},
{
"epoch": 0.3801773604590506,
"grad_norm": 0.17862026393413544,
"learning_rate": 0.0007404546870133916,
"loss": 1.7973225116729736,
"step": 1822
},
{
"epoch": 0.38059467918622847,
"grad_norm": 0.1792299896478653,
"learning_rate": 0.0007399563998754283,
"loss": 1.7597755193710327,
"step": 1824
},
{
"epoch": 0.38101199791340634,
"grad_norm": 0.17475061118602753,
"learning_rate": 0.000739458112737465,
"loss": 1.6212079524993896,
"step": 1826
},
{
"epoch": 0.38142931664058427,
"grad_norm": 0.18791435658931732,
"learning_rate": 0.0007389598255995018,
"loss": 1.6010181903839111,
"step": 1828
},
{
"epoch": 0.38184663536776214,
"grad_norm": 0.16160285472869873,
"learning_rate": 0.0007384615384615385,
"loss": 1.584136962890625,
"step": 1830
},
{
"epoch": 0.38226395409494,
"grad_norm": 0.19821852445602417,
"learning_rate": 0.0007379632513235752,
"loss": 1.7771517038345337,
"step": 1832
},
{
"epoch": 0.3826812728221179,
"grad_norm": 0.17471517622470856,
"learning_rate": 0.000737464964185612,
"loss": 1.6534874439239502,
"step": 1834
},
{
"epoch": 0.38309859154929576,
"grad_norm": 0.20125603675842285,
"learning_rate": 0.0007369666770476487,
"loss": 1.7214417457580566,
"step": 1836
},
{
"epoch": 0.38351591027647364,
"grad_norm": 0.1912640929222107,
"learning_rate": 0.0007364683899096855,
"loss": 1.7104030847549438,
"step": 1838
},
{
"epoch": 0.38393322900365157,
"grad_norm": 0.18999674916267395,
"learning_rate": 0.0007359701027717222,
"loss": 1.513830304145813,
"step": 1840
},
{
"epoch": 0.38435054773082944,
"grad_norm": 0.20635536313056946,
"learning_rate": 0.0007354718156337589,
"loss": 1.537522554397583,
"step": 1842
},
{
"epoch": 0.3847678664580073,
"grad_norm": 0.17322100698947906,
"learning_rate": 0.0007349735284957957,
"loss": 1.663376808166504,
"step": 1844
},
{
"epoch": 0.3851851851851852,
"grad_norm": 0.1880483329296112,
"learning_rate": 0.0007344752413578324,
"loss": 1.7355417013168335,
"step": 1846
},
{
"epoch": 0.38560250391236306,
"grad_norm": 0.1827288568019867,
"learning_rate": 0.0007339769542198693,
"loss": 1.597900390625,
"step": 1848
},
{
"epoch": 0.38601982263954093,
"grad_norm": 0.18197232484817505,
"learning_rate": 0.000733478667081906,
"loss": 1.5560450553894043,
"step": 1850
},
{
"epoch": 0.3864371413667188,
"grad_norm": 0.17846433818340302,
"learning_rate": 0.0007329803799439427,
"loss": 1.702115535736084,
"step": 1852
},
{
"epoch": 0.38685446009389673,
"grad_norm": 0.18424789607524872,
"learning_rate": 0.0007324820928059795,
"loss": 1.6574522256851196,
"step": 1854
},
{
"epoch": 0.3872717788210746,
"grad_norm": 0.16318754851818085,
"learning_rate": 0.0007319838056680162,
"loss": 1.6719655990600586,
"step": 1856
},
{
"epoch": 0.3876890975482525,
"grad_norm": 0.201651930809021,
"learning_rate": 0.0007314855185300529,
"loss": 1.7175278663635254,
"step": 1858
},
{
"epoch": 0.38810641627543035,
"grad_norm": 0.2002033144235611,
"learning_rate": 0.0007309872313920897,
"loss": 1.6978487968444824,
"step": 1860
},
{
"epoch": 0.38852373500260823,
"grad_norm": 0.18376575410366058,
"learning_rate": 0.0007304889442541264,
"loss": 1.8257746696472168,
"step": 1862
},
{
"epoch": 0.3889410537297861,
"grad_norm": 0.18807926774024963,
"learning_rate": 0.0007299906571161633,
"loss": 1.6768311262130737,
"step": 1864
},
{
"epoch": 0.38935837245696403,
"grad_norm": 0.16683189570903778,
"learning_rate": 0.0007294923699782,
"loss": 1.6995952129364014,
"step": 1866
},
{
"epoch": 0.3897756911841419,
"grad_norm": 0.190398707985878,
"learning_rate": 0.0007289940828402367,
"loss": 1.6539530754089355,
"step": 1868
},
{
"epoch": 0.3901930099113198,
"grad_norm": 0.19362810254096985,
"learning_rate": 0.0007284957957022735,
"loss": 1.5737992525100708,
"step": 1870
},
{
"epoch": 0.39061032863849765,
"grad_norm": 0.16772130131721497,
"learning_rate": 0.0007279975085643102,
"loss": 1.6366090774536133,
"step": 1872
},
{
"epoch": 0.3910276473656755,
"grad_norm": 0.1983155608177185,
"learning_rate": 0.0007274992214263469,
"loss": 1.602610468864441,
"step": 1874
},
{
"epoch": 0.3914449660928534,
"grad_norm": 0.2138281911611557,
"learning_rate": 0.0007270009342883837,
"loss": 1.5655598640441895,
"step": 1876
},
{
"epoch": 0.3918622848200313,
"grad_norm": 0.20030027627944946,
"learning_rate": 0.0007265026471504205,
"loss": 1.7486549615859985,
"step": 1878
},
{
"epoch": 0.3922796035472092,
"grad_norm": 0.16378532350063324,
"learning_rate": 0.0007260043600124572,
"loss": 1.6830294132232666,
"step": 1880
},
{
"epoch": 0.3926969222743871,
"grad_norm": 0.19713258743286133,
"learning_rate": 0.000725506072874494,
"loss": 1.7935724258422852,
"step": 1882
},
{
"epoch": 0.39311424100156495,
"grad_norm": 0.20065288245677948,
"learning_rate": 0.0007250077857365307,
"loss": 1.6355443000793457,
"step": 1884
},
{
"epoch": 0.3935315597287428,
"grad_norm": 0.17801856994628906,
"learning_rate": 0.0007245094985985675,
"loss": 1.6839393377304077,
"step": 1886
},
{
"epoch": 0.3939488784559207,
"grad_norm": 0.16672945022583008,
"learning_rate": 0.0007240112114606042,
"loss": 1.7417818307876587,
"step": 1888
},
{
"epoch": 0.39436619718309857,
"grad_norm": 0.17643491923809052,
"learning_rate": 0.0007235129243226409,
"loss": 1.7232372760772705,
"step": 1890
},
{
"epoch": 0.3947835159102765,
"grad_norm": 0.17067725956439972,
"learning_rate": 0.0007230146371846778,
"loss": 1.7092854976654053,
"step": 1892
},
{
"epoch": 0.39520083463745437,
"grad_norm": 0.16026510298252106,
"learning_rate": 0.0007225163500467145,
"loss": 1.6849822998046875,
"step": 1894
},
{
"epoch": 0.39561815336463224,
"grad_norm": 0.18624243140220642,
"learning_rate": 0.0007220180629087512,
"loss": 1.6231648921966553,
"step": 1896
},
{
"epoch": 0.3960354720918101,
"grad_norm": 0.1890266239643097,
"learning_rate": 0.000721519775770788,
"loss": 1.7189944982528687,
"step": 1898
},
{
"epoch": 0.396452790818988,
"grad_norm": 0.17692142724990845,
"learning_rate": 0.0007210214886328247,
"loss": 1.7012109756469727,
"step": 1900
},
{
"epoch": 0.39687010954616586,
"grad_norm": 0.18538671731948853,
"learning_rate": 0.0007205232014948615,
"loss": 1.6796350479125977,
"step": 1902
},
{
"epoch": 0.3972874282733438,
"grad_norm": 0.17001305520534515,
"learning_rate": 0.0007200249143568982,
"loss": 1.680551290512085,
"step": 1904
},
{
"epoch": 0.39770474700052166,
"grad_norm": 0.19359631836414337,
"learning_rate": 0.000719526627218935,
"loss": 1.604306936264038,
"step": 1906
},
{
"epoch": 0.39812206572769954,
"grad_norm": 0.15615129470825195,
"learning_rate": 0.0007190283400809718,
"loss": 1.6380712985992432,
"step": 1908
},
{
"epoch": 0.3985393844548774,
"grad_norm": 0.20027422904968262,
"learning_rate": 0.0007185300529430085,
"loss": 1.6788406372070312,
"step": 1910
},
{
"epoch": 0.3989567031820553,
"grad_norm": 0.22814518213272095,
"learning_rate": 0.0007180317658050452,
"loss": 1.5780839920043945,
"step": 1912
},
{
"epoch": 0.39937402190923316,
"grad_norm": 0.22108320891857147,
"learning_rate": 0.000717533478667082,
"loss": 1.6069958209991455,
"step": 1914
},
{
"epoch": 0.3997913406364111,
"grad_norm": 0.16581694781780243,
"learning_rate": 0.0007170351915291187,
"loss": 1.6984632015228271,
"step": 1916
},
{
"epoch": 0.40020865936358896,
"grad_norm": 0.18057137727737427,
"learning_rate": 0.0007165369043911554,
"loss": 1.7250497341156006,
"step": 1918
},
{
"epoch": 0.40020865936358896,
"eval_loss": 1.6550198793411255,
"eval_runtime": 25.5778,
"eval_samples_per_second": 60.599,
"eval_steps_per_second": 7.585,
"step": 1918
},
{
"epoch": 0.40062597809076683,
"grad_norm": 0.19742168486118317,
"learning_rate": 0.0007160386172531921,
"loss": 1.6215667724609375,
"step": 1920
},
{
"epoch": 0.4010432968179447,
"grad_norm": 0.16541939973831177,
"learning_rate": 0.0007155403301152289,
"loss": 1.5849136114120483,
"step": 1922
},
{
"epoch": 0.4014606155451226,
"grad_norm": 0.1955988109111786,
"learning_rate": 0.0007150420429772657,
"loss": 1.7365734577178955,
"step": 1924
},
{
"epoch": 0.40187793427230045,
"grad_norm": 0.18706025183200836,
"learning_rate": 0.0007145437558393024,
"loss": 1.5803372859954834,
"step": 1926
},
{
"epoch": 0.4022952529994783,
"grad_norm": 0.21752716600894928,
"learning_rate": 0.0007140454687013392,
"loss": 1.5979130268096924,
"step": 1928
},
{
"epoch": 0.40271257172665625,
"grad_norm": 0.18118444085121155,
"learning_rate": 0.0007135471815633759,
"loss": 1.790116786956787,
"step": 1930
},
{
"epoch": 0.4031298904538341,
"grad_norm": 0.15723809599876404,
"learning_rate": 0.0007130488944254126,
"loss": 1.7410356998443604,
"step": 1932
},
{
"epoch": 0.403547209181012,
"grad_norm": 0.1724429726600647,
"learning_rate": 0.0007125506072874494,
"loss": 1.7528969049453735,
"step": 1934
},
{
"epoch": 0.4039645279081899,
"grad_norm": 0.1871008723974228,
"learning_rate": 0.0007120523201494862,
"loss": 1.7563276290893555,
"step": 1936
},
{
"epoch": 0.40438184663536775,
"grad_norm": 0.19913703203201294,
"learning_rate": 0.0007115540330115229,
"loss": 1.585667371749878,
"step": 1938
},
{
"epoch": 0.4047991653625456,
"grad_norm": 0.18781612813472748,
"learning_rate": 0.0007110557458735597,
"loss": 1.7286334037780762,
"step": 1940
},
{
"epoch": 0.40521648408972355,
"grad_norm": 0.20625922083854675,
"learning_rate": 0.0007105574587355964,
"loss": 1.7253109216690063,
"step": 1942
},
{
"epoch": 0.4056338028169014,
"grad_norm": 0.19266554713249207,
"learning_rate": 0.0007100591715976331,
"loss": 1.8188170194625854,
"step": 1944
},
{
"epoch": 0.4060511215440793,
"grad_norm": 0.18233270943164825,
"learning_rate": 0.0007095608844596699,
"loss": 1.7483798265457153,
"step": 1946
},
{
"epoch": 0.40646844027125717,
"grad_norm": 0.16827167570590973,
"learning_rate": 0.0007090625973217066,
"loss": 1.779498815536499,
"step": 1948
},
{
"epoch": 0.40688575899843504,
"grad_norm": 0.2003028690814972,
"learning_rate": 0.0007085643101837435,
"loss": 1.7313119173049927,
"step": 1950
},
{
"epoch": 0.4073030777256129,
"grad_norm": 0.17419478297233582,
"learning_rate": 0.0007080660230457802,
"loss": 1.685150146484375,
"step": 1952
},
{
"epoch": 0.40772039645279085,
"grad_norm": 0.23210090398788452,
"learning_rate": 0.0007075677359078169,
"loss": 1.6654866933822632,
"step": 1954
},
{
"epoch": 0.4081377151799687,
"grad_norm": 0.18305149674415588,
"learning_rate": 0.0007070694487698537,
"loss": 1.6528769731521606,
"step": 1956
},
{
"epoch": 0.4085550339071466,
"grad_norm": 0.2183254361152649,
"learning_rate": 0.0007065711616318904,
"loss": 1.6183055639266968,
"step": 1958
},
{
"epoch": 0.40897235263432447,
"grad_norm": 0.19013535976409912,
"learning_rate": 0.0007060728744939271,
"loss": 1.642290472984314,
"step": 1960
},
{
"epoch": 0.40938967136150234,
"grad_norm": 0.16879412531852722,
"learning_rate": 0.0007055745873559639,
"loss": 1.5484110116958618,
"step": 1962
},
{
"epoch": 0.4098069900886802,
"grad_norm": 0.1673821210861206,
"learning_rate": 0.0007050763002180006,
"loss": 1.612941861152649,
"step": 1964
},
{
"epoch": 0.4102243088158581,
"grad_norm": 0.19346307218074799,
"learning_rate": 0.0007045780130800375,
"loss": 1.601862907409668,
"step": 1966
},
{
"epoch": 0.410641627543036,
"grad_norm": 0.19376583397388458,
"learning_rate": 0.0007040797259420742,
"loss": 1.551114797592163,
"step": 1968
},
{
"epoch": 0.4110589462702139,
"grad_norm": 0.18905408680438995,
"learning_rate": 0.0007035814388041109,
"loss": 1.6617710590362549,
"step": 1970
},
{
"epoch": 0.41147626499739176,
"grad_norm": 0.18663017451763153,
"learning_rate": 0.0007030831516661477,
"loss": 1.6053431034088135,
"step": 1972
},
{
"epoch": 0.41189358372456963,
"grad_norm": 0.1955297738313675,
"learning_rate": 0.0007025848645281844,
"loss": 1.6625816822052002,
"step": 1974
},
{
"epoch": 0.4123109024517475,
"grad_norm": 0.17214395105838776,
"learning_rate": 0.0007020865773902211,
"loss": 1.6572118997573853,
"step": 1976
},
{
"epoch": 0.4127282211789254,
"grad_norm": 0.18428188562393188,
"learning_rate": 0.0007015882902522579,
"loss": 1.7975879907608032,
"step": 1978
},
{
"epoch": 0.4131455399061033,
"grad_norm": 0.19191627204418182,
"learning_rate": 0.0007010900031142947,
"loss": 1.723305344581604,
"step": 1980
},
{
"epoch": 0.4135628586332812,
"grad_norm": 0.20642754435539246,
"learning_rate": 0.0007005917159763315,
"loss": 1.698765516281128,
"step": 1982
},
{
"epoch": 0.41398017736045906,
"grad_norm": 0.2158510535955429,
"learning_rate": 0.0007000934288383682,
"loss": 1.7140588760375977,
"step": 1984
},
{
"epoch": 0.41439749608763693,
"grad_norm": 0.23603455722332,
"learning_rate": 0.0006995951417004049,
"loss": 1.6636567115783691,
"step": 1986
},
{
"epoch": 0.4148148148148148,
"grad_norm": 0.17696522176265717,
"learning_rate": 0.0006990968545624417,
"loss": 1.7558321952819824,
"step": 1988
},
{
"epoch": 0.4152321335419927,
"grad_norm": 0.21585774421691895,
"learning_rate": 0.0006985985674244784,
"loss": 1.707481861114502,
"step": 1990
},
{
"epoch": 0.4156494522691706,
"grad_norm": 0.1892940104007721,
"learning_rate": 0.0006981002802865151,
"loss": 1.56398344039917,
"step": 1992
},
{
"epoch": 0.4160667709963485,
"grad_norm": 0.1834668070077896,
"learning_rate": 0.000697601993148552,
"loss": 1.7686594724655151,
"step": 1994
},
{
"epoch": 0.41648408972352635,
"grad_norm": 0.18568262457847595,
"learning_rate": 0.0006971037060105887,
"loss": 1.5007394552230835,
"step": 1996
},
{
"epoch": 0.4169014084507042,
"grad_norm": 0.18123599886894226,
"learning_rate": 0.0006966054188726254,
"loss": 1.6267486810684204,
"step": 1998
},
{
"epoch": 0.4173187271778821,
"grad_norm": 0.16701748967170715,
"learning_rate": 0.0006961071317346622,
"loss": 1.7028529644012451,
"step": 2000
},
{
"epoch": 0.41773604590505997,
"grad_norm": 0.1927339732646942,
"learning_rate": 0.0006956088445966988,
"loss": 1.598311185836792,
"step": 2002
},
{
"epoch": 0.41815336463223785,
"grad_norm": 0.1722911149263382,
"learning_rate": 0.0006951105574587356,
"loss": 1.531990647315979,
"step": 2004
},
{
"epoch": 0.4185706833594158,
"grad_norm": 0.19591574370861053,
"learning_rate": 0.0006946122703207723,
"loss": 1.6635020971298218,
"step": 2006
},
{
"epoch": 0.41898800208659365,
"grad_norm": 0.22266089916229248,
"learning_rate": 0.000694113983182809,
"loss": 1.6365724802017212,
"step": 2008
},
{
"epoch": 0.4194053208137715,
"grad_norm": 0.19425643980503082,
"learning_rate": 0.0006936156960448459,
"loss": 1.63710618019104,
"step": 2010
},
{
"epoch": 0.4198226395409494,
"grad_norm": 0.18465062975883484,
"learning_rate": 0.0006931174089068826,
"loss": 1.648695468902588,
"step": 2012
},
{
"epoch": 0.42023995826812727,
"grad_norm": 0.19431091845035553,
"learning_rate": 0.0006926191217689194,
"loss": 1.5516412258148193,
"step": 2014
},
{
"epoch": 0.42065727699530514,
"grad_norm": 0.216373011469841,
"learning_rate": 0.0006921208346309561,
"loss": 1.692888855934143,
"step": 2016
},
{
"epoch": 0.42107459572248307,
"grad_norm": 0.17031949758529663,
"learning_rate": 0.0006916225474929928,
"loss": 1.5910251140594482,
"step": 2018
},
{
"epoch": 0.42149191444966094,
"grad_norm": 0.1789386123418808,
"learning_rate": 0.0006911242603550296,
"loss": 1.7299878597259521,
"step": 2020
},
{
"epoch": 0.4219092331768388,
"grad_norm": 0.17683827877044678,
"learning_rate": 0.0006906259732170663,
"loss": 1.726607322692871,
"step": 2022
},
{
"epoch": 0.4223265519040167,
"grad_norm": 0.17248451709747314,
"learning_rate": 0.000690127686079103,
"loss": 1.6691478490829468,
"step": 2024
},
{
"epoch": 0.42274387063119456,
"grad_norm": 0.20521403849124908,
"learning_rate": 0.0006896293989411399,
"loss": 1.6416796445846558,
"step": 2026
},
{
"epoch": 0.42316118935837244,
"grad_norm": 0.2102445363998413,
"learning_rate": 0.0006891311118031766,
"loss": 1.744062066078186,
"step": 2028
},
{
"epoch": 0.42357850808555036,
"grad_norm": 0.2281162440776825,
"learning_rate": 0.0006886328246652134,
"loss": 1.7833259105682373,
"step": 2030
},
{
"epoch": 0.42399582681272824,
"grad_norm": 0.17696285247802734,
"learning_rate": 0.0006881345375272501,
"loss": 1.6629830598831177,
"step": 2032
},
{
"epoch": 0.4244131455399061,
"grad_norm": 0.1717655509710312,
"learning_rate": 0.0006876362503892868,
"loss": 1.5669326782226562,
"step": 2034
},
{
"epoch": 0.424830464267084,
"grad_norm": 0.1734437793493271,
"learning_rate": 0.0006871379632513236,
"loss": 1.6308287382125854,
"step": 2036
},
{
"epoch": 0.42524778299426186,
"grad_norm": 0.18259912729263306,
"learning_rate": 0.0006866396761133603,
"loss": 1.7051451206207275,
"step": 2038
},
{
"epoch": 0.42566510172143973,
"grad_norm": 0.19569163024425507,
"learning_rate": 0.0006861413889753971,
"loss": 1.7503201961517334,
"step": 2040
},
{
"epoch": 0.4260824204486176,
"grad_norm": 0.20505301654338837,
"learning_rate": 0.0006856431018374339,
"loss": 1.739957332611084,
"step": 2042
},
{
"epoch": 0.42649973917579553,
"grad_norm": 0.16601890325546265,
"learning_rate": 0.0006851448146994706,
"loss": 1.6939730644226074,
"step": 2044
},
{
"epoch": 0.4269170579029734,
"grad_norm": 0.14138959348201752,
"learning_rate": 0.0006846465275615074,
"loss": 1.6229454278945923,
"step": 2046
},
{
"epoch": 0.4273343766301513,
"grad_norm": 0.16821424663066864,
"learning_rate": 0.0006841482404235441,
"loss": 1.569868564605713,
"step": 2048
},
{
"epoch": 0.42775169535732915,
"grad_norm": 0.18612134456634521,
"learning_rate": 0.0006836499532855808,
"loss": 1.7251687049865723,
"step": 2050
},
{
"epoch": 0.428169014084507,
"grad_norm": 0.21419104933738708,
"learning_rate": 0.0006831516661476177,
"loss": 1.5867853164672852,
"step": 2052
},
{
"epoch": 0.4285863328116849,
"grad_norm": 0.17739352583885193,
"learning_rate": 0.0006826533790096544,
"loss": 1.6989272832870483,
"step": 2054
},
{
"epoch": 0.42900365153886283,
"grad_norm": 0.18007194995880127,
"learning_rate": 0.0006821550918716911,
"loss": 1.6633107662200928,
"step": 2056
},
{
"epoch": 0.4294209702660407,
"grad_norm": 0.18534772098064423,
"learning_rate": 0.0006816568047337279,
"loss": 1.524631381034851,
"step": 2058
},
{
"epoch": 0.4298382889932186,
"grad_norm": 0.23682451248168945,
"learning_rate": 0.0006811585175957646,
"loss": 1.5820634365081787,
"step": 2060
},
{
"epoch": 0.43025560772039645,
"grad_norm": 0.17606110870838165,
"learning_rate": 0.0006806602304578013,
"loss": 1.6821006536483765,
"step": 2062
},
{
"epoch": 0.4306729264475743,
"grad_norm": 0.17417310178279877,
"learning_rate": 0.0006801619433198381,
"loss": 1.5983318090438843,
"step": 2064
},
{
"epoch": 0.4310902451747522,
"grad_norm": 0.21326765418052673,
"learning_rate": 0.0006796636561818748,
"loss": 1.6636732816696167,
"step": 2066
},
{
"epoch": 0.4315075639019301,
"grad_norm": 0.1700189858675003,
"learning_rate": 0.0006791653690439117,
"loss": 1.6755625009536743,
"step": 2068
},
{
"epoch": 0.431924882629108,
"grad_norm": 0.17494486272335052,
"learning_rate": 0.0006786670819059484,
"loss": 1.5736558437347412,
"step": 2070
},
{
"epoch": 0.43234220135628587,
"grad_norm": 0.17146234214305878,
"learning_rate": 0.0006781687947679851,
"loss": 1.6380850076675415,
"step": 2072
},
{
"epoch": 0.43275952008346374,
"grad_norm": 0.19481350481510162,
"learning_rate": 0.0006776705076300219,
"loss": 1.5706202983856201,
"step": 2074
},
{
"epoch": 0.4331768388106416,
"grad_norm": 0.18794581294059753,
"learning_rate": 0.0006771722204920586,
"loss": 1.6174395084381104,
"step": 2076
},
{
"epoch": 0.4335941575378195,
"grad_norm": 0.20505531132221222,
"learning_rate": 0.0006766739333540953,
"loss": 1.7102923393249512,
"step": 2078
},
{
"epoch": 0.43401147626499736,
"grad_norm": 0.21512311697006226,
"learning_rate": 0.0006761756462161321,
"loss": 1.6858521699905396,
"step": 2080
},
{
"epoch": 0.4344287949921753,
"grad_norm": 0.1825442761182785,
"learning_rate": 0.0006756773590781689,
"loss": 1.5891410112380981,
"step": 2082
},
{
"epoch": 0.43484611371935317,
"grad_norm": 0.18736852705478668,
"learning_rate": 0.0006751790719402056,
"loss": 1.6347702741622925,
"step": 2084
},
{
"epoch": 0.43526343244653104,
"grad_norm": 0.17808939516544342,
"learning_rate": 0.0006746807848022423,
"loss": 1.8280189037322998,
"step": 2086
},
{
"epoch": 0.4356807511737089,
"grad_norm": 0.21274828910827637,
"learning_rate": 0.000674182497664279,
"loss": 1.5847911834716797,
"step": 2088
},
{
"epoch": 0.4360980699008868,
"grad_norm": 0.18459489941596985,
"learning_rate": 0.0006736842105263158,
"loss": 1.7398196458816528,
"step": 2090
},
{
"epoch": 0.43651538862806466,
"grad_norm": 0.17764605581760406,
"learning_rate": 0.0006731859233883525,
"loss": 1.6091114282608032,
"step": 2092
},
{
"epoch": 0.4369327073552426,
"grad_norm": 0.19259013235569,
"learning_rate": 0.0006726876362503893,
"loss": 1.6295032501220703,
"step": 2094
},
{
"epoch": 0.43735002608242046,
"grad_norm": 0.19112874567508698,
"learning_rate": 0.000672189349112426,
"loss": 1.6889574527740479,
"step": 2096
},
{
"epoch": 0.43776734480959834,
"grad_norm": 0.19554616510868073,
"learning_rate": 0.0006716910619744628,
"loss": 1.551025390625,
"step": 2098
},
{
"epoch": 0.4381846635367762,
"grad_norm": 0.17289364337921143,
"learning_rate": 0.0006711927748364996,
"loss": 1.6089515686035156,
"step": 2100
},
{
"epoch": 0.4386019822639541,
"grad_norm": 0.18652550876140594,
"learning_rate": 0.0006706944876985363,
"loss": 1.6269158124923706,
"step": 2102
},
{
"epoch": 0.43901930099113196,
"grad_norm": 0.18318471312522888,
"learning_rate": 0.000670196200560573,
"loss": 1.8091142177581787,
"step": 2104
},
{
"epoch": 0.4394366197183099,
"grad_norm": 0.17808561027050018,
"learning_rate": 0.0006696979134226098,
"loss": 1.6483501195907593,
"step": 2106
},
{
"epoch": 0.43985393844548776,
"grad_norm": 0.17564748227596283,
"learning_rate": 0.0006691996262846465,
"loss": 1.6110496520996094,
"step": 2108
},
{
"epoch": 0.44027125717266563,
"grad_norm": 0.1854456514120102,
"learning_rate": 0.0006687013391466833,
"loss": 1.6509326696395874,
"step": 2110
},
{
"epoch": 0.4406885758998435,
"grad_norm": 0.19521023333072662,
"learning_rate": 0.0006682030520087201,
"loss": 1.4685437679290771,
"step": 2112
},
{
"epoch": 0.4411058946270214,
"grad_norm": 0.19183216989040375,
"learning_rate": 0.0006677047648707568,
"loss": 1.615735411643982,
"step": 2114
},
{
"epoch": 0.44152321335419925,
"grad_norm": 0.17928355932235718,
"learning_rate": 0.0006672064777327936,
"loss": 1.6966636180877686,
"step": 2116
},
{
"epoch": 0.4419405320813771,
"grad_norm": 0.19644534587860107,
"learning_rate": 0.0006667081905948303,
"loss": 1.6603734493255615,
"step": 2118
},
{
"epoch": 0.44235785080855505,
"grad_norm": 0.1943538933992386,
"learning_rate": 0.000666209903456867,
"loss": 1.65966796875,
"step": 2120
},
{
"epoch": 0.4427751695357329,
"grad_norm": 0.1803293526172638,
"learning_rate": 0.0006657116163189038,
"loss": 1.750153660774231,
"step": 2122
},
{
"epoch": 0.4431924882629108,
"grad_norm": 0.19837968051433563,
"learning_rate": 0.0006652133291809405,
"loss": 1.7525608539581299,
"step": 2124
},
{
"epoch": 0.4436098069900887,
"grad_norm": 0.2169271558523178,
"learning_rate": 0.0006647150420429772,
"loss": 1.7094480991363525,
"step": 2126
},
{
"epoch": 0.44402712571726655,
"grad_norm": 0.1747976690530777,
"learning_rate": 0.0006642167549050141,
"loss": 1.710147738456726,
"step": 2128
},
{
"epoch": 0.4444444444444444,
"grad_norm": 0.17337706685066223,
"learning_rate": 0.0006637184677670508,
"loss": 1.7052322626113892,
"step": 2130
},
{
"epoch": 0.44486176317162235,
"grad_norm": 0.19824403524398804,
"learning_rate": 0.0006632201806290876,
"loss": 1.5822726488113403,
"step": 2132
},
{
"epoch": 0.4452790818988002,
"grad_norm": 0.19013693928718567,
"learning_rate": 0.0006627218934911243,
"loss": 1.691605806350708,
"step": 2134
},
{
"epoch": 0.4456964006259781,
"grad_norm": 0.19300545752048492,
"learning_rate": 0.000662223606353161,
"loss": 1.6828343868255615,
"step": 2136
},
{
"epoch": 0.44611371935315597,
"grad_norm": 0.1866510808467865,
"learning_rate": 0.0006617253192151978,
"loss": 1.5603797435760498,
"step": 2138
},
{
"epoch": 0.44653103808033384,
"grad_norm": 0.18131285905838013,
"learning_rate": 0.0006612270320772345,
"loss": 1.762597680091858,
"step": 2140
},
{
"epoch": 0.4469483568075117,
"grad_norm": 0.22835023701190948,
"learning_rate": 0.0006607287449392713,
"loss": 1.5227411985397339,
"step": 2142
},
{
"epoch": 0.44736567553468964,
"grad_norm": 0.1832299381494522,
"learning_rate": 0.0006602304578013081,
"loss": 1.6503472328186035,
"step": 2144
},
{
"epoch": 0.4477829942618675,
"grad_norm": 0.1619878113269806,
"learning_rate": 0.0006597321706633448,
"loss": 1.530448317527771,
"step": 2146
},
{
"epoch": 0.4482003129890454,
"grad_norm": 0.1806347370147705,
"learning_rate": 0.0006592338835253816,
"loss": 1.720436453819275,
"step": 2148
},
{
"epoch": 0.44861763171622326,
"grad_norm": 0.17783842980861664,
"learning_rate": 0.0006587355963874183,
"loss": 1.6536171436309814,
"step": 2150
},
{
"epoch": 0.44903495044340114,
"grad_norm": 0.19388899207115173,
"learning_rate": 0.000658237309249455,
"loss": 1.5561476945877075,
"step": 2152
},
{
"epoch": 0.449452269170579,
"grad_norm": 0.170505091547966,
"learning_rate": 0.0006577390221114918,
"loss": 1.5716391801834106,
"step": 2154
},
{
"epoch": 0.44986958789775694,
"grad_norm": 0.2037661373615265,
"learning_rate": 0.0006572407349735286,
"loss": 1.655495524406433,
"step": 2156
},
{
"epoch": 0.4502869066249348,
"grad_norm": 0.16902558505535126,
"learning_rate": 0.0006567424478355653,
"loss": 1.6777594089508057,
"step": 2158
},
{
"epoch": 0.4507042253521127,
"grad_norm": 0.20422546565532684,
"learning_rate": 0.0006562441606976021,
"loss": 1.6523692607879639,
"step": 2160
},
{
"epoch": 0.45112154407929056,
"grad_norm": 0.18302936851978302,
"learning_rate": 0.0006557458735596388,
"loss": 1.552883267402649,
"step": 2162
},
{
"epoch": 0.45153886280646843,
"grad_norm": 0.19708263874053955,
"learning_rate": 0.0006552475864216755,
"loss": 1.5769011974334717,
"step": 2164
},
{
"epoch": 0.4519561815336463,
"grad_norm": 0.22629764676094055,
"learning_rate": 0.0006547492992837123,
"loss": 1.4636987447738647,
"step": 2166
},
{
"epoch": 0.4523735002608242,
"grad_norm": 0.1784856617450714,
"learning_rate": 0.0006542510121457489,
"loss": 1.8124885559082031,
"step": 2168
},
{
"epoch": 0.4527908189880021,
"grad_norm": 0.16386978328227997,
"learning_rate": 0.0006537527250077858,
"loss": 1.7226455211639404,
"step": 2170
},
{
"epoch": 0.45320813771518,
"grad_norm": 0.2245175838470459,
"learning_rate": 0.0006532544378698225,
"loss": 1.6121279001235962,
"step": 2172
},
{
"epoch": 0.45362545644235786,
"grad_norm": 0.2042732536792755,
"learning_rate": 0.0006527561507318593,
"loss": 1.8244189023971558,
"step": 2174
},
{
"epoch": 0.45404277516953573,
"grad_norm": 0.237874835729599,
"learning_rate": 0.000652257863593896,
"loss": 1.4849438667297363,
"step": 2176
},
{
"epoch": 0.4544600938967136,
"grad_norm": 0.18323439359664917,
"learning_rate": 0.0006517595764559327,
"loss": 1.7365816831588745,
"step": 2178
},
{
"epoch": 0.4548774126238915,
"grad_norm": 0.2310006320476532,
"learning_rate": 0.0006512612893179695,
"loss": 1.719451665878296,
"step": 2180
},
{
"epoch": 0.4552947313510694,
"grad_norm": 0.2078419029712677,
"learning_rate": 0.0006507630021800062,
"loss": 1.5895004272460938,
"step": 2182
},
{
"epoch": 0.4557120500782473,
"grad_norm": 0.20061245560646057,
"learning_rate": 0.0006502647150420429,
"loss": 1.8829911947250366,
"step": 2184
},
{
"epoch": 0.45612936880542515,
"grad_norm": 0.20012111961841583,
"learning_rate": 0.0006497664279040798,
"loss": 1.4763175249099731,
"step": 2186
},
{
"epoch": 0.456546687532603,
"grad_norm": 0.17549832165241241,
"learning_rate": 0.0006492681407661165,
"loss": 1.6976118087768555,
"step": 2188
},
{
"epoch": 0.4569640062597809,
"grad_norm": 0.1684069186449051,
"learning_rate": 0.0006487698536281532,
"loss": 1.682408094406128,
"step": 2190
},
{
"epoch": 0.45738132498695877,
"grad_norm": 0.19265960156917572,
"learning_rate": 0.00064827156649019,
"loss": 1.6903444528579712,
"step": 2192
},
{
"epoch": 0.4577986437141367,
"grad_norm": 0.20462718605995178,
"learning_rate": 0.0006477732793522267,
"loss": 1.5667544603347778,
"step": 2194
},
{
"epoch": 0.4582159624413146,
"grad_norm": 0.23280535638332367,
"learning_rate": 0.0006472749922142635,
"loss": 1.7492413520812988,
"step": 2196
},
{
"epoch": 0.45863328116849245,
"grad_norm": 0.17211668193340302,
"learning_rate": 0.0006467767050763002,
"loss": 1.71237051486969,
"step": 2198
},
{
"epoch": 0.4590505998956703,
"grad_norm": 0.2092355340719223,
"learning_rate": 0.000646278417938337,
"loss": 1.6101242303848267,
"step": 2200
},
{
"epoch": 0.4594679186228482,
"grad_norm": 0.2288168966770172,
"learning_rate": 0.0006457801308003738,
"loss": 1.525240182876587,
"step": 2202
},
{
"epoch": 0.45988523735002607,
"grad_norm": 0.19022385776042938,
"learning_rate": 0.0006452818436624105,
"loss": 1.6549973487854004,
"step": 2204
},
{
"epoch": 0.46030255607720394,
"grad_norm": 0.1893918663263321,
"learning_rate": 0.0006447835565244472,
"loss": 1.603515386581421,
"step": 2206
},
{
"epoch": 0.46071987480438187,
"grad_norm": 0.19583308696746826,
"learning_rate": 0.000644285269386484,
"loss": 1.792559266090393,
"step": 2208
},
{
"epoch": 0.46113719353155974,
"grad_norm": 0.17401592433452606,
"learning_rate": 0.0006437869822485207,
"loss": 1.781632661819458,
"step": 2210
},
{
"epoch": 0.4615545122587376,
"grad_norm": 0.164097860455513,
"learning_rate": 0.0006432886951105575,
"loss": 1.6397689580917358,
"step": 2212
},
{
"epoch": 0.4619718309859155,
"grad_norm": 0.18655383586883545,
"learning_rate": 0.0006427904079725943,
"loss": 1.65058171749115,
"step": 2214
},
{
"epoch": 0.46238914971309336,
"grad_norm": 0.19977721571922302,
"learning_rate": 0.000642292120834631,
"loss": 1.6740310192108154,
"step": 2216
},
{
"epoch": 0.46280646844027123,
"grad_norm": 0.1789490431547165,
"learning_rate": 0.0006417938336966678,
"loss": 1.660505771636963,
"step": 2218
},
{
"epoch": 0.46322378716744916,
"grad_norm": 0.15630024671554565,
"learning_rate": 0.0006412955465587045,
"loss": 1.7098737955093384,
"step": 2220
},
{
"epoch": 0.46364110589462704,
"grad_norm": 0.17152395844459534,
"learning_rate": 0.0006407972594207412,
"loss": 1.7534066438674927,
"step": 2222
},
{
"epoch": 0.4640584246218049,
"grad_norm": 0.21707962453365326,
"learning_rate": 0.000640298972282778,
"loss": 1.6748636960983276,
"step": 2224
},
{
"epoch": 0.4644757433489828,
"grad_norm": 0.1711374819278717,
"learning_rate": 0.0006398006851448147,
"loss": 1.6252702474594116,
"step": 2226
},
{
"epoch": 0.46489306207616066,
"grad_norm": 0.2005840241909027,
"learning_rate": 0.0006393023980068514,
"loss": 1.7125120162963867,
"step": 2228
},
{
"epoch": 0.46531038080333853,
"grad_norm": 0.18972571194171906,
"learning_rate": 0.0006388041108688883,
"loss": 1.7711200714111328,
"step": 2230
},
{
"epoch": 0.46572769953051646,
"grad_norm": 0.18839924037456512,
"learning_rate": 0.000638305823730925,
"loss": 1.6517610549926758,
"step": 2232
},
{
"epoch": 0.46614501825769433,
"grad_norm": 0.17488519847393036,
"learning_rate": 0.0006378075365929618,
"loss": 1.86976957321167,
"step": 2234
},
{
"epoch": 0.4665623369848722,
"grad_norm": 0.2039034068584442,
"learning_rate": 0.0006373092494549985,
"loss": 1.6738086938858032,
"step": 2236
},
{
"epoch": 0.4669796557120501,
"grad_norm": 0.20377559959888458,
"learning_rate": 0.0006368109623170352,
"loss": 1.6221966743469238,
"step": 2238
},
{
"epoch": 0.46739697443922795,
"grad_norm": 0.21467648446559906,
"learning_rate": 0.000636312675179072,
"loss": 1.606877326965332,
"step": 2240
},
{
"epoch": 0.4678142931664058,
"grad_norm": 0.17877544462680817,
"learning_rate": 0.0006358143880411087,
"loss": 1.6395072937011719,
"step": 2242
},
{
"epoch": 0.4682316118935837,
"grad_norm": 0.1867547184228897,
"learning_rate": 0.0006353161009031455,
"loss": 1.6619255542755127,
"step": 2244
},
{
"epoch": 0.46864893062076163,
"grad_norm": 0.16417239606380463,
"learning_rate": 0.0006348178137651823,
"loss": 1.8050012588500977,
"step": 2246
},
{
"epoch": 0.4690662493479395,
"grad_norm": 0.19289712607860565,
"learning_rate": 0.000634319526627219,
"loss": 1.7720729112625122,
"step": 2248
},
{
"epoch": 0.4694835680751174,
"grad_norm": 0.19002565741539001,
"learning_rate": 0.0006338212394892558,
"loss": 1.6322200298309326,
"step": 2250
},
{
"epoch": 0.46990088680229525,
"grad_norm": 0.1756228655576706,
"learning_rate": 0.0006333229523512924,
"loss": 1.6642818450927734,
"step": 2252
},
{
"epoch": 0.4703182055294731,
"grad_norm": 0.19345125555992126,
"learning_rate": 0.0006328246652133291,
"loss": 1.7236707210540771,
"step": 2254
},
{
"epoch": 0.470735524256651,
"grad_norm": 0.18979834020137787,
"learning_rate": 0.0006323263780753659,
"loss": 1.5878114700317383,
"step": 2256
},
{
"epoch": 0.4711528429838289,
"grad_norm": 0.19021959602832794,
"learning_rate": 0.0006318280909374027,
"loss": 1.6624784469604492,
"step": 2258
},
{
"epoch": 0.4715701617110068,
"grad_norm": 0.1626739799976349,
"learning_rate": 0.0006313298037994395,
"loss": 1.696656346321106,
"step": 2260
},
{
"epoch": 0.47198748043818467,
"grad_norm": 0.17045506834983826,
"learning_rate": 0.0006308315166614762,
"loss": 1.6809720993041992,
"step": 2262
},
{
"epoch": 0.47240479916536254,
"grad_norm": 0.17728687822818756,
"learning_rate": 0.0006303332295235129,
"loss": 1.4981613159179688,
"step": 2264
},
{
"epoch": 0.4728221178925404,
"grad_norm": 0.18320240080356598,
"learning_rate": 0.0006298349423855497,
"loss": 1.7004302740097046,
"step": 2266
},
{
"epoch": 0.4732394366197183,
"grad_norm": 0.1768762320280075,
"learning_rate": 0.0006293366552475864,
"loss": 1.6750084161758423,
"step": 2268
},
{
"epoch": 0.4736567553468962,
"grad_norm": 0.18989863991737366,
"learning_rate": 0.0006288383681096231,
"loss": 1.5724411010742188,
"step": 2270
},
{
"epoch": 0.4740740740740741,
"grad_norm": 0.18472252786159515,
"learning_rate": 0.00062834008097166,
"loss": 1.6495447158813477,
"step": 2272
},
{
"epoch": 0.47449139280125197,
"grad_norm": 0.16814455389976501,
"learning_rate": 0.0006278417938336967,
"loss": 1.7860522270202637,
"step": 2274
},
{
"epoch": 0.47490871152842984,
"grad_norm": 0.17506344616413116,
"learning_rate": 0.0006273435066957335,
"loss": 1.7075011730194092,
"step": 2276
},
{
"epoch": 0.4753260302556077,
"grad_norm": 0.18617728352546692,
"learning_rate": 0.0006268452195577702,
"loss": 1.8390833139419556,
"step": 2278
},
{
"epoch": 0.4757433489827856,
"grad_norm": 0.1951189637184143,
"learning_rate": 0.0006263469324198069,
"loss": 1.7569984197616577,
"step": 2280
},
{
"epoch": 0.47616066770996346,
"grad_norm": 0.19304701685905457,
"learning_rate": 0.0006258486452818437,
"loss": 1.666843295097351,
"step": 2282
},
{
"epoch": 0.4765779864371414,
"grad_norm": 0.17116114497184753,
"learning_rate": 0.0006253503581438804,
"loss": 1.7839329242706299,
"step": 2284
},
{
"epoch": 0.47699530516431926,
"grad_norm": 0.16400864720344543,
"learning_rate": 0.0006248520710059171,
"loss": 1.6923694610595703,
"step": 2286
},
{
"epoch": 0.47741262389149713,
"grad_norm": 0.1887049525976181,
"learning_rate": 0.000624353783867954,
"loss": 1.7406247854232788,
"step": 2288
},
{
"epoch": 0.477829942618675,
"grad_norm": 0.19211868941783905,
"learning_rate": 0.0006238554967299907,
"loss": 1.6580698490142822,
"step": 2290
},
{
"epoch": 0.4782472613458529,
"grad_norm": 0.1673884242773056,
"learning_rate": 0.0006233572095920274,
"loss": 1.6956758499145508,
"step": 2292
},
{
"epoch": 0.47866458007303075,
"grad_norm": 0.18888750672340393,
"learning_rate": 0.0006228589224540642,
"loss": 1.77895987033844,
"step": 2294
},
{
"epoch": 0.4790818988002087,
"grad_norm": 0.19842085242271423,
"learning_rate": 0.0006223606353161009,
"loss": 1.6926261186599731,
"step": 2296
},
{
"epoch": 0.47949921752738656,
"grad_norm": 0.17038029432296753,
"learning_rate": 0.0006218623481781377,
"loss": 1.7498395442962646,
"step": 2298
},
{
"epoch": 0.47991653625456443,
"grad_norm": 0.18308082222938538,
"learning_rate": 0.0006213640610401744,
"loss": 1.7081809043884277,
"step": 2300
},
{
"epoch": 0.4803338549817423,
"grad_norm": 0.18448707461357117,
"learning_rate": 0.0006208657739022112,
"loss": 1.6018449068069458,
"step": 2302
},
{
"epoch": 0.4807511737089202,
"grad_norm": 0.2225142866373062,
"learning_rate": 0.000620367486764248,
"loss": 1.7542376518249512,
"step": 2304
},
{
"epoch": 0.48116849243609805,
"grad_norm": 0.1678779125213623,
"learning_rate": 0.0006198691996262847,
"loss": 1.6199181079864502,
"step": 2306
},
{
"epoch": 0.481585811163276,
"grad_norm": 0.17093871533870697,
"learning_rate": 0.0006193709124883214,
"loss": 1.7297284603118896,
"step": 2308
},
{
"epoch": 0.48200312989045385,
"grad_norm": 0.17756441235542297,
"learning_rate": 0.0006188726253503582,
"loss": 1.671688437461853,
"step": 2310
},
{
"epoch": 0.4824204486176317,
"grad_norm": 0.19993822276592255,
"learning_rate": 0.0006183743382123949,
"loss": 1.6771029233932495,
"step": 2312
},
{
"epoch": 0.4828377673448096,
"grad_norm": 0.17967967689037323,
"learning_rate": 0.0006178760510744317,
"loss": 1.620607852935791,
"step": 2314
},
{
"epoch": 0.48325508607198747,
"grad_norm": 0.20832481980323792,
"learning_rate": 0.0006173777639364685,
"loss": 1.6548066139221191,
"step": 2316
},
{
"epoch": 0.48367240479916535,
"grad_norm": 0.17591466009616852,
"learning_rate": 0.0006168794767985052,
"loss": 1.6800392866134644,
"step": 2318
},
{
"epoch": 0.4840897235263432,
"grad_norm": 0.20305797457695007,
"learning_rate": 0.000616381189660542,
"loss": 1.7078672647476196,
"step": 2320
},
{
"epoch": 0.48450704225352115,
"grad_norm": 0.20098401606082916,
"learning_rate": 0.0006158829025225787,
"loss": 1.7006711959838867,
"step": 2322
},
{
"epoch": 0.484924360980699,
"grad_norm": 0.1678527593612671,
"learning_rate": 0.0006153846153846154,
"loss": 1.7028697729110718,
"step": 2324
},
{
"epoch": 0.4853416797078769,
"grad_norm": 0.2127094864845276,
"learning_rate": 0.0006148863282466522,
"loss": 1.6439032554626465,
"step": 2326
},
{
"epoch": 0.48575899843505477,
"grad_norm": 0.2123558670282364,
"learning_rate": 0.0006143880411086889,
"loss": 1.7013537883758545,
"step": 2328
},
{
"epoch": 0.48617631716223264,
"grad_norm": 0.17157968878746033,
"learning_rate": 0.0006138897539707258,
"loss": 1.7817928791046143,
"step": 2330
},
{
"epoch": 0.4865936358894105,
"grad_norm": 0.18738074600696564,
"learning_rate": 0.0006133914668327625,
"loss": 1.7224258184432983,
"step": 2332
},
{
"epoch": 0.48701095461658844,
"grad_norm": 0.18726903200149536,
"learning_rate": 0.0006128931796947991,
"loss": 1.6396877765655518,
"step": 2334
},
{
"epoch": 0.4874282733437663,
"grad_norm": 0.19470545649528503,
"learning_rate": 0.0006123948925568359,
"loss": 1.613906979560852,
"step": 2336
},
{
"epoch": 0.4878455920709442,
"grad_norm": 0.16727350652217865,
"learning_rate": 0.0006118966054188726,
"loss": 1.652796983718872,
"step": 2338
},
{
"epoch": 0.48826291079812206,
"grad_norm": 0.1922510415315628,
"learning_rate": 0.0006113983182809094,
"loss": 1.7772053480148315,
"step": 2340
},
{
"epoch": 0.48868022952529994,
"grad_norm": 0.1855093240737915,
"learning_rate": 0.0006109000311429461,
"loss": 1.7090497016906738,
"step": 2342
},
{
"epoch": 0.4890975482524778,
"grad_norm": 0.21971635520458221,
"learning_rate": 0.0006104017440049828,
"loss": 1.515670895576477,
"step": 2344
},
{
"epoch": 0.48951486697965574,
"grad_norm": 0.17876079678535461,
"learning_rate": 0.0006099034568670197,
"loss": 1.6303476095199585,
"step": 2346
},
{
"epoch": 0.4899321857068336,
"grad_norm": 0.17011287808418274,
"learning_rate": 0.0006094051697290564,
"loss": 1.5526978969573975,
"step": 2348
},
{
"epoch": 0.4903495044340115,
"grad_norm": 0.17210261523723602,
"learning_rate": 0.0006089068825910931,
"loss": 1.6197071075439453,
"step": 2350
},
{
"epoch": 0.49076682316118936,
"grad_norm": 0.16901074349880219,
"learning_rate": 0.0006084085954531299,
"loss": 1.7013108730316162,
"step": 2352
},
{
"epoch": 0.49118414188836723,
"grad_norm": 0.1713145226240158,
"learning_rate": 0.0006079103083151666,
"loss": 1.6729518175125122,
"step": 2354
},
{
"epoch": 0.4916014606155451,
"grad_norm": 0.19876129925251007,
"learning_rate": 0.0006074120211772033,
"loss": 1.5788373947143555,
"step": 2356
},
{
"epoch": 0.492018779342723,
"grad_norm": 0.20095910131931305,
"learning_rate": 0.0006069137340392401,
"loss": 1.687571406364441,
"step": 2358
},
{
"epoch": 0.4924360980699009,
"grad_norm": 0.17344386875629425,
"learning_rate": 0.0006064154469012768,
"loss": 1.6565451622009277,
"step": 2360
},
{
"epoch": 0.4928534167970788,
"grad_norm": 0.190763920545578,
"learning_rate": 0.0006059171597633137,
"loss": 1.6927180290222168,
"step": 2362
},
{
"epoch": 0.49327073552425665,
"grad_norm": 0.19455622136592865,
"learning_rate": 0.0006054188726253504,
"loss": 1.6182314157485962,
"step": 2364
},
{
"epoch": 0.4936880542514345,
"grad_norm": 0.19459623098373413,
"learning_rate": 0.0006049205854873871,
"loss": 1.6601393222808838,
"step": 2366
},
{
"epoch": 0.4941053729786124,
"grad_norm": 0.18286044895648956,
"learning_rate": 0.0006044222983494239,
"loss": 1.6309341192245483,
"step": 2368
},
{
"epoch": 0.4945226917057903,
"grad_norm": 0.1740557700395584,
"learning_rate": 0.0006039240112114606,
"loss": 1.7038424015045166,
"step": 2370
},
{
"epoch": 0.4949400104329682,
"grad_norm": 0.17317251861095428,
"learning_rate": 0.0006034257240734973,
"loss": 1.624521017074585,
"step": 2372
},
{
"epoch": 0.4953573291601461,
"grad_norm": 0.17656317353248596,
"learning_rate": 0.0006029274369355341,
"loss": 1.783949375152588,
"step": 2374
},
{
"epoch": 0.49577464788732395,
"grad_norm": 0.2020646184682846,
"learning_rate": 0.0006024291497975709,
"loss": 1.679700493812561,
"step": 2376
},
{
"epoch": 0.4961919666145018,
"grad_norm": 0.1891888529062271,
"learning_rate": 0.0006019308626596077,
"loss": 1.6682631969451904,
"step": 2378
},
{
"epoch": 0.4966092853416797,
"grad_norm": 0.20177747309207916,
"learning_rate": 0.0006014325755216444,
"loss": 1.5069735050201416,
"step": 2380
},
{
"epoch": 0.49702660406885757,
"grad_norm": 0.18581867218017578,
"learning_rate": 0.0006009342883836811,
"loss": 1.7082576751708984,
"step": 2382
},
{
"epoch": 0.4974439227960355,
"grad_norm": 0.2088656723499298,
"learning_rate": 0.0006004360012457179,
"loss": 1.5859789848327637,
"step": 2384
},
{
"epoch": 0.49786124152321337,
"grad_norm": 0.1886608749628067,
"learning_rate": 0.0005999377141077546,
"loss": 1.7043393850326538,
"step": 2386
},
{
"epoch": 0.49827856025039124,
"grad_norm": 0.1901412457227707,
"learning_rate": 0.0005994394269697913,
"loss": 1.551528811454773,
"step": 2388
},
{
"epoch": 0.4986958789775691,
"grad_norm": 0.17218685150146484,
"learning_rate": 0.0005989411398318282,
"loss": 1.5922272205352783,
"step": 2390
},
{
"epoch": 0.499113197704747,
"grad_norm": 0.19282056391239166,
"learning_rate": 0.0005984428526938649,
"loss": 1.619492530822754,
"step": 2392
},
{
"epoch": 0.49953051643192486,
"grad_norm": 0.18915361166000366,
"learning_rate": 0.0005979445655559017,
"loss": 1.7186191082000732,
"step": 2394
},
{
"epoch": 0.49994783515910274,
"grad_norm": 0.17565329372882843,
"learning_rate": 0.0005974462784179384,
"loss": 1.6823314428329468,
"step": 2396
},
{
"epoch": 0.5003651538862807,
"grad_norm": 0.19733518362045288,
"learning_rate": 0.0005969479912799751,
"loss": 1.6992161273956299,
"step": 2398
},
{
"epoch": 0.5007824726134585,
"grad_norm": 0.1652766913175583,
"learning_rate": 0.0005964497041420119,
"loss": 1.6654176712036133,
"step": 2400
},
{
"epoch": 0.5011997913406364,
"grad_norm": 0.1735621690750122,
"learning_rate": 0.0005959514170040486,
"loss": 1.6445595026016235,
"step": 2402
},
{
"epoch": 0.5016171100678143,
"grad_norm": 0.17494626343250275,
"learning_rate": 0.0005954531298660854,
"loss": 1.6333580017089844,
"step": 2404
},
{
"epoch": 0.5020344287949922,
"grad_norm": 0.16876085102558136,
"learning_rate": 0.0005949548427281222,
"loss": 1.6972699165344238,
"step": 2406
},
{
"epoch": 0.50245174752217,
"grad_norm": 0.18788889050483704,
"learning_rate": 0.0005944565555901589,
"loss": 1.5049173831939697,
"step": 2408
},
{
"epoch": 0.5028690662493479,
"grad_norm": 0.15353694558143616,
"learning_rate": 0.0005939582684521956,
"loss": 1.7257428169250488,
"step": 2410
},
{
"epoch": 0.5032863849765258,
"grad_norm": 0.1653592586517334,
"learning_rate": 0.0005934599813142324,
"loss": 1.6432178020477295,
"step": 2412
},
{
"epoch": 0.5037037037037037,
"grad_norm": 0.22428563237190247,
"learning_rate": 0.0005929616941762691,
"loss": 1.6636204719543457,
"step": 2414
},
{
"epoch": 0.5041210224308816,
"grad_norm": 0.22870789468288422,
"learning_rate": 0.0005924634070383059,
"loss": 1.6800665855407715,
"step": 2416
},
{
"epoch": 0.5045383411580595,
"grad_norm": 0.15653257071971893,
"learning_rate": 0.0005919651199003425,
"loss": 1.5888636112213135,
"step": 2418
},
{
"epoch": 0.5049556598852374,
"grad_norm": 0.1919258087873459,
"learning_rate": 0.0005914668327623793,
"loss": 1.7502076625823975,
"step": 2420
},
{
"epoch": 0.5053729786124153,
"grad_norm": 0.18305595219135284,
"learning_rate": 0.0005909685456244161,
"loss": 1.7078887224197388,
"step": 2422
},
{
"epoch": 0.5057902973395931,
"grad_norm": 0.18780159950256348,
"learning_rate": 0.0005904702584864528,
"loss": 1.6881355047225952,
"step": 2424
},
{
"epoch": 0.506207616066771,
"grad_norm": 0.17898254096508026,
"learning_rate": 0.0005899719713484896,
"loss": 1.5203887224197388,
"step": 2426
},
{
"epoch": 0.5066249347939489,
"grad_norm": 0.17234547436237335,
"learning_rate": 0.0005894736842105263,
"loss": 1.7683515548706055,
"step": 2428
},
{
"epoch": 0.5070422535211268,
"grad_norm": 0.19153040647506714,
"learning_rate": 0.000588975397072563,
"loss": 1.6274034976959229,
"step": 2430
},
{
"epoch": 0.5074595722483046,
"grad_norm": 0.18104514479637146,
"learning_rate": 0.0005884771099345998,
"loss": 1.6722689867019653,
"step": 2432
},
{
"epoch": 0.5078768909754825,
"grad_norm": 0.18306469917297363,
"learning_rate": 0.0005879788227966366,
"loss": 1.5766488313674927,
"step": 2434
},
{
"epoch": 0.5082942097026604,
"grad_norm": 0.1755378693342209,
"learning_rate": 0.0005874805356586733,
"loss": 1.6823691129684448,
"step": 2436
},
{
"epoch": 0.5087115284298382,
"grad_norm": 0.17623284459114075,
"learning_rate": 0.0005869822485207101,
"loss": 1.7034425735473633,
"step": 2438
},
{
"epoch": 0.5091288471570161,
"grad_norm": 0.15370145440101624,
"learning_rate": 0.0005864839613827468,
"loss": 1.6950407028198242,
"step": 2440
},
{
"epoch": 0.5095461658841941,
"grad_norm": 0.21810156106948853,
"learning_rate": 0.0005859856742447836,
"loss": 1.51209557056427,
"step": 2442
},
{
"epoch": 0.509963484611372,
"grad_norm": 0.17423439025878906,
"learning_rate": 0.0005854873871068203,
"loss": 1.5613700151443481,
"step": 2444
},
{
"epoch": 0.5103808033385498,
"grad_norm": 0.17340213060379028,
"learning_rate": 0.000584989099968857,
"loss": 1.6352014541625977,
"step": 2446
},
{
"epoch": 0.5107981220657277,
"grad_norm": 0.19288502633571625,
"learning_rate": 0.0005844908128308939,
"loss": 1.7463655471801758,
"step": 2448
},
{
"epoch": 0.5112154407929056,
"grad_norm": 0.2064741998910904,
"learning_rate": 0.0005839925256929306,
"loss": 1.749002456665039,
"step": 2450
},
{
"epoch": 0.5116327595200835,
"grad_norm": 0.19839169085025787,
"learning_rate": 0.0005834942385549673,
"loss": 1.5387158393859863,
"step": 2452
},
{
"epoch": 0.5120500782472613,
"grad_norm": 0.2169366478919983,
"learning_rate": 0.0005829959514170041,
"loss": 1.716029405593872,
"step": 2454
},
{
"epoch": 0.5124673969744392,
"grad_norm": 0.1778964102268219,
"learning_rate": 0.0005824976642790408,
"loss": 1.6789476871490479,
"step": 2456
},
{
"epoch": 0.5128847157016171,
"grad_norm": 0.17874129116535187,
"learning_rate": 0.0005819993771410776,
"loss": 1.684020757675171,
"step": 2458
},
{
"epoch": 0.513302034428795,
"grad_norm": 0.18823137879371643,
"learning_rate": 0.0005815010900031143,
"loss": 1.5745269060134888,
"step": 2460
},
{
"epoch": 0.5137193531559728,
"grad_norm": 0.17944714426994324,
"learning_rate": 0.000581002802865151,
"loss": 1.6303616762161255,
"step": 2462
},
{
"epoch": 0.5141366718831507,
"grad_norm": 0.18502040207386017,
"learning_rate": 0.0005805045157271879,
"loss": 1.5294238328933716,
"step": 2464
},
{
"epoch": 0.5145539906103287,
"grad_norm": 0.16828668117523193,
"learning_rate": 0.0005800062285892246,
"loss": 1.7226977348327637,
"step": 2466
},
{
"epoch": 0.5149713093375066,
"grad_norm": 0.17077459394931793,
"learning_rate": 0.0005795079414512613,
"loss": 1.6081149578094482,
"step": 2468
},
{
"epoch": 0.5153886280646844,
"grad_norm": 0.2032557874917984,
"learning_rate": 0.0005790096543132981,
"loss": 1.6381583213806152,
"step": 2470
},
{
"epoch": 0.5158059467918623,
"grad_norm": 0.18319398164749146,
"learning_rate": 0.0005785113671753348,
"loss": 1.7247673273086548,
"step": 2472
},
{
"epoch": 0.5162232655190402,
"grad_norm": 0.19449344277381897,
"learning_rate": 0.0005780130800373715,
"loss": 1.747827410697937,
"step": 2474
},
{
"epoch": 0.5166405842462181,
"grad_norm": 0.17509305477142334,
"learning_rate": 0.0005775147928994083,
"loss": 1.7604385614395142,
"step": 2476
},
{
"epoch": 0.5170579029733959,
"grad_norm": 0.1723404973745346,
"learning_rate": 0.0005770165057614451,
"loss": 1.7105096578598022,
"step": 2478
},
{
"epoch": 0.5174752217005738,
"grad_norm": 0.21351687610149384,
"learning_rate": 0.0005765182186234819,
"loss": 1.5708328485488892,
"step": 2480
},
{
"epoch": 0.5178925404277517,
"grad_norm": 0.20518575608730316,
"learning_rate": 0.0005760199314855186,
"loss": 1.8066588640213013,
"step": 2482
},
{
"epoch": 0.5183098591549296,
"grad_norm": 0.18570895493030548,
"learning_rate": 0.0005755216443475553,
"loss": 1.7550853490829468,
"step": 2484
},
{
"epoch": 0.5187271778821074,
"grad_norm": 0.18616485595703125,
"learning_rate": 0.0005750233572095921,
"loss": 1.5576244592666626,
"step": 2486
},
{
"epoch": 0.5191444966092853,
"grad_norm": 0.16710995137691498,
"learning_rate": 0.0005745250700716288,
"loss": 1.6978063583374023,
"step": 2488
},
{
"epoch": 0.5195618153364632,
"grad_norm": 0.1688261330127716,
"learning_rate": 0.0005740267829336655,
"loss": 1.6968060731887817,
"step": 2490
},
{
"epoch": 0.5199791340636412,
"grad_norm": 0.18001560866832733,
"learning_rate": 0.0005735284957957024,
"loss": 1.6870050430297852,
"step": 2492
},
{
"epoch": 0.520396452790819,
"grad_norm": 0.16801725327968597,
"learning_rate": 0.0005730302086577391,
"loss": 1.5540623664855957,
"step": 2494
},
{
"epoch": 0.5208137715179969,
"grad_norm": 0.17628654837608337,
"learning_rate": 0.0005725319215197759,
"loss": 1.6644922494888306,
"step": 2496
},
{
"epoch": 0.5212310902451748,
"grad_norm": 0.21176116168498993,
"learning_rate": 0.0005720336343818126,
"loss": 1.6676300764083862,
"step": 2498
},
{
"epoch": 0.5216484089723527,
"grad_norm": 0.18427042663097382,
"learning_rate": 0.0005715353472438493,
"loss": 1.7856544256210327,
"step": 2500
},
{
"epoch": 0.5220657276995305,
"grad_norm": 0.1743779182434082,
"learning_rate": 0.000571037060105886,
"loss": 1.7691344022750854,
"step": 2502
},
{
"epoch": 0.5224830464267084,
"grad_norm": 0.2030993551015854,
"learning_rate": 0.0005705387729679227,
"loss": 1.6127548217773438,
"step": 2504
},
{
"epoch": 0.5229003651538863,
"grad_norm": 0.17827993631362915,
"learning_rate": 0.0005700404858299596,
"loss": 1.642172932624817,
"step": 2506
},
{
"epoch": 0.5233176838810641,
"grad_norm": 0.2054392695426941,
"learning_rate": 0.0005695421986919963,
"loss": 1.6635738611221313,
"step": 2508
},
{
"epoch": 0.523735002608242,
"grad_norm": 0.18387217819690704,
"learning_rate": 0.000569043911554033,
"loss": 1.7192704677581787,
"step": 2510
},
{
"epoch": 0.5241523213354199,
"grad_norm": 0.17396944761276245,
"learning_rate": 0.0005685456244160698,
"loss": 1.665990948677063,
"step": 2512
},
{
"epoch": 0.5245696400625978,
"grad_norm": 0.19370155036449432,
"learning_rate": 0.0005680473372781065,
"loss": 1.515099287033081,
"step": 2514
},
{
"epoch": 0.5249869587897756,
"grad_norm": 0.17007134854793549,
"learning_rate": 0.0005675490501401432,
"loss": 1.6382145881652832,
"step": 2516
},
{
"epoch": 0.5254042775169536,
"grad_norm": 0.16325214505195618,
"learning_rate": 0.00056705076300218,
"loss": 1.7413040399551392,
"step": 2518
},
{
"epoch": 0.5258215962441315,
"grad_norm": 0.20152249932289124,
"learning_rate": 0.0005665524758642167,
"loss": 1.6860828399658203,
"step": 2520
},
{
"epoch": 0.5262389149713094,
"grad_norm": 0.18658874928951263,
"learning_rate": 0.0005660541887262536,
"loss": 1.5169949531555176,
"step": 2522
},
{
"epoch": 0.5266562336984872,
"grad_norm": 0.17777974903583527,
"learning_rate": 0.0005655559015882903,
"loss": 1.694572925567627,
"step": 2524
},
{
"epoch": 0.5270735524256651,
"grad_norm": 0.17755714058876038,
"learning_rate": 0.000565057614450327,
"loss": 1.6401901245117188,
"step": 2526
},
{
"epoch": 0.527490871152843,
"grad_norm": 0.18403002619743347,
"learning_rate": 0.0005645593273123638,
"loss": 1.6839842796325684,
"step": 2528
},
{
"epoch": 0.5279081898800209,
"grad_norm": 0.21825653314590454,
"learning_rate": 0.0005640610401744005,
"loss": 1.6376888751983643,
"step": 2530
},
{
"epoch": 0.5283255086071987,
"grad_norm": 0.20095300674438477,
"learning_rate": 0.0005635627530364372,
"loss": 1.7161227464675903,
"step": 2532
},
{
"epoch": 0.5287428273343766,
"grad_norm": 0.16237320005893707,
"learning_rate": 0.000563064465898474,
"loss": 1.7353678941726685,
"step": 2534
},
{
"epoch": 0.5291601460615545,
"grad_norm": 0.17728379368782043,
"learning_rate": 0.0005625661787605108,
"loss": 1.6828662157058716,
"step": 2536
},
{
"epoch": 0.5295774647887324,
"grad_norm": 0.18883806467056274,
"learning_rate": 0.0005620678916225475,
"loss": 1.6642425060272217,
"step": 2538
},
{
"epoch": 0.5299947835159102,
"grad_norm": 0.16482634842395782,
"learning_rate": 0.0005615696044845843,
"loss": 1.6938458681106567,
"step": 2540
},
{
"epoch": 0.5304121022430882,
"grad_norm": 0.16626358032226562,
"learning_rate": 0.000561071317346621,
"loss": 1.621595859527588,
"step": 2542
},
{
"epoch": 0.5308294209702661,
"grad_norm": 0.18850813806056976,
"learning_rate": 0.0005605730302086578,
"loss": 1.6721577644348145,
"step": 2544
},
{
"epoch": 0.531246739697444,
"grad_norm": 0.1878683865070343,
"learning_rate": 0.0005600747430706945,
"loss": 1.6365470886230469,
"step": 2546
},
{
"epoch": 0.5316640584246218,
"grad_norm": 0.21200984716415405,
"learning_rate": 0.0005595764559327312,
"loss": 1.7305500507354736,
"step": 2548
},
{
"epoch": 0.5320813771517997,
"grad_norm": 0.22448937594890594,
"learning_rate": 0.0005590781687947681,
"loss": 1.6665229797363281,
"step": 2550
},
{
"epoch": 0.5324986958789776,
"grad_norm": 0.16661383211612701,
"learning_rate": 0.0005585798816568048,
"loss": 1.7415140867233276,
"step": 2552
},
{
"epoch": 0.5329160146061555,
"grad_norm": 0.2437662035226822,
"learning_rate": 0.0005580815945188415,
"loss": 1.7086694240570068,
"step": 2554
},
{
"epoch": 0.5333333333333333,
"grad_norm": 0.17701959609985352,
"learning_rate": 0.0005575833073808783,
"loss": 1.6117677688598633,
"step": 2556
},
{
"epoch": 0.5337506520605112,
"grad_norm": 0.18963120877742767,
"learning_rate": 0.000557085020242915,
"loss": 1.6729681491851807,
"step": 2558
},
{
"epoch": 0.5341679707876891,
"grad_norm": 0.20314951241016388,
"learning_rate": 0.0005565867331049518,
"loss": 1.5021780729293823,
"step": 2560
},
{
"epoch": 0.534585289514867,
"grad_norm": 0.1569170206785202,
"learning_rate": 0.0005560884459669885,
"loss": 1.6840097904205322,
"step": 2562
},
{
"epoch": 0.5350026082420448,
"grad_norm": 0.23780283331871033,
"learning_rate": 0.0005555901588290252,
"loss": 1.4439613819122314,
"step": 2564
},
{
"epoch": 0.5354199269692227,
"grad_norm": 0.18453800678253174,
"learning_rate": 0.0005550918716910621,
"loss": 1.6953411102294922,
"step": 2566
},
{
"epoch": 0.5358372456964007,
"grad_norm": 0.19585901498794556,
"learning_rate": 0.0005545935845530988,
"loss": 1.630983591079712,
"step": 2568
},
{
"epoch": 0.5362545644235786,
"grad_norm": 0.18498928844928741,
"learning_rate": 0.0005540952974151355,
"loss": 1.636730670928955,
"step": 2570
},
{
"epoch": 0.5366718831507564,
"grad_norm": 0.19784347712993622,
"learning_rate": 0.0005535970102771723,
"loss": 1.5415608882904053,
"step": 2572
},
{
"epoch": 0.5370892018779343,
"grad_norm": 0.15536262094974518,
"learning_rate": 0.000553098723139209,
"loss": 1.6956651210784912,
"step": 2574
},
{
"epoch": 0.5375065206051122,
"grad_norm": 0.18716371059417725,
"learning_rate": 0.0005526004360012457,
"loss": 1.6503138542175293,
"step": 2576
},
{
"epoch": 0.53792383933229,
"grad_norm": 0.22692058980464935,
"learning_rate": 0.0005521021488632825,
"loss": 1.6187713146209717,
"step": 2578
},
{
"epoch": 0.5383411580594679,
"grad_norm": 0.19440312683582306,
"learning_rate": 0.0005516038617253193,
"loss": 1.707751989364624,
"step": 2580
},
{
"epoch": 0.5387584767866458,
"grad_norm": 0.18238282203674316,
"learning_rate": 0.0005511055745873561,
"loss": 1.5600261688232422,
"step": 2582
},
{
"epoch": 0.5391757955138237,
"grad_norm": 0.20360791683197021,
"learning_rate": 0.0005506072874493927,
"loss": 1.7330106496810913,
"step": 2584
},
{
"epoch": 0.5395931142410015,
"grad_norm": 0.18367339670658112,
"learning_rate": 0.0005501090003114295,
"loss": 1.645473837852478,
"step": 2586
},
{
"epoch": 0.5400104329681794,
"grad_norm": 0.185563862323761,
"learning_rate": 0.0005496107131734662,
"loss": 1.6239993572235107,
"step": 2588
},
{
"epoch": 0.5404277516953573,
"grad_norm": 0.1883496344089508,
"learning_rate": 0.0005491124260355029,
"loss": 1.7215344905853271,
"step": 2590
},
{
"epoch": 0.5408450704225352,
"grad_norm": 0.18414616584777832,
"learning_rate": 0.0005486141388975397,
"loss": 1.6290435791015625,
"step": 2592
},
{
"epoch": 0.5412623891497131,
"grad_norm": 0.20297639071941376,
"learning_rate": 0.0005481158517595765,
"loss": 1.6351189613342285,
"step": 2594
},
{
"epoch": 0.541679707876891,
"grad_norm": 0.19413942098617554,
"learning_rate": 0.0005476175646216132,
"loss": 1.6903860569000244,
"step": 2596
},
{
"epoch": 0.5420970266040689,
"grad_norm": 0.19679142534732819,
"learning_rate": 0.00054711927748365,
"loss": 1.6611744165420532,
"step": 2598
},
{
"epoch": 0.5425143453312468,
"grad_norm": 0.18017566204071045,
"learning_rate": 0.0005466209903456867,
"loss": 1.5086408853530884,
"step": 2600
},
{
"epoch": 0.5429316640584246,
"grad_norm": 0.2049633264541626,
"learning_rate": 0.0005461227032077234,
"loss": 1.6443570852279663,
"step": 2602
},
{
"epoch": 0.5433489827856025,
"grad_norm": 0.204730823636055,
"learning_rate": 0.0005456244160697602,
"loss": 1.697648525238037,
"step": 2604
},
{
"epoch": 0.5437663015127804,
"grad_norm": 0.20954586565494537,
"learning_rate": 0.0005451261289317969,
"loss": 1.5692315101623535,
"step": 2606
},
{
"epoch": 0.5441836202399583,
"grad_norm": 0.17089051008224487,
"learning_rate": 0.0005446278417938338,
"loss": 1.708683967590332,
"step": 2608
},
{
"epoch": 0.5446009389671361,
"grad_norm": 0.17690244317054749,
"learning_rate": 0.0005441295546558705,
"loss": 1.7338215112686157,
"step": 2610
},
{
"epoch": 0.545018257694314,
"grad_norm": 0.17400382459163666,
"learning_rate": 0.0005436312675179072,
"loss": 1.7206354141235352,
"step": 2612
},
{
"epoch": 0.5454355764214919,
"grad_norm": 0.1846373975276947,
"learning_rate": 0.000543132980379944,
"loss": 1.5830307006835938,
"step": 2614
},
{
"epoch": 0.5458528951486697,
"grad_norm": 0.1858123540878296,
"learning_rate": 0.0005426346932419807,
"loss": 1.5219602584838867,
"step": 2616
},
{
"epoch": 0.5462702138758477,
"grad_norm": 0.1746957004070282,
"learning_rate": 0.0005421364061040174,
"loss": 1.6628625392913818,
"step": 2618
},
{
"epoch": 0.5466875326030256,
"grad_norm": 0.19798243045806885,
"learning_rate": 0.0005416381189660542,
"loss": 1.6740467548370361,
"step": 2620
},
{
"epoch": 0.5471048513302035,
"grad_norm": 0.1535359025001526,
"learning_rate": 0.0005411398318280909,
"loss": 1.5913817882537842,
"step": 2622
},
{
"epoch": 0.5475221700573814,
"grad_norm": 0.17042136192321777,
"learning_rate": 0.0005406415446901278,
"loss": 1.6836981773376465,
"step": 2624
},
{
"epoch": 0.5479394887845592,
"grad_norm": 0.21833325922489166,
"learning_rate": 0.0005401432575521645,
"loss": 1.5287742614746094,
"step": 2626
},
{
"epoch": 0.5483568075117371,
"grad_norm": 0.18353882431983948,
"learning_rate": 0.0005396449704142012,
"loss": 1.6056004762649536,
"step": 2628
},
{
"epoch": 0.548774126238915,
"grad_norm": 0.19037789106369019,
"learning_rate": 0.000539146683276238,
"loss": 1.5553654432296753,
"step": 2630
},
{
"epoch": 0.5491914449660928,
"grad_norm": 0.1939574033021927,
"learning_rate": 0.0005386483961382747,
"loss": 1.6138291358947754,
"step": 2632
},
{
"epoch": 0.5496087636932707,
"grad_norm": 0.1935683786869049,
"learning_rate": 0.0005381501090003114,
"loss": 1.5878697633743286,
"step": 2634
},
{
"epoch": 0.5500260824204486,
"grad_norm": 0.1670914888381958,
"learning_rate": 0.0005376518218623482,
"loss": 1.6278353929519653,
"step": 2636
},
{
"epoch": 0.5504434011476265,
"grad_norm": 0.18019060790538788,
"learning_rate": 0.000537153534724385,
"loss": 1.6293913125991821,
"step": 2638
},
{
"epoch": 0.5508607198748043,
"grad_norm": 0.19435915350914001,
"learning_rate": 0.0005366552475864217,
"loss": 1.7511961460113525,
"step": 2640
},
{
"epoch": 0.5512780386019822,
"grad_norm": 0.19905701279640198,
"learning_rate": 0.0005361569604484585,
"loss": 1.4236350059509277,
"step": 2642
},
{
"epoch": 0.5516953573291602,
"grad_norm": 0.18854565918445587,
"learning_rate": 0.0005356586733104952,
"loss": 1.6020326614379883,
"step": 2644
},
{
"epoch": 0.5521126760563381,
"grad_norm": 0.2206660807132721,
"learning_rate": 0.000535160386172532,
"loss": 1.6151522397994995,
"step": 2646
},
{
"epoch": 0.5525299947835159,
"grad_norm": 0.2025166153907776,
"learning_rate": 0.0005346620990345687,
"loss": 1.6654980182647705,
"step": 2648
},
{
"epoch": 0.5529473135106938,
"grad_norm": 0.18366900086402893,
"learning_rate": 0.0005341638118966054,
"loss": 1.7074745893478394,
"step": 2650
},
{
"epoch": 0.5533646322378717,
"grad_norm": 0.19196325540542603,
"learning_rate": 0.0005336655247586423,
"loss": 1.6605453491210938,
"step": 2652
},
{
"epoch": 0.5537819509650496,
"grad_norm": 0.19255101680755615,
"learning_rate": 0.000533167237620679,
"loss": 1.6116297245025635,
"step": 2654
},
{
"epoch": 0.5541992696922274,
"grad_norm": 0.2326728105545044,
"learning_rate": 0.0005326689504827157,
"loss": 1.593017578125,
"step": 2656
},
{
"epoch": 0.5546165884194053,
"grad_norm": 0.17418722808361053,
"learning_rate": 0.0005321706633447525,
"loss": 1.6630582809448242,
"step": 2658
},
{
"epoch": 0.5550339071465832,
"grad_norm": 0.18985475599765778,
"learning_rate": 0.0005316723762067892,
"loss": 1.615740418434143,
"step": 2660
},
{
"epoch": 0.5554512258737611,
"grad_norm": 0.1669873595237732,
"learning_rate": 0.000531174089068826,
"loss": 1.8242087364196777,
"step": 2662
},
{
"epoch": 0.5558685446009389,
"grad_norm": 0.16505779325962067,
"learning_rate": 0.0005306758019308627,
"loss": 1.5801950693130493,
"step": 2664
},
{
"epoch": 0.5562858633281168,
"grad_norm": 0.18953555822372437,
"learning_rate": 0.0005301775147928994,
"loss": 1.6009870767593384,
"step": 2666
},
{
"epoch": 0.5567031820552947,
"grad_norm": 0.16053496301174164,
"learning_rate": 0.0005296792276549362,
"loss": 1.741046667098999,
"step": 2668
},
{
"epoch": 0.5571205007824727,
"grad_norm": 0.18851445615291595,
"learning_rate": 0.0005291809405169729,
"loss": 1.6481761932373047,
"step": 2670
},
{
"epoch": 0.5575378195096505,
"grad_norm": 0.18606077134609222,
"learning_rate": 0.0005286826533790097,
"loss": 1.6813530921936035,
"step": 2672
},
{
"epoch": 0.5579551382368284,
"grad_norm": 0.1917671114206314,
"learning_rate": 0.0005281843662410464,
"loss": 1.7276756763458252,
"step": 2674
},
{
"epoch": 0.5583724569640063,
"grad_norm": 0.2070527821779251,
"learning_rate": 0.0005276860791030831,
"loss": 1.6686636209487915,
"step": 2676
},
{
"epoch": 0.5587897756911842,
"grad_norm": 0.18664246797561646,
"learning_rate": 0.0005271877919651199,
"loss": 1.6385400295257568,
"step": 2678
},
{
"epoch": 0.559207094418362,
"grad_norm": 0.17125488817691803,
"learning_rate": 0.0005266895048271566,
"loss": 1.6458193063735962,
"step": 2680
},
{
"epoch": 0.5596244131455399,
"grad_norm": 0.17900311946868896,
"learning_rate": 0.0005261912176891933,
"loss": 1.7152608633041382,
"step": 2682
},
{
"epoch": 0.5600417318727178,
"grad_norm": 0.22325071692466736,
"learning_rate": 0.0005256929305512302,
"loss": 1.5353511571884155,
"step": 2684
},
{
"epoch": 0.5604590505998956,
"grad_norm": 0.18414054811000824,
"learning_rate": 0.0005251946434132669,
"loss": 1.5491306781768799,
"step": 2686
},
{
"epoch": 0.5608763693270735,
"grad_norm": 0.20302103459835052,
"learning_rate": 0.0005246963562753037,
"loss": 1.6303091049194336,
"step": 2688
},
{
"epoch": 0.5612936880542514,
"grad_norm": 0.172519251704216,
"learning_rate": 0.0005241980691373404,
"loss": 1.7245980501174927,
"step": 2690
},
{
"epoch": 0.5617110067814293,
"grad_norm": 0.1944926381111145,
"learning_rate": 0.0005236997819993771,
"loss": 1.6948201656341553,
"step": 2692
},
{
"epoch": 0.5621283255086073,
"grad_norm": 0.16919668018817902,
"learning_rate": 0.0005232014948614139,
"loss": 1.624497652053833,
"step": 2694
},
{
"epoch": 0.5625456442357851,
"grad_norm": 0.18296246230602264,
"learning_rate": 0.0005227032077234506,
"loss": 1.7612922191619873,
"step": 2696
},
{
"epoch": 0.562962962962963,
"grad_norm": 0.22652164101600647,
"learning_rate": 0.0005222049205854874,
"loss": 1.5955829620361328,
"step": 2698
},
{
"epoch": 0.5633802816901409,
"grad_norm": 0.18193946778774261,
"learning_rate": 0.0005217066334475242,
"loss": 1.803607702255249,
"step": 2700
},
{
"epoch": 0.5637976004173187,
"grad_norm": 0.1849241554737091,
"learning_rate": 0.0005212083463095609,
"loss": 1.700540542602539,
"step": 2702
},
{
"epoch": 0.5642149191444966,
"grad_norm": 0.16596536338329315,
"learning_rate": 0.0005207100591715976,
"loss": 1.6775227785110474,
"step": 2704
},
{
"epoch": 0.5646322378716745,
"grad_norm": 0.17335398495197296,
"learning_rate": 0.0005202117720336344,
"loss": 1.5570757389068604,
"step": 2706
},
{
"epoch": 0.5650495565988524,
"grad_norm": 0.17634885013103485,
"learning_rate": 0.0005197134848956711,
"loss": 1.5774015188217163,
"step": 2708
},
{
"epoch": 0.5654668753260302,
"grad_norm": 0.19291938841342926,
"learning_rate": 0.000519215197757708,
"loss": 1.5656869411468506,
"step": 2710
},
{
"epoch": 0.5658841940532081,
"grad_norm": 0.17061962187290192,
"learning_rate": 0.0005187169106197447,
"loss": 1.609908103942871,
"step": 2712
},
{
"epoch": 0.566301512780386,
"grad_norm": 0.1720571666955948,
"learning_rate": 0.0005182186234817814,
"loss": 1.7176014184951782,
"step": 2714
},
{
"epoch": 0.5667188315075639,
"grad_norm": 0.16794128715991974,
"learning_rate": 0.0005177203363438182,
"loss": 1.6396007537841797,
"step": 2716
},
{
"epoch": 0.5671361502347417,
"grad_norm": 0.17661422491073608,
"learning_rate": 0.0005172220492058549,
"loss": 1.7083172798156738,
"step": 2718
},
{
"epoch": 0.5675534689619197,
"grad_norm": 0.19686675071716309,
"learning_rate": 0.0005167237620678916,
"loss": 1.7523099184036255,
"step": 2720
},
{
"epoch": 0.5679707876890976,
"grad_norm": 0.1818755865097046,
"learning_rate": 0.0005162254749299284,
"loss": 1.7118453979492188,
"step": 2722
},
{
"epoch": 0.5683881064162755,
"grad_norm": 0.17738768458366394,
"learning_rate": 0.0005157271877919651,
"loss": 1.699044108390808,
"step": 2724
},
{
"epoch": 0.5688054251434533,
"grad_norm": 0.20979493856430054,
"learning_rate": 0.000515228900654002,
"loss": 1.7189500331878662,
"step": 2726
},
{
"epoch": 0.5692227438706312,
"grad_norm": 0.21231696009635925,
"learning_rate": 0.0005147306135160387,
"loss": 1.7079036235809326,
"step": 2728
},
{
"epoch": 0.5696400625978091,
"grad_norm": 0.17614737153053284,
"learning_rate": 0.0005142323263780754,
"loss": 1.69643235206604,
"step": 2730
},
{
"epoch": 0.570057381324987,
"grad_norm": 0.20790863037109375,
"learning_rate": 0.0005137340392401122,
"loss": 1.6363704204559326,
"step": 2732
},
{
"epoch": 0.5704747000521648,
"grad_norm": 0.21195682883262634,
"learning_rate": 0.0005132357521021489,
"loss": 1.6397018432617188,
"step": 2734
},
{
"epoch": 0.5708920187793427,
"grad_norm": 0.2375713437795639,
"learning_rate": 0.0005127374649641856,
"loss": 1.6042075157165527,
"step": 2736
},
{
"epoch": 0.5713093375065206,
"grad_norm": 0.17576976120471954,
"learning_rate": 0.0005122391778262224,
"loss": 1.5349665880203247,
"step": 2738
},
{
"epoch": 0.5717266562336984,
"grad_norm": 0.16033388674259186,
"learning_rate": 0.0005117408906882592,
"loss": 1.7466727495193481,
"step": 2740
},
{
"epoch": 0.5721439749608763,
"grad_norm": 0.16839449107646942,
"learning_rate": 0.0005112426035502959,
"loss": 1.6488425731658936,
"step": 2742
},
{
"epoch": 0.5725612936880543,
"grad_norm": 0.15866801142692566,
"learning_rate": 0.0005107443164123327,
"loss": 1.5046507120132446,
"step": 2744
},
{
"epoch": 0.5729786124152322,
"grad_norm": 0.18288397789001465,
"learning_rate": 0.0005102460292743694,
"loss": 1.64163076877594,
"step": 2746
},
{
"epoch": 0.57339593114241,
"grad_norm": 0.24673451483249664,
"learning_rate": 0.0005097477421364062,
"loss": 1.308207392692566,
"step": 2748
},
{
"epoch": 0.5738132498695879,
"grad_norm": 0.21381480991840363,
"learning_rate": 0.0005092494549984429,
"loss": 1.6450862884521484,
"step": 2750
},
{
"epoch": 0.5742305685967658,
"grad_norm": 0.21157296001911163,
"learning_rate": 0.0005087511678604796,
"loss": 1.5671970844268799,
"step": 2752
},
{
"epoch": 0.5746478873239437,
"grad_norm": 0.19186222553253174,
"learning_rate": 0.0005082528807225163,
"loss": 1.6529240608215332,
"step": 2754
},
{
"epoch": 0.5750652060511215,
"grad_norm": 0.2344437688589096,
"learning_rate": 0.0005077545935845531,
"loss": 1.7823957204818726,
"step": 2756
},
{
"epoch": 0.5754825247782994,
"grad_norm": 0.18494389951229095,
"learning_rate": 0.0005072563064465899,
"loss": 1.6493897438049316,
"step": 2758
},
{
"epoch": 0.5758998435054773,
"grad_norm": 0.21340534090995789,
"learning_rate": 0.0005067580193086266,
"loss": 1.5306980609893799,
"step": 2760
},
{
"epoch": 0.5763171622326552,
"grad_norm": 0.20003263652324677,
"learning_rate": 0.0005062597321706633,
"loss": 1.6893612146377563,
"step": 2762
},
{
"epoch": 0.576734480959833,
"grad_norm": 0.19722574949264526,
"learning_rate": 0.0005057614450327001,
"loss": 1.5928115844726562,
"step": 2764
},
{
"epoch": 0.5771517996870109,
"grad_norm": 0.2228558510541916,
"learning_rate": 0.0005052631578947368,
"loss": 1.545350432395935,
"step": 2766
},
{
"epoch": 0.5775691184141888,
"grad_norm": 0.17985211312770844,
"learning_rate": 0.0005047648707567735,
"loss": 1.5899455547332764,
"step": 2768
},
{
"epoch": 0.5779864371413668,
"grad_norm": 0.16370762884616852,
"learning_rate": 0.0005042665836188104,
"loss": 1.606586217880249,
"step": 2770
},
{
"epoch": 0.5784037558685446,
"grad_norm": 0.17988185584545135,
"learning_rate": 0.0005037682964808471,
"loss": 1.7284839153289795,
"step": 2772
},
{
"epoch": 0.5788210745957225,
"grad_norm": 0.17850655317306519,
"learning_rate": 0.0005032700093428839,
"loss": 1.7124626636505127,
"step": 2774
},
{
"epoch": 0.5792383933229004,
"grad_norm": 0.1940651386976242,
"learning_rate": 0.0005027717222049206,
"loss": 1.6954848766326904,
"step": 2776
},
{
"epoch": 0.5796557120500783,
"grad_norm": 0.16563449800014496,
"learning_rate": 0.0005022734350669573,
"loss": 1.7141410112380981,
"step": 2778
},
{
"epoch": 0.5800730307772561,
"grad_norm": 0.2002563178539276,
"learning_rate": 0.0005017751479289941,
"loss": 1.6658897399902344,
"step": 2780
},
{
"epoch": 0.580490349504434,
"grad_norm": 0.2039714753627777,
"learning_rate": 0.0005012768607910308,
"loss": 1.6996421813964844,
"step": 2782
},
{
"epoch": 0.5809076682316119,
"grad_norm": 0.2018059343099594,
"learning_rate": 0.0005007785736530675,
"loss": 1.5792217254638672,
"step": 2784
},
{
"epoch": 0.5813249869587898,
"grad_norm": 0.19589859247207642,
"learning_rate": 0.0005002802865151044,
"loss": 1.6206440925598145,
"step": 2786
},
{
"epoch": 0.5817423056859676,
"grad_norm": 0.18769143521785736,
"learning_rate": 0.0004997819993771411,
"loss": 1.64694344997406,
"step": 2788
},
{
"epoch": 0.5821596244131455,
"grad_norm": 0.2359609603881836,
"learning_rate": 0.0004992837122391779,
"loss": 1.4787166118621826,
"step": 2790
},
{
"epoch": 0.5825769431403234,
"grad_norm": 0.1819252371788025,
"learning_rate": 0.0004987854251012146,
"loss": 1.4756464958190918,
"step": 2792
},
{
"epoch": 0.5829942618675013,
"grad_norm": 0.19079644978046417,
"learning_rate": 0.0004982871379632513,
"loss": 1.6275835037231445,
"step": 2794
},
{
"epoch": 0.5834115805946792,
"grad_norm": 0.19106005132198334,
"learning_rate": 0.0004977888508252881,
"loss": 1.5918662548065186,
"step": 2796
},
{
"epoch": 0.5838288993218571,
"grad_norm": 0.23186901211738586,
"learning_rate": 0.0004972905636873248,
"loss": 1.5772104263305664,
"step": 2798
},
{
"epoch": 0.584246218049035,
"grad_norm": 0.1947910040616989,
"learning_rate": 0.0004967922765493616,
"loss": 1.7197240591049194,
"step": 2800
},
{
"epoch": 0.5846635367762129,
"grad_norm": 0.16244561970233917,
"learning_rate": 0.0004962939894113984,
"loss": 1.715684413909912,
"step": 2802
},
{
"epoch": 0.5850808555033907,
"grad_norm": 0.16783419251441956,
"learning_rate": 0.0004957957022734351,
"loss": 1.582826852798462,
"step": 2804
},
{
"epoch": 0.5854981742305686,
"grad_norm": 0.21026046574115753,
"learning_rate": 0.0004952974151354718,
"loss": 1.5453834533691406,
"step": 2806
},
{
"epoch": 0.5859154929577465,
"grad_norm": 0.23682411015033722,
"learning_rate": 0.0004947991279975086,
"loss": 1.697816252708435,
"step": 2808
},
{
"epoch": 0.5863328116849243,
"grad_norm": 0.19026942551136017,
"learning_rate": 0.0004943008408595453,
"loss": 1.66938316822052,
"step": 2810
},
{
"epoch": 0.5867501304121022,
"grad_norm": 0.15915033221244812,
"learning_rate": 0.0004938025537215821,
"loss": 1.6958580017089844,
"step": 2812
},
{
"epoch": 0.5871674491392801,
"grad_norm": 0.19176815450191498,
"learning_rate": 0.0004933042665836189,
"loss": 1.724020004272461,
"step": 2814
},
{
"epoch": 0.587584767866458,
"grad_norm": 0.17253069579601288,
"learning_rate": 0.0004928059794456556,
"loss": 1.6651839017868042,
"step": 2816
},
{
"epoch": 0.5880020865936358,
"grad_norm": 0.17185230553150177,
"learning_rate": 0.0004923076923076924,
"loss": 1.5924116373062134,
"step": 2818
},
{
"epoch": 0.5884194053208138,
"grad_norm": 0.20549938082695007,
"learning_rate": 0.0004918094051697291,
"loss": 1.878544807434082,
"step": 2820
},
{
"epoch": 0.5888367240479917,
"grad_norm": 0.17337928712368011,
"learning_rate": 0.0004913111180317658,
"loss": 1.5607351064682007,
"step": 2822
},
{
"epoch": 0.5892540427751696,
"grad_norm": 0.18207484483718872,
"learning_rate": 0.0004908128308938026,
"loss": 1.5620158910751343,
"step": 2824
},
{
"epoch": 0.5896713615023474,
"grad_norm": 0.21272793412208557,
"learning_rate": 0.0004903145437558393,
"loss": 1.557486891746521,
"step": 2826
},
{
"epoch": 0.5900886802295253,
"grad_norm": 0.16419966518878937,
"learning_rate": 0.0004898162566178762,
"loss": 1.646332025527954,
"step": 2828
},
{
"epoch": 0.5905059989567032,
"grad_norm": 0.20964190363883972,
"learning_rate": 0.0004893179694799129,
"loss": 1.6490819454193115,
"step": 2830
},
{
"epoch": 0.5909233176838811,
"grad_norm": 0.19490443170070648,
"learning_rate": 0.0004888196823419496,
"loss": 1.7113087177276611,
"step": 2832
},
{
"epoch": 0.5913406364110589,
"grad_norm": 0.18936558067798615,
"learning_rate": 0.0004883213952039864,
"loss": 1.6554588079452515,
"step": 2834
},
{
"epoch": 0.5917579551382368,
"grad_norm": 0.18321694433689117,
"learning_rate": 0.000487823108066023,
"loss": 1.5866620540618896,
"step": 2836
},
{
"epoch": 0.5921752738654147,
"grad_norm": 0.1745002418756485,
"learning_rate": 0.00048732482092805977,
"loss": 1.6719403266906738,
"step": 2838
},
{
"epoch": 0.5925925925925926,
"grad_norm": 0.1661233752965927,
"learning_rate": 0.00048682653379009653,
"loss": 1.6308550834655762,
"step": 2840
},
{
"epoch": 0.5930099113197704,
"grad_norm": 0.20850799977779388,
"learning_rate": 0.0004863282466521333,
"loss": 1.6365723609924316,
"step": 2842
},
{
"epoch": 0.5934272300469483,
"grad_norm": 0.1738293319940567,
"learning_rate": 0.00048582995951417,
"loss": 1.5891228914260864,
"step": 2844
},
{
"epoch": 0.5938445487741263,
"grad_norm": 0.16520492732524872,
"learning_rate": 0.0004853316723762068,
"loss": 1.7726880311965942,
"step": 2846
},
{
"epoch": 0.5942618675013042,
"grad_norm": 0.2029978185892105,
"learning_rate": 0.00048483338523824354,
"loss": 1.5459405183792114,
"step": 2848
},
{
"epoch": 0.594679186228482,
"grad_norm": 0.17381776869297028,
"learning_rate": 0.0004843350981002803,
"loss": 1.7816643714904785,
"step": 2850
},
{
"epoch": 0.5950965049556599,
"grad_norm": 0.19975237548351288,
"learning_rate": 0.000483836810962317,
"loss": 1.5473201274871826,
"step": 2852
},
{
"epoch": 0.5955138236828378,
"grad_norm": 0.18777495622634888,
"learning_rate": 0.0004833385238243538,
"loss": 1.6488192081451416,
"step": 2854
},
{
"epoch": 0.5959311424100157,
"grad_norm": 0.1772993803024292,
"learning_rate": 0.00048284023668639054,
"loss": 1.6942178010940552,
"step": 2856
},
{
"epoch": 0.5963484611371935,
"grad_norm": 0.2237555831670761,
"learning_rate": 0.0004823419495484273,
"loss": 1.6374008655548096,
"step": 2858
},
{
"epoch": 0.5967657798643714,
"grad_norm": 0.22892020642757416,
"learning_rate": 0.000481843662410464,
"loss": 1.5714294910430908,
"step": 2860
},
{
"epoch": 0.5971830985915493,
"grad_norm": 0.19660325348377228,
"learning_rate": 0.0004813453752725008,
"loss": 1.6836249828338623,
"step": 2862
},
{
"epoch": 0.5976004173187272,
"grad_norm": 0.20764879882335663,
"learning_rate": 0.00048084708813453755,
"loss": 1.663045883178711,
"step": 2864
},
{
"epoch": 0.598017736045905,
"grad_norm": 0.1717427372932434,
"learning_rate": 0.0004803488009965743,
"loss": 1.7530301809310913,
"step": 2866
},
{
"epoch": 0.5984350547730829,
"grad_norm": 0.2102285474538803,
"learning_rate": 0.000479850513858611,
"loss": 1.6049411296844482,
"step": 2868
},
{
"epoch": 0.5988523735002608,
"grad_norm": 0.17871743440628052,
"learning_rate": 0.0004793522267206478,
"loss": 1.7112499475479126,
"step": 2870
},
{
"epoch": 0.5992696922274388,
"grad_norm": 0.1878865659236908,
"learning_rate": 0.00047885393958268456,
"loss": 1.634444236755371,
"step": 2872
},
{
"epoch": 0.5996870109546166,
"grad_norm": 0.18738122284412384,
"learning_rate": 0.00047835565244472127,
"loss": 1.6382564306259155,
"step": 2874
},
{
"epoch": 0.6001043296817945,
"grad_norm": 0.17012278735637665,
"learning_rate": 0.00047785736530675803,
"loss": 1.5668729543685913,
"step": 2876
},
{
"epoch": 0.6003129890453834,
"eval_loss": 1.6334973573684692,
"eval_runtime": 25.5462,
"eval_samples_per_second": 60.674,
"eval_steps_per_second": 7.594,
"step": 2877
},
{
"epoch": 0.6005216484089724,
"grad_norm": 0.16799584031105042,
"learning_rate": 0.0004773590781687948,
"loss": 1.6411515474319458,
"step": 2878
},
{
"epoch": 0.6009389671361502,
"grad_norm": 0.21026895940303802,
"learning_rate": 0.00047686079103083156,
"loss": 1.5678623914718628,
"step": 2880
},
{
"epoch": 0.6013562858633281,
"grad_norm": 0.19277074933052063,
"learning_rate": 0.00047636250389286827,
"loss": 1.7933660745620728,
"step": 2882
},
{
"epoch": 0.601773604590506,
"grad_norm": 0.17304041981697083,
"learning_rate": 0.00047586421675490504,
"loss": 1.6832016706466675,
"step": 2884
},
{
"epoch": 0.6021909233176839,
"grad_norm": 0.17196182906627655,
"learning_rate": 0.0004753659296169418,
"loss": 1.5628772974014282,
"step": 2886
},
{
"epoch": 0.6026082420448617,
"grad_norm": 0.18928901851177216,
"learning_rate": 0.00047486764247897857,
"loss": 1.625766396522522,
"step": 2888
},
{
"epoch": 0.6030255607720396,
"grad_norm": 0.2047906070947647,
"learning_rate": 0.0004743693553410153,
"loss": 1.4966003894805908,
"step": 2890
},
{
"epoch": 0.6034428794992175,
"grad_norm": 0.17776289582252502,
"learning_rate": 0.00047387106820305204,
"loss": 1.7307140827178955,
"step": 2892
},
{
"epoch": 0.6038601982263954,
"grad_norm": 0.1764279305934906,
"learning_rate": 0.0004733727810650888,
"loss": 1.7942173480987549,
"step": 2894
},
{
"epoch": 0.6042775169535733,
"grad_norm": 0.1908806711435318,
"learning_rate": 0.00047287449392712557,
"loss": 1.4785549640655518,
"step": 2896
},
{
"epoch": 0.6046948356807512,
"grad_norm": 0.1910744160413742,
"learning_rate": 0.0004723762067891623,
"loss": 1.695818305015564,
"step": 2898
},
{
"epoch": 0.6051121544079291,
"grad_norm": 0.1757185459136963,
"learning_rate": 0.00047187791965119905,
"loss": 1.6474026441574097,
"step": 2900
},
{
"epoch": 0.605529473135107,
"grad_norm": 0.18998144567012787,
"learning_rate": 0.0004713796325132358,
"loss": 1.6418334245681763,
"step": 2902
},
{
"epoch": 0.6059467918622848,
"grad_norm": 0.19940714538097382,
"learning_rate": 0.0004708813453752726,
"loss": 1.650315284729004,
"step": 2904
},
{
"epoch": 0.6063641105894627,
"grad_norm": 0.19629891216754913,
"learning_rate": 0.0004703830582373093,
"loss": 1.6142354011535645,
"step": 2906
},
{
"epoch": 0.6067814293166406,
"grad_norm": 0.19428136944770813,
"learning_rate": 0.00046988477109934605,
"loss": 1.5597343444824219,
"step": 2908
},
{
"epoch": 0.6071987480438185,
"grad_norm": 0.1555328220129013,
"learning_rate": 0.0004693864839613828,
"loss": 1.6591336727142334,
"step": 2910
},
{
"epoch": 0.6076160667709963,
"grad_norm": 0.17966602742671967,
"learning_rate": 0.0004688881968234196,
"loss": 1.544377088546753,
"step": 2912
},
{
"epoch": 0.6080333854981742,
"grad_norm": 0.18002371490001678,
"learning_rate": 0.0004683899096854563,
"loss": 1.77950119972229,
"step": 2914
},
{
"epoch": 0.6084507042253521,
"grad_norm": 0.1758643537759781,
"learning_rate": 0.00046789162254749306,
"loss": 1.751016616821289,
"step": 2916
},
{
"epoch": 0.60886802295253,
"grad_norm": 0.1944553703069687,
"learning_rate": 0.0004673933354095297,
"loss": 1.6245421171188354,
"step": 2918
},
{
"epoch": 0.6092853416797078,
"grad_norm": 0.16478294134140015,
"learning_rate": 0.0004668950482715665,
"loss": 1.6343562602996826,
"step": 2920
},
{
"epoch": 0.6097026604068858,
"grad_norm": 0.18271125853061676,
"learning_rate": 0.00046639676113360325,
"loss": 1.5838673114776611,
"step": 2922
},
{
"epoch": 0.6101199791340637,
"grad_norm": 0.1791974902153015,
"learning_rate": 0.00046589847399563996,
"loss": 1.5538192987442017,
"step": 2924
},
{
"epoch": 0.6105372978612416,
"grad_norm": 0.1821575164794922,
"learning_rate": 0.0004654001868576767,
"loss": 1.7037311792373657,
"step": 2926
},
{
"epoch": 0.6109546165884194,
"grad_norm": 0.17619025707244873,
"learning_rate": 0.0004649018997197135,
"loss": 1.686895728111267,
"step": 2928
},
{
"epoch": 0.6113719353155973,
"grad_norm": 0.15952201187610626,
"learning_rate": 0.00046440361258175025,
"loss": 1.6833465099334717,
"step": 2930
},
{
"epoch": 0.6117892540427752,
"grad_norm": 0.23567429184913635,
"learning_rate": 0.00046390532544378696,
"loss": 1.4512883424758911,
"step": 2932
},
{
"epoch": 0.612206572769953,
"grad_norm": 0.19210046529769897,
"learning_rate": 0.0004634070383058237,
"loss": 1.7234846353530884,
"step": 2934
},
{
"epoch": 0.6126238914971309,
"grad_norm": 0.1734471470117569,
"learning_rate": 0.0004629087511678605,
"loss": 1.7545769214630127,
"step": 2936
},
{
"epoch": 0.6130412102243088,
"grad_norm": 0.18496385216712952,
"learning_rate": 0.0004624104640298972,
"loss": 1.572465181350708,
"step": 2938
},
{
"epoch": 0.6134585289514867,
"grad_norm": 0.2221444547176361,
"learning_rate": 0.00046191217689193397,
"loss": 1.519800066947937,
"step": 2940
},
{
"epoch": 0.6138758476786645,
"grad_norm": 0.24368079006671906,
"learning_rate": 0.00046141388975397073,
"loss": 1.6060869693756104,
"step": 2942
},
{
"epoch": 0.6142931664058424,
"grad_norm": 0.16458842158317566,
"learning_rate": 0.0004609156026160075,
"loss": 1.6020828485488892,
"step": 2944
},
{
"epoch": 0.6147104851330203,
"grad_norm": 0.1970018744468689,
"learning_rate": 0.0004604173154780442,
"loss": 1.6586238145828247,
"step": 2946
},
{
"epoch": 0.6151278038601983,
"grad_norm": 0.1822272688150406,
"learning_rate": 0.00045991902834008097,
"loss": 1.720170497894287,
"step": 2948
},
{
"epoch": 0.6155451225873761,
"grad_norm": 0.17016743123531342,
"learning_rate": 0.00045942074120211774,
"loss": 1.5761950016021729,
"step": 2950
},
{
"epoch": 0.615962441314554,
"grad_norm": 0.2120160609483719,
"learning_rate": 0.0004589224540641545,
"loss": 1.6407349109649658,
"step": 2952
},
{
"epoch": 0.6163797600417319,
"grad_norm": 0.16476835310459137,
"learning_rate": 0.0004584241669261912,
"loss": 1.7095156908035278,
"step": 2954
},
{
"epoch": 0.6167970787689098,
"grad_norm": 0.1890491098165512,
"learning_rate": 0.000457925879788228,
"loss": 1.591587781906128,
"step": 2956
},
{
"epoch": 0.6172143974960876,
"grad_norm": 0.18666179478168488,
"learning_rate": 0.00045742759265026474,
"loss": 1.5489561557769775,
"step": 2958
},
{
"epoch": 0.6176317162232655,
"grad_norm": 0.2605527341365814,
"learning_rate": 0.0004569293055123015,
"loss": 1.5896910429000854,
"step": 2960
},
{
"epoch": 0.6180490349504434,
"grad_norm": 0.16350412368774414,
"learning_rate": 0.0004564310183743382,
"loss": 1.6582103967666626,
"step": 2962
},
{
"epoch": 0.6184663536776213,
"grad_norm": 0.18763825297355652,
"learning_rate": 0.000455932731236375,
"loss": 1.7946631908416748,
"step": 2964
},
{
"epoch": 0.6188836724047991,
"grad_norm": 0.22416430711746216,
"learning_rate": 0.00045543444409841175,
"loss": 1.6434133052825928,
"step": 2966
},
{
"epoch": 0.619300991131977,
"grad_norm": 0.17112663388252258,
"learning_rate": 0.0004549361569604485,
"loss": 1.5859663486480713,
"step": 2968
},
{
"epoch": 0.6197183098591549,
"grad_norm": 0.22274653613567352,
"learning_rate": 0.0004544378698224852,
"loss": 1.6013418436050415,
"step": 2970
},
{
"epoch": 0.6201356285863329,
"grad_norm": 0.19702008366584778,
"learning_rate": 0.000453939582684522,
"loss": 1.6691360473632812,
"step": 2972
},
{
"epoch": 0.6205529473135107,
"grad_norm": 0.1698775291442871,
"learning_rate": 0.00045344129554655875,
"loss": 1.8140881061553955,
"step": 2974
},
{
"epoch": 0.6209702660406886,
"grad_norm": 0.1693205088376999,
"learning_rate": 0.0004529430084085955,
"loss": 1.6525713205337524,
"step": 2976
},
{
"epoch": 0.6213875847678665,
"grad_norm": 0.1770809441804886,
"learning_rate": 0.00045244472127063223,
"loss": 1.5939429998397827,
"step": 2978
},
{
"epoch": 0.6218049034950444,
"grad_norm": 0.1748695820569992,
"learning_rate": 0.000451946434132669,
"loss": 1.5831851959228516,
"step": 2980
},
{
"epoch": 0.6222222222222222,
"grad_norm": 0.19646704196929932,
"learning_rate": 0.00045144814699470576,
"loss": 1.6287367343902588,
"step": 2982
},
{
"epoch": 0.6226395409494001,
"grad_norm": 0.1683228760957718,
"learning_rate": 0.00045094985985674247,
"loss": 1.6986901760101318,
"step": 2984
},
{
"epoch": 0.623056859676578,
"grad_norm": 0.1828620731830597,
"learning_rate": 0.00045045157271877924,
"loss": 1.7912113666534424,
"step": 2986
},
{
"epoch": 0.6234741784037559,
"grad_norm": 0.16508342325687408,
"learning_rate": 0.000449953285580816,
"loss": 1.5801665782928467,
"step": 2988
},
{
"epoch": 0.6238914971309337,
"grad_norm": 0.15644466876983643,
"learning_rate": 0.00044945499844285277,
"loss": 1.6213196516036987,
"step": 2990
},
{
"epoch": 0.6243088158581116,
"grad_norm": 0.16748785972595215,
"learning_rate": 0.0004489567113048895,
"loss": 1.6507726907730103,
"step": 2992
},
{
"epoch": 0.6247261345852895,
"grad_norm": 0.18665555119514465,
"learning_rate": 0.00044845842416692624,
"loss": 1.7788985967636108,
"step": 2994
},
{
"epoch": 0.6251434533124673,
"grad_norm": 0.21414385735988617,
"learning_rate": 0.000447960137028963,
"loss": 1.587815284729004,
"step": 2996
},
{
"epoch": 0.6255607720396453,
"grad_norm": 0.2360076606273651,
"learning_rate": 0.00044746184989099977,
"loss": 1.6898614168167114,
"step": 2998
},
{
"epoch": 0.6259780907668232,
"grad_norm": 0.19727914035320282,
"learning_rate": 0.0004469635627530365,
"loss": 1.5027776956558228,
"step": 3000
},
{
"epoch": 0.6263954094940011,
"grad_norm": 0.16130492091178894,
"learning_rate": 0.00044646527561507314,
"loss": 1.6677963733673096,
"step": 3002
},
{
"epoch": 0.626812728221179,
"grad_norm": 0.17396186292171478,
"learning_rate": 0.0004459669884771099,
"loss": 1.7318880558013916,
"step": 3004
},
{
"epoch": 0.6272300469483568,
"grad_norm": 0.16842906177043915,
"learning_rate": 0.00044546870133914667,
"loss": 1.7320445775985718,
"step": 3006
},
{
"epoch": 0.6276473656755347,
"grad_norm": 0.17343546450138092,
"learning_rate": 0.00044497041420118343,
"loss": 1.5647716522216797,
"step": 3008
},
{
"epoch": 0.6280646844027126,
"grad_norm": 0.18864287436008453,
"learning_rate": 0.00044447212706322014,
"loss": 1.6539498567581177,
"step": 3010
},
{
"epoch": 0.6284820031298904,
"grad_norm": 0.18099819123744965,
"learning_rate": 0.0004439738399252569,
"loss": 1.6597203016281128,
"step": 3012
},
{
"epoch": 0.6288993218570683,
"grad_norm": 0.1965656727552414,
"learning_rate": 0.0004434755527872937,
"loss": 1.5832552909851074,
"step": 3014
},
{
"epoch": 0.6293166405842462,
"grad_norm": 0.1879548877477646,
"learning_rate": 0.00044297726564933044,
"loss": 1.5848406553268433,
"step": 3016
},
{
"epoch": 0.6297339593114241,
"grad_norm": 0.16935324668884277,
"learning_rate": 0.00044247897851136715,
"loss": 1.6081706285476685,
"step": 3018
},
{
"epoch": 0.6301512780386019,
"grad_norm": 0.17128311097621918,
"learning_rate": 0.0004419806913734039,
"loss": 1.7153724431991577,
"step": 3020
},
{
"epoch": 0.6305685967657798,
"grad_norm": 0.18790321052074432,
"learning_rate": 0.0004414824042354407,
"loss": 1.7321152687072754,
"step": 3022
},
{
"epoch": 0.6309859154929578,
"grad_norm": 0.17196448147296906,
"learning_rate": 0.00044098411709747744,
"loss": 1.6450070142745972,
"step": 3024
},
{
"epoch": 0.6314032342201357,
"grad_norm": 0.181528240442276,
"learning_rate": 0.00044048582995951415,
"loss": 1.6748415231704712,
"step": 3026
},
{
"epoch": 0.6318205529473135,
"grad_norm": 0.18286748230457306,
"learning_rate": 0.0004399875428215509,
"loss": 1.7617762088775635,
"step": 3028
},
{
"epoch": 0.6322378716744914,
"grad_norm": 0.16749003529548645,
"learning_rate": 0.0004394892556835877,
"loss": 1.6863735914230347,
"step": 3030
},
{
"epoch": 0.6326551904016693,
"grad_norm": 0.22944116592407227,
"learning_rate": 0.00043899096854562445,
"loss": 1.5296680927276611,
"step": 3032
},
{
"epoch": 0.6330725091288472,
"grad_norm": 0.1561669260263443,
"learning_rate": 0.00043849268140766116,
"loss": 1.5262645483016968,
"step": 3034
},
{
"epoch": 0.633489827856025,
"grad_norm": 0.16601932048797607,
"learning_rate": 0.0004379943942696979,
"loss": 1.5379654169082642,
"step": 3036
},
{
"epoch": 0.6339071465832029,
"grad_norm": 0.17442798614501953,
"learning_rate": 0.0004374961071317347,
"loss": 1.7634894847869873,
"step": 3038
},
{
"epoch": 0.6343244653103808,
"grad_norm": 0.1849924623966217,
"learning_rate": 0.00043699781999377146,
"loss": 1.6289749145507812,
"step": 3040
},
{
"epoch": 0.6347417840375587,
"grad_norm": 0.17964428663253784,
"learning_rate": 0.00043649953285580817,
"loss": 1.7756881713867188,
"step": 3042
},
{
"epoch": 0.6351591027647365,
"grad_norm": 0.1700710654258728,
"learning_rate": 0.00043600124571784493,
"loss": 1.654688835144043,
"step": 3044
},
{
"epoch": 0.6355764214919144,
"grad_norm": 0.18866358697414398,
"learning_rate": 0.0004355029585798817,
"loss": 1.6230889558792114,
"step": 3046
},
{
"epoch": 0.6359937402190924,
"grad_norm": 0.23964236676692963,
"learning_rate": 0.0004350046714419184,
"loss": 1.4981260299682617,
"step": 3048
},
{
"epoch": 0.6364110589462703,
"grad_norm": 0.1868370622396469,
"learning_rate": 0.00043450638430395517,
"loss": 1.6155753135681152,
"step": 3050
},
{
"epoch": 0.6368283776734481,
"grad_norm": 0.17351564764976501,
"learning_rate": 0.00043400809716599194,
"loss": 1.6848652362823486,
"step": 3052
},
{
"epoch": 0.637245696400626,
"grad_norm": 0.1712159514427185,
"learning_rate": 0.0004335098100280287,
"loss": 1.689025640487671,
"step": 3054
},
{
"epoch": 0.6376630151278039,
"grad_norm": 0.2635113596916199,
"learning_rate": 0.0004330115228900654,
"loss": 1.5704872608184814,
"step": 3056
},
{
"epoch": 0.6380803338549818,
"grad_norm": 0.20459075272083282,
"learning_rate": 0.0004325132357521022,
"loss": 1.628517985343933,
"step": 3058
},
{
"epoch": 0.6384976525821596,
"grad_norm": 0.19031086564064026,
"learning_rate": 0.00043201494861413894,
"loss": 1.6251946687698364,
"step": 3060
},
{
"epoch": 0.6389149713093375,
"grad_norm": 0.2103298157453537,
"learning_rate": 0.0004315166614761757,
"loss": 1.6923656463623047,
"step": 3062
},
{
"epoch": 0.6393322900365154,
"grad_norm": 0.15932463109493256,
"learning_rate": 0.0004310183743382124,
"loss": 1.6129586696624756,
"step": 3064
},
{
"epoch": 0.6397496087636932,
"grad_norm": 0.17592453956604004,
"learning_rate": 0.0004305200872002492,
"loss": 1.536402702331543,
"step": 3066
},
{
"epoch": 0.6401669274908711,
"grad_norm": 0.16769124567508698,
"learning_rate": 0.00043002180006228595,
"loss": 1.7355247735977173,
"step": 3068
},
{
"epoch": 0.640584246218049,
"grad_norm": 0.17105732858181,
"learning_rate": 0.0004295235129243227,
"loss": 1.634315013885498,
"step": 3070
},
{
"epoch": 0.6410015649452269,
"grad_norm": 0.17621351778507233,
"learning_rate": 0.0004290252257863594,
"loss": 1.6633837223052979,
"step": 3072
},
{
"epoch": 0.6414188836724048,
"grad_norm": 0.1565503627061844,
"learning_rate": 0.0004285269386483962,
"loss": 1.7071964740753174,
"step": 3074
},
{
"epoch": 0.6418362023995827,
"grad_norm": 0.19788403809070587,
"learning_rate": 0.00042802865151043295,
"loss": 1.5612205266952515,
"step": 3076
},
{
"epoch": 0.6422535211267606,
"grad_norm": 0.17223519086837769,
"learning_rate": 0.0004275303643724697,
"loss": 1.6041910648345947,
"step": 3078
},
{
"epoch": 0.6426708398539385,
"grad_norm": 0.17532415688037872,
"learning_rate": 0.00042703207723450643,
"loss": 1.779238224029541,
"step": 3080
},
{
"epoch": 0.6430881585811163,
"grad_norm": 0.21478240191936493,
"learning_rate": 0.0004265337900965432,
"loss": 1.5742828845977783,
"step": 3082
},
{
"epoch": 0.6435054773082942,
"grad_norm": 0.20716965198516846,
"learning_rate": 0.00042603550295857996,
"loss": 1.6765398979187012,
"step": 3084
},
{
"epoch": 0.6439227960354721,
"grad_norm": 0.22852440178394318,
"learning_rate": 0.0004255372158206166,
"loss": 1.6046693325042725,
"step": 3086
},
{
"epoch": 0.64434011476265,
"grad_norm": 0.18507982790470123,
"learning_rate": 0.0004250389286826534,
"loss": 1.5526434183120728,
"step": 3088
},
{
"epoch": 0.6447574334898278,
"grad_norm": 0.17970949411392212,
"learning_rate": 0.0004245406415446901,
"loss": 1.5689854621887207,
"step": 3090
},
{
"epoch": 0.6451747522170057,
"grad_norm": 0.18663153052330017,
"learning_rate": 0.00042404235440672686,
"loss": 1.7312533855438232,
"step": 3092
},
{
"epoch": 0.6455920709441836,
"grad_norm": 0.20818455517292023,
"learning_rate": 0.0004235440672687636,
"loss": 1.6105419397354126,
"step": 3094
},
{
"epoch": 0.6460093896713615,
"grad_norm": 0.15261031687259674,
"learning_rate": 0.0004230457801308004,
"loss": 1.723541021347046,
"step": 3096
},
{
"epoch": 0.6464267083985393,
"grad_norm": 0.19954752922058105,
"learning_rate": 0.0004225474929928371,
"loss": 1.619447946548462,
"step": 3098
},
{
"epoch": 0.6468440271257173,
"grad_norm": 0.20010371506214142,
"learning_rate": 0.00042204920585487386,
"loss": 1.6042195558547974,
"step": 3100
},
{
"epoch": 0.6472613458528952,
"grad_norm": 0.16776025295257568,
"learning_rate": 0.0004215509187169106,
"loss": 1.804923176765442,
"step": 3102
},
{
"epoch": 0.6476786645800731,
"grad_norm": 0.19969533383846283,
"learning_rate": 0.0004210526315789474,
"loss": 1.5725339651107788,
"step": 3104
},
{
"epoch": 0.6480959833072509,
"grad_norm": 0.18492990732192993,
"learning_rate": 0.0004205543444409841,
"loss": 1.6459952592849731,
"step": 3106
},
{
"epoch": 0.6485133020344288,
"grad_norm": 0.19893598556518555,
"learning_rate": 0.00042005605730302087,
"loss": 1.7250248193740845,
"step": 3108
},
{
"epoch": 0.6489306207616067,
"grad_norm": 0.19987525045871735,
"learning_rate": 0.00041955777016505763,
"loss": 1.687385082244873,
"step": 3110
},
{
"epoch": 0.6493479394887846,
"grad_norm": 0.27105215191841125,
"learning_rate": 0.00041905948302709434,
"loss": 1.5936906337738037,
"step": 3112
},
{
"epoch": 0.6497652582159624,
"grad_norm": 0.19652622938156128,
"learning_rate": 0.0004185611958891311,
"loss": 1.7589534521102905,
"step": 3114
},
{
"epoch": 0.6501825769431403,
"grad_norm": 0.1824251115322113,
"learning_rate": 0.00041806290875116787,
"loss": 1.63213050365448,
"step": 3116
},
{
"epoch": 0.6505998956703182,
"grad_norm": 0.18713854253292084,
"learning_rate": 0.00041756462161320464,
"loss": 1.6345391273498535,
"step": 3118
},
{
"epoch": 0.651017214397496,
"grad_norm": 0.17366652190685272,
"learning_rate": 0.00041706633447524135,
"loss": 1.5746169090270996,
"step": 3120
},
{
"epoch": 0.6514345331246739,
"grad_norm": 0.18607227504253387,
"learning_rate": 0.0004165680473372781,
"loss": 1.6285501718521118,
"step": 3122
},
{
"epoch": 0.6518518518518519,
"grad_norm": 0.17764310538768768,
"learning_rate": 0.0004160697601993149,
"loss": 1.5877246856689453,
"step": 3124
},
{
"epoch": 0.6522691705790298,
"grad_norm": 0.18224091827869415,
"learning_rate": 0.00041557147306135164,
"loss": 1.6738301515579224,
"step": 3126
},
{
"epoch": 0.6526864893062077,
"grad_norm": 0.2037447690963745,
"learning_rate": 0.00041507318592338835,
"loss": 1.7492774724960327,
"step": 3128
},
{
"epoch": 0.6531038080333855,
"grad_norm": 0.19194239377975464,
"learning_rate": 0.0004145748987854251,
"loss": 1.7368258237838745,
"step": 3130
},
{
"epoch": 0.6535211267605634,
"grad_norm": 0.16388799250125885,
"learning_rate": 0.0004140766116474619,
"loss": 1.6800915002822876,
"step": 3132
},
{
"epoch": 0.6539384454877413,
"grad_norm": 0.19569306075572968,
"learning_rate": 0.00041357832450949865,
"loss": 1.6806323528289795,
"step": 3134
},
{
"epoch": 0.6543557642149191,
"grad_norm": 0.19107472896575928,
"learning_rate": 0.00041308003737153536,
"loss": 1.5883365869522095,
"step": 3136
},
{
"epoch": 0.654773082942097,
"grad_norm": 0.17638075351715088,
"learning_rate": 0.0004125817502335721,
"loss": 1.636826992034912,
"step": 3138
},
{
"epoch": 0.6551904016692749,
"grad_norm": 0.19983480870723724,
"learning_rate": 0.0004120834630956089,
"loss": 1.6066926717758179,
"step": 3140
},
{
"epoch": 0.6556077203964528,
"grad_norm": 0.1928826868534088,
"learning_rate": 0.00041158517595764565,
"loss": 1.6404352188110352,
"step": 3142
},
{
"epoch": 0.6560250391236306,
"grad_norm": 0.19987136125564575,
"learning_rate": 0.00041108688881968237,
"loss": 1.7169287204742432,
"step": 3144
},
{
"epoch": 0.6564423578508085,
"grad_norm": 0.1764020472764969,
"learning_rate": 0.00041058860168171913,
"loss": 1.6418157815933228,
"step": 3146
},
{
"epoch": 0.6568596765779864,
"grad_norm": 0.1640889197587967,
"learning_rate": 0.0004100903145437559,
"loss": 1.641084909439087,
"step": 3148
},
{
"epoch": 0.6572769953051644,
"grad_norm": 0.1755753457546234,
"learning_rate": 0.00040959202740579266,
"loss": 1.692739725112915,
"step": 3150
},
{
"epoch": 0.6576943140323422,
"grad_norm": 0.17704693973064423,
"learning_rate": 0.00040909374026782937,
"loss": 1.5750787258148193,
"step": 3152
},
{
"epoch": 0.6581116327595201,
"grad_norm": 0.17037749290466309,
"learning_rate": 0.00040859545312986614,
"loss": 1.7550424337387085,
"step": 3154
},
{
"epoch": 0.658528951486698,
"grad_norm": 0.19859790802001953,
"learning_rate": 0.0004080971659919029,
"loss": 1.6859033107757568,
"step": 3156
},
{
"epoch": 0.6589462702138759,
"grad_norm": 0.18174374103546143,
"learning_rate": 0.0004075988788539396,
"loss": 1.6949917078018188,
"step": 3158
},
{
"epoch": 0.6593635889410537,
"grad_norm": 0.18143464624881744,
"learning_rate": 0.0004071005917159764,
"loss": 1.66156005859375,
"step": 3160
},
{
"epoch": 0.6597809076682316,
"grad_norm": 0.18153072893619537,
"learning_rate": 0.00040660230457801314,
"loss": 1.502579689025879,
"step": 3162
},
{
"epoch": 0.6601982263954095,
"grad_norm": 0.19244757294654846,
"learning_rate": 0.0004061040174400499,
"loss": 1.6462054252624512,
"step": 3164
},
{
"epoch": 0.6606155451225874,
"grad_norm": 0.18178164958953857,
"learning_rate": 0.0004056057303020866,
"loss": 1.523628830909729,
"step": 3166
},
{
"epoch": 0.6610328638497652,
"grad_norm": 0.1802728772163391,
"learning_rate": 0.00040510744316412333,
"loss": 1.7697495222091675,
"step": 3168
},
{
"epoch": 0.6614501825769431,
"grad_norm": 0.22055105865001678,
"learning_rate": 0.00040460915602616004,
"loss": 1.6208723783493042,
"step": 3170
},
{
"epoch": 0.661867501304121,
"grad_norm": 0.22543752193450928,
"learning_rate": 0.0004041108688881968,
"loss": 1.6431186199188232,
"step": 3172
},
{
"epoch": 0.662284820031299,
"grad_norm": 0.1864599883556366,
"learning_rate": 0.00040361258175023357,
"loss": 1.689392328262329,
"step": 3174
},
{
"epoch": 0.6627021387584768,
"grad_norm": 0.20197515189647675,
"learning_rate": 0.0004031142946122703,
"loss": 1.6715501546859741,
"step": 3176
},
{
"epoch": 0.6631194574856547,
"grad_norm": 0.18714426457881927,
"learning_rate": 0.00040261600747430704,
"loss": 1.7265963554382324,
"step": 3178
},
{
"epoch": 0.6635367762128326,
"grad_norm": 0.17732010781764984,
"learning_rate": 0.0004021177203363438,
"loss": 1.7067770957946777,
"step": 3180
},
{
"epoch": 0.6639540949400105,
"grad_norm": 0.20761427283287048,
"learning_rate": 0.0004016194331983806,
"loss": 1.4576196670532227,
"step": 3182
},
{
"epoch": 0.6643714136671883,
"grad_norm": 0.16294176876544952,
"learning_rate": 0.0004011211460604173,
"loss": 1.7235853672027588,
"step": 3184
},
{
"epoch": 0.6647887323943662,
"grad_norm": 0.20277062058448792,
"learning_rate": 0.00040062285892245405,
"loss": 1.5236420631408691,
"step": 3186
},
{
"epoch": 0.6652060511215441,
"grad_norm": 0.21320201456546783,
"learning_rate": 0.0004001245717844908,
"loss": 1.6254127025604248,
"step": 3188
},
{
"epoch": 0.665623369848722,
"grad_norm": 0.1975260078907013,
"learning_rate": 0.0003996262846465276,
"loss": 1.716719627380371,
"step": 3190
},
{
"epoch": 0.6660406885758998,
"grad_norm": 0.17935079336166382,
"learning_rate": 0.0003991279975085643,
"loss": 1.7938005924224854,
"step": 3192
},
{
"epoch": 0.6664580073030777,
"grad_norm": 0.19730965793132782,
"learning_rate": 0.00039862971037060106,
"loss": 1.5672612190246582,
"step": 3194
},
{
"epoch": 0.6668753260302556,
"grad_norm": 0.16387251019477844,
"learning_rate": 0.0003981314232326378,
"loss": 1.551000952720642,
"step": 3196
},
{
"epoch": 0.6672926447574334,
"grad_norm": 0.18270333111286163,
"learning_rate": 0.0003976331360946746,
"loss": 1.6228139400482178,
"step": 3198
},
{
"epoch": 0.6677099634846114,
"grad_norm": 0.19708923995494843,
"learning_rate": 0.0003971348489567113,
"loss": 1.6735169887542725,
"step": 3200
},
{
"epoch": 0.6681272822117893,
"grad_norm": 0.20654243230819702,
"learning_rate": 0.00039663656181874806,
"loss": 1.555161476135254,
"step": 3202
},
{
"epoch": 0.6685446009389672,
"grad_norm": 0.17933540046215057,
"learning_rate": 0.0003961382746807848,
"loss": 1.6258504390716553,
"step": 3204
},
{
"epoch": 0.668961919666145,
"grad_norm": 0.24534529447555542,
"learning_rate": 0.0003956399875428216,
"loss": 1.6228227615356445,
"step": 3206
},
{
"epoch": 0.6693792383933229,
"grad_norm": 0.2009853720664978,
"learning_rate": 0.0003951417004048583,
"loss": 1.673019289970398,
"step": 3208
},
{
"epoch": 0.6697965571205008,
"grad_norm": 0.18508794903755188,
"learning_rate": 0.00039464341326689507,
"loss": 1.5413572788238525,
"step": 3210
},
{
"epoch": 0.6702138758476787,
"grad_norm": 0.1912604719400406,
"learning_rate": 0.00039414512612893183,
"loss": 1.593103051185608,
"step": 3212
},
{
"epoch": 0.6706311945748565,
"grad_norm": 0.1695089191198349,
"learning_rate": 0.0003936468389909686,
"loss": 1.6970627307891846,
"step": 3214
},
{
"epoch": 0.6710485133020344,
"grad_norm": 0.19945579767227173,
"learning_rate": 0.0003931485518530053,
"loss": 1.754443883895874,
"step": 3216
},
{
"epoch": 0.6714658320292123,
"grad_norm": 0.20188401639461517,
"learning_rate": 0.00039265026471504207,
"loss": 1.630683422088623,
"step": 3218
},
{
"epoch": 0.6718831507563902,
"grad_norm": 0.18985597789287567,
"learning_rate": 0.00039215197757707884,
"loss": 1.5402309894561768,
"step": 3220
},
{
"epoch": 0.672300469483568,
"grad_norm": 0.17440970242023468,
"learning_rate": 0.00039165369043911555,
"loss": 1.7405648231506348,
"step": 3222
},
{
"epoch": 0.6727177882107459,
"grad_norm": 0.1797809600830078,
"learning_rate": 0.0003911554033011523,
"loss": 1.7436715364456177,
"step": 3224
},
{
"epoch": 0.6731351069379239,
"grad_norm": 0.18043005466461182,
"learning_rate": 0.0003906571161631891,
"loss": 1.6429089307785034,
"step": 3226
},
{
"epoch": 0.6735524256651018,
"grad_norm": 0.19399769604206085,
"learning_rate": 0.00039015882902522584,
"loss": 1.715533971786499,
"step": 3228
},
{
"epoch": 0.6739697443922796,
"grad_norm": 0.21315521001815796,
"learning_rate": 0.00038966054188726255,
"loss": 1.5649092197418213,
"step": 3230
},
{
"epoch": 0.6743870631194575,
"grad_norm": 0.17036151885986328,
"learning_rate": 0.00038916225474929926,
"loss": 1.6871109008789062,
"step": 3232
},
{
"epoch": 0.6748043818466354,
"grad_norm": 0.21142089366912842,
"learning_rate": 0.00038866396761133603,
"loss": 1.5012352466583252,
"step": 3234
},
{
"epoch": 0.6752217005738133,
"grad_norm": 0.1776968389749527,
"learning_rate": 0.0003881656804733728,
"loss": 1.715356707572937,
"step": 3236
},
{
"epoch": 0.6756390193009911,
"grad_norm": 0.167434424161911,
"learning_rate": 0.00038766739333540956,
"loss": 1.761622428894043,
"step": 3238
},
{
"epoch": 0.676056338028169,
"grad_norm": 0.19718925654888153,
"learning_rate": 0.00038716910619744627,
"loss": 1.716399908065796,
"step": 3240
},
{
"epoch": 0.6764736567553469,
"grad_norm": 0.18885953724384308,
"learning_rate": 0.00038667081905948303,
"loss": 1.545933485031128,
"step": 3242
},
{
"epoch": 0.6768909754825247,
"grad_norm": 0.1755514144897461,
"learning_rate": 0.0003861725319215198,
"loss": 1.606783151626587,
"step": 3244
},
{
"epoch": 0.6773082942097026,
"grad_norm": 0.19303053617477417,
"learning_rate": 0.00038567424478355656,
"loss": 1.5963878631591797,
"step": 3246
},
{
"epoch": 0.6777256129368805,
"grad_norm": 0.16974513232707977,
"learning_rate": 0.0003851759576455933,
"loss": 1.658547043800354,
"step": 3248
},
{
"epoch": 0.6781429316640585,
"grad_norm": 0.21708205342292786,
"learning_rate": 0.00038467767050763004,
"loss": 1.4607315063476562,
"step": 3250
},
{
"epoch": 0.6785602503912364,
"grad_norm": 0.21690596640110016,
"learning_rate": 0.0003841793833696668,
"loss": 1.6367452144622803,
"step": 3252
},
{
"epoch": 0.6789775691184142,
"grad_norm": 0.21128971874713898,
"learning_rate": 0.0003836810962317035,
"loss": 1.663558006286621,
"step": 3254
},
{
"epoch": 0.6793948878455921,
"grad_norm": 0.1995653212070465,
"learning_rate": 0.0003831828090937403,
"loss": 1.718706488609314,
"step": 3256
},
{
"epoch": 0.67981220657277,
"grad_norm": 0.17688687145709991,
"learning_rate": 0.00038268452195577705,
"loss": 1.6101213693618774,
"step": 3258
},
{
"epoch": 0.6802295252999478,
"grad_norm": 0.17573122680187225,
"learning_rate": 0.0003821862348178138,
"loss": 1.6461971998214722,
"step": 3260
},
{
"epoch": 0.6806468440271257,
"grad_norm": 0.16721221804618835,
"learning_rate": 0.0003816879476798505,
"loss": 1.688936710357666,
"step": 3262
},
{
"epoch": 0.6810641627543036,
"grad_norm": 0.19553811848163605,
"learning_rate": 0.0003811896605418873,
"loss": 1.7398422956466675,
"step": 3264
},
{
"epoch": 0.6814814814814815,
"grad_norm": 0.17369849979877472,
"learning_rate": 0.00038069137340392405,
"loss": 1.6012886762619019,
"step": 3266
},
{
"epoch": 0.6818988002086593,
"grad_norm": 0.18755407631397247,
"learning_rate": 0.0003801930862659608,
"loss": 1.5039241313934326,
"step": 3268
},
{
"epoch": 0.6823161189358372,
"grad_norm": 0.1913551688194275,
"learning_rate": 0.0003796947991279975,
"loss": 1.6101913452148438,
"step": 3270
},
{
"epoch": 0.6827334376630151,
"grad_norm": 0.19303348660469055,
"learning_rate": 0.00037919651199003424,
"loss": 1.6337401866912842,
"step": 3272
},
{
"epoch": 0.683150756390193,
"grad_norm": 0.21885934472084045,
"learning_rate": 0.000378698224852071,
"loss": 1.6370759010314941,
"step": 3274
},
{
"epoch": 0.6835680751173709,
"grad_norm": 0.16648758947849274,
"learning_rate": 0.00037819993771410777,
"loss": 1.5797863006591797,
"step": 3276
},
{
"epoch": 0.6839853938445488,
"grad_norm": 0.1816018521785736,
"learning_rate": 0.00037770165057614453,
"loss": 1.61228346824646,
"step": 3278
},
{
"epoch": 0.6844027125717267,
"grad_norm": 0.21701984107494354,
"learning_rate": 0.00037720336343818124,
"loss": 1.5739727020263672,
"step": 3280
},
{
"epoch": 0.6848200312989046,
"grad_norm": 0.2241283357143402,
"learning_rate": 0.000376705076300218,
"loss": 1.674843668937683,
"step": 3282
},
{
"epoch": 0.6852373500260824,
"grad_norm": 0.17072917520999908,
"learning_rate": 0.00037620678916225477,
"loss": 1.6810898780822754,
"step": 3284
},
{
"epoch": 0.6856546687532603,
"grad_norm": 0.1833476424217224,
"learning_rate": 0.0003757085020242915,
"loss": 1.663867712020874,
"step": 3286
},
{
"epoch": 0.6860719874804382,
"grad_norm": 0.16291077435016632,
"learning_rate": 0.00037521021488632825,
"loss": 1.611328125,
"step": 3288
},
{
"epoch": 0.6864893062076161,
"grad_norm": 0.21334248781204224,
"learning_rate": 0.000374711927748365,
"loss": 1.6744343042373657,
"step": 3290
},
{
"epoch": 0.6869066249347939,
"grad_norm": 0.18318229913711548,
"learning_rate": 0.0003742136406104018,
"loss": 1.687151312828064,
"step": 3292
},
{
"epoch": 0.6873239436619718,
"grad_norm": 0.2117055207490921,
"learning_rate": 0.0003737153534724385,
"loss": 1.5922801494598389,
"step": 3294
},
{
"epoch": 0.6877412623891497,
"grad_norm": 0.17517757415771484,
"learning_rate": 0.00037321706633447525,
"loss": 1.7506741285324097,
"step": 3296
},
{
"epoch": 0.6881585811163276,
"grad_norm": 0.18518812954425812,
"learning_rate": 0.000372718779196512,
"loss": 1.6906851530075073,
"step": 3298
},
{
"epoch": 0.6885758998435054,
"grad_norm": 0.1808939278125763,
"learning_rate": 0.0003722204920585488,
"loss": 1.5298922061920166,
"step": 3300
},
{
"epoch": 0.6889932185706834,
"grad_norm": 0.1775629073381424,
"learning_rate": 0.0003717222049205855,
"loss": 1.6584842205047607,
"step": 3302
},
{
"epoch": 0.6894105372978613,
"grad_norm": 0.17619644105434418,
"learning_rate": 0.00037122391778262226,
"loss": 1.658012866973877,
"step": 3304
},
{
"epoch": 0.6898278560250392,
"grad_norm": 0.18246251344680786,
"learning_rate": 0.000370725630644659,
"loss": 1.6148688793182373,
"step": 3306
},
{
"epoch": 0.690245174752217,
"grad_norm": 0.17772823572158813,
"learning_rate": 0.0003702273435066958,
"loss": 1.6593983173370361,
"step": 3308
},
{
"epoch": 0.6906624934793949,
"grad_norm": 0.17234240472316742,
"learning_rate": 0.0003697290563687325,
"loss": 1.6848957538604736,
"step": 3310
},
{
"epoch": 0.6910798122065728,
"grad_norm": 0.18917346000671387,
"learning_rate": 0.00036923076923076927,
"loss": 1.5930238962173462,
"step": 3312
},
{
"epoch": 0.6914971309337506,
"grad_norm": 0.17682476341724396,
"learning_rate": 0.000368732482092806,
"loss": 1.7607903480529785,
"step": 3314
},
{
"epoch": 0.6919144496609285,
"grad_norm": 0.17324304580688477,
"learning_rate": 0.00036823419495484274,
"loss": 1.6287239789962769,
"step": 3316
},
{
"epoch": 0.6923317683881064,
"grad_norm": 0.18196740746498108,
"learning_rate": 0.00036773590781687945,
"loss": 1.6950976848602295,
"step": 3318
},
{
"epoch": 0.6927490871152843,
"grad_norm": 0.18560899794101715,
"learning_rate": 0.0003672376206789162,
"loss": 1.6717556715011597,
"step": 3320
},
{
"epoch": 0.6931664058424621,
"grad_norm": 0.1800072193145752,
"learning_rate": 0.000366739333540953,
"loss": 1.6935051679611206,
"step": 3322
},
{
"epoch": 0.69358372456964,
"grad_norm": 0.1855444610118866,
"learning_rate": 0.00036624104640298975,
"loss": 1.6905603408813477,
"step": 3324
},
{
"epoch": 0.694001043296818,
"grad_norm": 0.20639169216156006,
"learning_rate": 0.00036574275926502646,
"loss": 1.5048364400863647,
"step": 3326
},
{
"epoch": 0.6944183620239959,
"grad_norm": 0.17336873710155487,
"learning_rate": 0.0003652444721270632,
"loss": 1.5049575567245483,
"step": 3328
},
{
"epoch": 0.6948356807511737,
"grad_norm": 0.18017591536045074,
"learning_rate": 0.0003647461849891,
"loss": 1.74604332447052,
"step": 3330
},
{
"epoch": 0.6952529994783516,
"grad_norm": 0.17093011736869812,
"learning_rate": 0.00036424789785113675,
"loss": 1.5602046251296997,
"step": 3332
},
{
"epoch": 0.6956703182055295,
"grad_norm": 0.23787932097911835,
"learning_rate": 0.00036374961071317346,
"loss": 1.6649483442306519,
"step": 3334
},
{
"epoch": 0.6960876369327074,
"grad_norm": 0.18329249322414398,
"learning_rate": 0.00036325132357521023,
"loss": 1.697551965713501,
"step": 3336
},
{
"epoch": 0.6965049556598852,
"grad_norm": 0.17169825732707977,
"learning_rate": 0.000362753036437247,
"loss": 1.6272770166397095,
"step": 3338
},
{
"epoch": 0.6969222743870631,
"grad_norm": 0.18326887488365173,
"learning_rate": 0.00036225474929928376,
"loss": 1.6268222332000732,
"step": 3340
},
{
"epoch": 0.697339593114241,
"grad_norm": 0.1736290156841278,
"learning_rate": 0.00036175646216132047,
"loss": 1.592427372932434,
"step": 3342
},
{
"epoch": 0.6977569118414189,
"grad_norm": 0.21996745467185974,
"learning_rate": 0.00036125817502335723,
"loss": 1.6270965337753296,
"step": 3344
},
{
"epoch": 0.6981742305685967,
"grad_norm": 0.1685713231563568,
"learning_rate": 0.000360759887885394,
"loss": 1.633894920349121,
"step": 3346
},
{
"epoch": 0.6985915492957746,
"grad_norm": 0.18318989872932434,
"learning_rate": 0.00036026160074743076,
"loss": 1.701606035232544,
"step": 3348
},
{
"epoch": 0.6990088680229525,
"grad_norm": 0.22740933299064636,
"learning_rate": 0.0003597633136094675,
"loss": 1.625685453414917,
"step": 3350
},
{
"epoch": 0.6994261867501305,
"grad_norm": 0.15949927270412445,
"learning_rate": 0.00035926502647150424,
"loss": 1.7595469951629639,
"step": 3352
},
{
"epoch": 0.6998435054773083,
"grad_norm": 0.1841728687286377,
"learning_rate": 0.000358766739333541,
"loss": 1.6015747785568237,
"step": 3354
},
{
"epoch": 0.7002608242044862,
"grad_norm": 0.20003744959831238,
"learning_rate": 0.0003582684521955777,
"loss": 1.5967538356781006,
"step": 3356
},
{
"epoch": 0.7006781429316641,
"grad_norm": 0.1791495531797409,
"learning_rate": 0.0003577701650576144,
"loss": 1.5644580125808716,
"step": 3358
},
{
"epoch": 0.701095461658842,
"grad_norm": 0.176228329539299,
"learning_rate": 0.0003572718779196512,
"loss": 1.608605980873108,
"step": 3360
},
{
"epoch": 0.7015127803860198,
"grad_norm": 0.220073863863945,
"learning_rate": 0.00035677359078168796,
"loss": 1.5692187547683716,
"step": 3362
},
{
"epoch": 0.7019300991131977,
"grad_norm": 0.20393012464046478,
"learning_rate": 0.0003562753036437247,
"loss": 1.7034521102905273,
"step": 3364
},
{
"epoch": 0.7023474178403756,
"grad_norm": 0.22776682674884796,
"learning_rate": 0.00035577701650576143,
"loss": 1.5576345920562744,
"step": 3366
},
{
"epoch": 0.7027647365675534,
"grad_norm": 0.18106496334075928,
"learning_rate": 0.0003552787293677982,
"loss": 1.726305365562439,
"step": 3368
},
{
"epoch": 0.7031820552947313,
"grad_norm": 0.1843574196100235,
"learning_rate": 0.00035478044222983496,
"loss": 1.6276588439941406,
"step": 3370
},
{
"epoch": 0.7035993740219092,
"grad_norm": 0.207363560795784,
"learning_rate": 0.0003542821550918717,
"loss": 1.6093859672546387,
"step": 3372
},
{
"epoch": 0.7040166927490871,
"grad_norm": 0.16735592484474182,
"learning_rate": 0.00035378386795390844,
"loss": 1.6532926559448242,
"step": 3374
},
{
"epoch": 0.7044340114762649,
"grad_norm": 0.21519218385219574,
"learning_rate": 0.0003532855808159452,
"loss": 1.7327213287353516,
"step": 3376
},
{
"epoch": 0.7048513302034429,
"grad_norm": 0.18955126404762268,
"learning_rate": 0.00035278729367798197,
"loss": 1.5333139896392822,
"step": 3378
},
{
"epoch": 0.7052686489306208,
"grad_norm": 0.1934884488582611,
"learning_rate": 0.00035228900654001873,
"loss": 1.6269757747650146,
"step": 3380
},
{
"epoch": 0.7056859676577987,
"grad_norm": 0.1872010976076126,
"learning_rate": 0.00035179071940205544,
"loss": 1.5592621564865112,
"step": 3382
},
{
"epoch": 0.7061032863849765,
"grad_norm": 0.17345593869686127,
"learning_rate": 0.0003512924322640922,
"loss": 1.501878261566162,
"step": 3384
},
{
"epoch": 0.7065206051121544,
"grad_norm": 0.1699255108833313,
"learning_rate": 0.00035079414512612897,
"loss": 1.7246781587600708,
"step": 3386
},
{
"epoch": 0.7069379238393323,
"grad_norm": 0.2052535116672516,
"learning_rate": 0.00035029585798816574,
"loss": 1.5708210468292236,
"step": 3388
},
{
"epoch": 0.7073552425665102,
"grad_norm": 0.1791800856590271,
"learning_rate": 0.00034979757085020245,
"loss": 1.6467716693878174,
"step": 3390
},
{
"epoch": 0.707772561293688,
"grad_norm": 0.17288070917129517,
"learning_rate": 0.0003492992837122392,
"loss": 1.7028530836105347,
"step": 3392
},
{
"epoch": 0.7081898800208659,
"grad_norm": 0.18910221755504608,
"learning_rate": 0.000348800996574276,
"loss": 1.6224675178527832,
"step": 3394
},
{
"epoch": 0.7086071987480438,
"grad_norm": 0.17708192765712738,
"learning_rate": 0.0003483027094363127,
"loss": 1.7361910343170166,
"step": 3396
},
{
"epoch": 0.7090245174752217,
"grad_norm": 0.18520548939704895,
"learning_rate": 0.0003478044222983494,
"loss": 1.7377831935882568,
"step": 3398
},
{
"epoch": 0.7094418362023995,
"grad_norm": 0.1852477788925171,
"learning_rate": 0.00034730613516038616,
"loss": 1.5747411251068115,
"step": 3400
},
{
"epoch": 0.7098591549295775,
"grad_norm": 0.17900055646896362,
"learning_rate": 0.00034680784802242293,
"loss": 1.6429157257080078,
"step": 3402
},
{
"epoch": 0.7102764736567554,
"grad_norm": 0.1578274518251419,
"learning_rate": 0.0003463095608844597,
"loss": 1.5116872787475586,
"step": 3404
},
{
"epoch": 0.7106937923839333,
"grad_norm": 0.18063633143901825,
"learning_rate": 0.0003458112737464964,
"loss": 1.6570004224777222,
"step": 3406
},
{
"epoch": 0.7111111111111111,
"grad_norm": 0.1643275022506714,
"learning_rate": 0.00034531298660853317,
"loss": 1.6581480503082275,
"step": 3408
},
{
"epoch": 0.711528429838289,
"grad_norm": 0.162019282579422,
"learning_rate": 0.00034481469947056993,
"loss": 1.7755917310714722,
"step": 3410
},
{
"epoch": 0.7119457485654669,
"grad_norm": 0.19159257411956787,
"learning_rate": 0.0003443164123326067,
"loss": 1.6659679412841797,
"step": 3412
},
{
"epoch": 0.7123630672926448,
"grad_norm": 0.23558324575424194,
"learning_rate": 0.0003438181251946434,
"loss": 1.5663334131240845,
"step": 3414
},
{
"epoch": 0.7127803860198226,
"grad_norm": 0.17369040846824646,
"learning_rate": 0.0003433198380566802,
"loss": 1.62667715549469,
"step": 3416
},
{
"epoch": 0.7131977047470005,
"grad_norm": 0.18028756976127625,
"learning_rate": 0.00034282155091871694,
"loss": 1.6441936492919922,
"step": 3418
},
{
"epoch": 0.7136150234741784,
"grad_norm": 0.16337233781814575,
"learning_rate": 0.0003423232637807537,
"loss": 1.7388122081756592,
"step": 3420
},
{
"epoch": 0.7140323422013563,
"grad_norm": 0.19569990038871765,
"learning_rate": 0.0003418249766427904,
"loss": 1.6868973970413208,
"step": 3422
},
{
"epoch": 0.7144496609285341,
"grad_norm": 0.20486384630203247,
"learning_rate": 0.0003413266895048272,
"loss": 1.5231859683990479,
"step": 3424
},
{
"epoch": 0.714866979655712,
"grad_norm": 0.18626397848129272,
"learning_rate": 0.00034082840236686395,
"loss": 1.4338966608047485,
"step": 3426
},
{
"epoch": 0.71528429838289,
"grad_norm": 0.18911179900169373,
"learning_rate": 0.00034033011522890066,
"loss": 1.697441577911377,
"step": 3428
},
{
"epoch": 0.7157016171100679,
"grad_norm": 0.17965462803840637,
"learning_rate": 0.0003398318280909374,
"loss": 1.6648627519607544,
"step": 3430
},
{
"epoch": 0.7161189358372457,
"grad_norm": 0.24312075972557068,
"learning_rate": 0.0003393335409529742,
"loss": 1.5659878253936768,
"step": 3432
},
{
"epoch": 0.7165362545644236,
"grad_norm": 0.2045416235923767,
"learning_rate": 0.00033883525381501095,
"loss": 1.592358946800232,
"step": 3434
},
{
"epoch": 0.7169535732916015,
"grad_norm": 0.17191605269908905,
"learning_rate": 0.00033833696667704766,
"loss": 1.6755856275558472,
"step": 3436
},
{
"epoch": 0.7173708920187793,
"grad_norm": 0.19783489406108856,
"learning_rate": 0.0003378386795390844,
"loss": 1.6227929592132568,
"step": 3438
},
{
"epoch": 0.7177882107459572,
"grad_norm": 0.2214718759059906,
"learning_rate": 0.00033734039240112114,
"loss": 1.6751933097839355,
"step": 3440
},
{
"epoch": 0.7182055294731351,
"grad_norm": 0.2062695324420929,
"learning_rate": 0.0003368421052631579,
"loss": 1.694725751876831,
"step": 3442
},
{
"epoch": 0.718622848200313,
"grad_norm": 0.20154893398284912,
"learning_rate": 0.00033634381812519467,
"loss": 1.6723582744598389,
"step": 3444
},
{
"epoch": 0.7190401669274908,
"grad_norm": 0.2059297114610672,
"learning_rate": 0.0003358455309872314,
"loss": 1.6096183061599731,
"step": 3446
},
{
"epoch": 0.7194574856546687,
"grad_norm": 0.1884816288948059,
"learning_rate": 0.00033534724384926814,
"loss": 1.5833548307418823,
"step": 3448
},
{
"epoch": 0.7198748043818466,
"grad_norm": 0.18161708116531372,
"learning_rate": 0.0003348489567113049,
"loss": 1.706489086151123,
"step": 3450
},
{
"epoch": 0.7202921231090245,
"grad_norm": 0.16823086142539978,
"learning_rate": 0.00033435066957334167,
"loss": 1.7213093042373657,
"step": 3452
},
{
"epoch": 0.7207094418362024,
"grad_norm": 0.19137415289878845,
"learning_rate": 0.0003338523824353784,
"loss": 1.6691241264343262,
"step": 3454
},
{
"epoch": 0.7211267605633803,
"grad_norm": 0.18943490087985992,
"learning_rate": 0.00033335409529741515,
"loss": 1.6035650968551636,
"step": 3456
},
{
"epoch": 0.7215440792905582,
"grad_norm": 0.17243346571922302,
"learning_rate": 0.0003328558081594519,
"loss": 1.7317566871643066,
"step": 3458
},
{
"epoch": 0.7219613980177361,
"grad_norm": 0.18171845376491547,
"learning_rate": 0.0003323575210214886,
"loss": 1.6930899620056152,
"step": 3460
},
{
"epoch": 0.7223787167449139,
"grad_norm": 0.2004927098751068,
"learning_rate": 0.0003318592338835254,
"loss": 1.6099839210510254,
"step": 3462
},
{
"epoch": 0.7227960354720918,
"grad_norm": 0.18558046221733093,
"learning_rate": 0.00033136094674556215,
"loss": 1.6914043426513672,
"step": 3464
},
{
"epoch": 0.7232133541992697,
"grad_norm": 0.17955829203128815,
"learning_rate": 0.0003308626596075989,
"loss": 1.5590879917144775,
"step": 3466
},
{
"epoch": 0.7236306729264476,
"grad_norm": 0.17141635715961456,
"learning_rate": 0.00033036437246963563,
"loss": 1.663912057876587,
"step": 3468
},
{
"epoch": 0.7240479916536254,
"grad_norm": 0.16364917159080505,
"learning_rate": 0.0003298660853316724,
"loss": 1.780855655670166,
"step": 3470
},
{
"epoch": 0.7244653103808033,
"grad_norm": 0.16918881237506866,
"learning_rate": 0.00032936779819370916,
"loss": 1.654486894607544,
"step": 3472
},
{
"epoch": 0.7248826291079812,
"grad_norm": 0.19285793602466583,
"learning_rate": 0.0003288695110557459,
"loss": 1.7523506879806519,
"step": 3474
},
{
"epoch": 0.725299947835159,
"grad_norm": 0.16608674824237823,
"learning_rate": 0.00032837122391778264,
"loss": 1.6734569072723389,
"step": 3476
},
{
"epoch": 0.725717266562337,
"grad_norm": 0.19353562593460083,
"learning_rate": 0.0003278729367798194,
"loss": 1.6555579900741577,
"step": 3478
},
{
"epoch": 0.7261345852895149,
"grad_norm": 0.16799671947956085,
"learning_rate": 0.00032737464964185617,
"loss": 1.6369271278381348,
"step": 3480
},
{
"epoch": 0.7265519040166928,
"grad_norm": 0.2019745111465454,
"learning_rate": 0.0003268763625038929,
"loss": 1.674414038658142,
"step": 3482
},
{
"epoch": 0.7269692227438707,
"grad_norm": 0.2025943249464035,
"learning_rate": 0.00032637807536592964,
"loss": 1.6427620649337769,
"step": 3484
},
{
"epoch": 0.7273865414710485,
"grad_norm": 0.2056577205657959,
"learning_rate": 0.00032587978822796635,
"loss": 1.5915733575820923,
"step": 3486
},
{
"epoch": 0.7278038601982264,
"grad_norm": 0.18525126576423645,
"learning_rate": 0.0003253815010900031,
"loss": 1.7508444786071777,
"step": 3488
},
{
"epoch": 0.7282211789254043,
"grad_norm": 0.1688654124736786,
"learning_rate": 0.0003248832139520399,
"loss": 1.5843013525009155,
"step": 3490
},
{
"epoch": 0.7286384976525822,
"grad_norm": 0.20414622128009796,
"learning_rate": 0.0003243849268140766,
"loss": 1.477527379989624,
"step": 3492
},
{
"epoch": 0.72905581637976,
"grad_norm": 0.19403406977653503,
"learning_rate": 0.00032388663967611336,
"loss": 1.672422170639038,
"step": 3494
},
{
"epoch": 0.7294731351069379,
"grad_norm": 0.19974422454833984,
"learning_rate": 0.0003233883525381501,
"loss": 1.6990084648132324,
"step": 3496
},
{
"epoch": 0.7298904538341158,
"grad_norm": 0.16629716753959656,
"learning_rate": 0.0003228900654001869,
"loss": 1.6100983619689941,
"step": 3498
},
{
"epoch": 0.7303077725612936,
"grad_norm": 0.1910194605588913,
"learning_rate": 0.0003223917782622236,
"loss": 1.686997652053833,
"step": 3500
},
{
"epoch": 0.7307250912884715,
"grad_norm": 0.19526292383670807,
"learning_rate": 0.00032189349112426036,
"loss": 1.658736228942871,
"step": 3502
},
{
"epoch": 0.7311424100156495,
"grad_norm": 0.20036041736602783,
"learning_rate": 0.00032139520398629713,
"loss": 1.5146386623382568,
"step": 3504
},
{
"epoch": 0.7315597287428274,
"grad_norm": 0.1627195030450821,
"learning_rate": 0.0003208969168483339,
"loss": 1.5579910278320312,
"step": 3506
},
{
"epoch": 0.7319770474700052,
"grad_norm": 0.17499464750289917,
"learning_rate": 0.0003203986297103706,
"loss": 1.6159790754318237,
"step": 3508
},
{
"epoch": 0.7323943661971831,
"grad_norm": 0.1754591464996338,
"learning_rate": 0.00031990034257240737,
"loss": 1.7475316524505615,
"step": 3510
},
{
"epoch": 0.732811684924361,
"grad_norm": 0.1966923177242279,
"learning_rate": 0.00031940205543444413,
"loss": 1.7484667301177979,
"step": 3512
},
{
"epoch": 0.7332290036515389,
"grad_norm": 0.18660548329353333,
"learning_rate": 0.0003189037682964809,
"loss": 1.645972490310669,
"step": 3514
},
{
"epoch": 0.7336463223787167,
"grad_norm": 0.19023196399211884,
"learning_rate": 0.0003184054811585176,
"loss": 1.6627051830291748,
"step": 3516
},
{
"epoch": 0.7340636411058946,
"grad_norm": 0.19592605531215668,
"learning_rate": 0.0003179071940205544,
"loss": 1.5341144800186157,
"step": 3518
},
{
"epoch": 0.7344809598330725,
"grad_norm": 0.16704915463924408,
"learning_rate": 0.00031740890688259114,
"loss": 1.6278053522109985,
"step": 3520
},
{
"epoch": 0.7348982785602504,
"grad_norm": 0.20632880926132202,
"learning_rate": 0.0003169106197446279,
"loss": 1.6913785934448242,
"step": 3522
},
{
"epoch": 0.7353155972874282,
"grad_norm": 0.21235673129558563,
"learning_rate": 0.00031641233260666456,
"loss": 1.6417800188064575,
"step": 3524
},
{
"epoch": 0.7357329160146061,
"grad_norm": 0.1678508073091507,
"learning_rate": 0.0003159140454687013,
"loss": 1.7229536771774292,
"step": 3526
},
{
"epoch": 0.7361502347417841,
"grad_norm": 0.1596875935792923,
"learning_rate": 0.0003154157583307381,
"loss": 1.6837574243545532,
"step": 3528
},
{
"epoch": 0.736567553468962,
"grad_norm": 0.19168215990066528,
"learning_rate": 0.00031491747119277486,
"loss": 1.5995575189590454,
"step": 3530
},
{
"epoch": 0.7369848721961398,
"grad_norm": 0.17190667986869812,
"learning_rate": 0.00031441918405481157,
"loss": 1.631792426109314,
"step": 3532
},
{
"epoch": 0.7374021909233177,
"grad_norm": 0.18951287865638733,
"learning_rate": 0.00031392089691684833,
"loss": 1.6929583549499512,
"step": 3534
},
{
"epoch": 0.7378195096504956,
"grad_norm": 0.23520375788211823,
"learning_rate": 0.0003134226097788851,
"loss": 1.6094257831573486,
"step": 3536
},
{
"epoch": 0.7382368283776735,
"grad_norm": 0.17844054102897644,
"learning_rate": 0.00031292432264092186,
"loss": 1.7344905138015747,
"step": 3538
},
{
"epoch": 0.7386541471048513,
"grad_norm": 0.2153760939836502,
"learning_rate": 0.00031242603550295857,
"loss": 1.4577974081039429,
"step": 3540
},
{
"epoch": 0.7390714658320292,
"grad_norm": 0.20724165439605713,
"learning_rate": 0.00031192774836499534,
"loss": 1.6709504127502441,
"step": 3542
},
{
"epoch": 0.7394887845592071,
"grad_norm": 0.18786104023456573,
"learning_rate": 0.0003114294612270321,
"loss": 1.6881275177001953,
"step": 3544
},
{
"epoch": 0.739906103286385,
"grad_norm": 0.18174445629119873,
"learning_rate": 0.00031093117408906887,
"loss": 1.6563602685928345,
"step": 3546
},
{
"epoch": 0.7403234220135628,
"grad_norm": 0.17231345176696777,
"learning_rate": 0.0003104328869511056,
"loss": 1.6132090091705322,
"step": 3548
},
{
"epoch": 0.7407407407407407,
"grad_norm": 0.1631447970867157,
"learning_rate": 0.00030993459981314234,
"loss": 1.7104825973510742,
"step": 3550
},
{
"epoch": 0.7411580594679186,
"grad_norm": 0.1798943430185318,
"learning_rate": 0.0003094363126751791,
"loss": 1.6832358837127686,
"step": 3552
},
{
"epoch": 0.7415753781950966,
"grad_norm": 0.18840263783931732,
"learning_rate": 0.00030893802553721587,
"loss": 1.6568399667739868,
"step": 3554
},
{
"epoch": 0.7419926969222744,
"grad_norm": 0.1632978916168213,
"learning_rate": 0.0003084397383992526,
"loss": 1.6325010061264038,
"step": 3556
},
{
"epoch": 0.7424100156494523,
"grad_norm": 0.1779744178056717,
"learning_rate": 0.00030794145126128935,
"loss": 1.7412676811218262,
"step": 3558
},
{
"epoch": 0.7428273343766302,
"grad_norm": 0.1876668930053711,
"learning_rate": 0.0003074431641233261,
"loss": 1.5542502403259277,
"step": 3560
},
{
"epoch": 0.743244653103808,
"grad_norm": 0.19083812832832336,
"learning_rate": 0.0003069448769853629,
"loss": 1.6276262998580933,
"step": 3562
},
{
"epoch": 0.7436619718309859,
"grad_norm": 0.1701236069202423,
"learning_rate": 0.00030644658984739953,
"loss": 1.727341651916504,
"step": 3564
},
{
"epoch": 0.7440792905581638,
"grad_norm": 0.22421248257160187,
"learning_rate": 0.0003059483027094363,
"loss": 1.5933177471160889,
"step": 3566
},
{
"epoch": 0.7444966092853417,
"grad_norm": 0.19811683893203735,
"learning_rate": 0.00030545001557147306,
"loss": 1.631873607635498,
"step": 3568
},
{
"epoch": 0.7449139280125195,
"grad_norm": 0.18560124933719635,
"learning_rate": 0.00030495172843350983,
"loss": 1.6574822664260864,
"step": 3570
},
{
"epoch": 0.7453312467396974,
"grad_norm": 0.17930057644844055,
"learning_rate": 0.00030445344129554654,
"loss": 1.4892685413360596,
"step": 3572
},
{
"epoch": 0.7457485654668753,
"grad_norm": 0.19537664949893951,
"learning_rate": 0.0003039551541575833,
"loss": 1.6422487497329712,
"step": 3574
},
{
"epoch": 0.7461658841940532,
"grad_norm": 0.1611482948064804,
"learning_rate": 0.00030345686701962007,
"loss": 1.5966274738311768,
"step": 3576
},
{
"epoch": 0.746583202921231,
"grad_norm": 0.19023479521274567,
"learning_rate": 0.00030295857988165683,
"loss": 1.514742374420166,
"step": 3578
},
{
"epoch": 0.747000521648409,
"grad_norm": 0.17993195354938507,
"learning_rate": 0.00030246029274369355,
"loss": 1.7231776714324951,
"step": 3580
},
{
"epoch": 0.7474178403755869,
"grad_norm": 0.20421218872070312,
"learning_rate": 0.0003019620056057303,
"loss": 1.7592166662216187,
"step": 3582
},
{
"epoch": 0.7478351591027648,
"grad_norm": 0.21028248965740204,
"learning_rate": 0.0003014637184677671,
"loss": 1.666697382926941,
"step": 3584
},
{
"epoch": 0.7482524778299426,
"grad_norm": 0.23155921697616577,
"learning_rate": 0.00030096543132980384,
"loss": 1.6484336853027344,
"step": 3586
},
{
"epoch": 0.7486697965571205,
"grad_norm": 0.20363196730613708,
"learning_rate": 0.00030046714419184055,
"loss": 1.651363492012024,
"step": 3588
},
{
"epoch": 0.7490871152842984,
"grad_norm": 0.25108081102371216,
"learning_rate": 0.0002999688570538773,
"loss": 1.5527845621109009,
"step": 3590
},
{
"epoch": 0.7495044340114763,
"grad_norm": 0.18479235470294952,
"learning_rate": 0.0002994705699159141,
"loss": 1.6575050354003906,
"step": 3592
},
{
"epoch": 0.7499217527386541,
"grad_norm": 0.16402573883533478,
"learning_rate": 0.00029897228277795085,
"loss": 1.6732909679412842,
"step": 3594
},
{
"epoch": 0.750339071465832,
"grad_norm": 0.17851699888706207,
"learning_rate": 0.00029847399563998756,
"loss": 1.6448041200637817,
"step": 3596
},
{
"epoch": 0.7507563901930099,
"grad_norm": 0.2052585482597351,
"learning_rate": 0.0002979757085020243,
"loss": 1.7649446725845337,
"step": 3598
},
{
"epoch": 0.7511737089201878,
"grad_norm": 0.19212980568408966,
"learning_rate": 0.0002974774213640611,
"loss": 1.4695961475372314,
"step": 3600
},
{
"epoch": 0.7515910276473656,
"grad_norm": 0.17334549129009247,
"learning_rate": 0.0002969791342260978,
"loss": 1.7598848342895508,
"step": 3602
},
{
"epoch": 0.7520083463745436,
"grad_norm": 0.17423833906650543,
"learning_rate": 0.00029648084708813456,
"loss": 1.758447527885437,
"step": 3604
},
{
"epoch": 0.7524256651017215,
"grad_norm": 0.16897940635681152,
"learning_rate": 0.00029598255995017127,
"loss": 1.610061526298523,
"step": 3606
},
{
"epoch": 0.7528429838288994,
"grad_norm": 0.19908249378204346,
"learning_rate": 0.00029548427281220804,
"loss": 1.5874207019805908,
"step": 3608
},
{
"epoch": 0.7532603025560772,
"grad_norm": 0.1795583814382553,
"learning_rate": 0.0002949859856742448,
"loss": 1.6818596124649048,
"step": 3610
},
{
"epoch": 0.7536776212832551,
"grad_norm": 0.1832699477672577,
"learning_rate": 0.0002944876985362815,
"loss": 1.6959115266799927,
"step": 3612
},
{
"epoch": 0.754094940010433,
"grad_norm": 0.17787104845046997,
"learning_rate": 0.0002939894113983183,
"loss": 1.7967486381530762,
"step": 3614
},
{
"epoch": 0.7545122587376109,
"grad_norm": 0.17039570212364197,
"learning_rate": 0.00029349112426035504,
"loss": 1.6880176067352295,
"step": 3616
},
{
"epoch": 0.7549295774647887,
"grad_norm": 0.17902342975139618,
"learning_rate": 0.0002929928371223918,
"loss": 1.733896017074585,
"step": 3618
},
{
"epoch": 0.7553468961919666,
"grad_norm": 0.19079959392547607,
"learning_rate": 0.0002924945499844285,
"loss": 1.5747251510620117,
"step": 3620
},
{
"epoch": 0.7557642149191445,
"grad_norm": 0.18268540501594543,
"learning_rate": 0.0002919962628464653,
"loss": 1.6724902391433716,
"step": 3622
},
{
"epoch": 0.7561815336463223,
"grad_norm": 0.19185388088226318,
"learning_rate": 0.00029149797570850205,
"loss": 1.610269546508789,
"step": 3624
},
{
"epoch": 0.7565988523735002,
"grad_norm": 0.1950681060552597,
"learning_rate": 0.0002909996885705388,
"loss": 1.6017721891403198,
"step": 3626
},
{
"epoch": 0.7570161711006781,
"grad_norm": 0.1675286740064621,
"learning_rate": 0.0002905014014325755,
"loss": 1.5421817302703857,
"step": 3628
},
{
"epoch": 0.7574334898278561,
"grad_norm": 0.17911790311336517,
"learning_rate": 0.0002900031142946123,
"loss": 1.7843542098999023,
"step": 3630
},
{
"epoch": 0.757850808555034,
"grad_norm": 0.20425382256507874,
"learning_rate": 0.00028950482715664905,
"loss": 1.7099859714508057,
"step": 3632
},
{
"epoch": 0.7582681272822118,
"grad_norm": 0.17926475405693054,
"learning_rate": 0.00028900654001868576,
"loss": 1.6763795614242554,
"step": 3634
},
{
"epoch": 0.7586854460093897,
"grad_norm": 0.17786246538162231,
"learning_rate": 0.00028850825288072253,
"loss": 1.584932565689087,
"step": 3636
},
{
"epoch": 0.7591027647365676,
"grad_norm": 0.17833611369132996,
"learning_rate": 0.0002880099657427593,
"loss": 1.5606003999710083,
"step": 3638
},
{
"epoch": 0.7595200834637454,
"grad_norm": 0.1954340934753418,
"learning_rate": 0.00028751167860479606,
"loss": 1.7637163400650024,
"step": 3640
},
{
"epoch": 0.7599374021909233,
"grad_norm": 0.2349027395248413,
"learning_rate": 0.00028701339146683277,
"loss": 1.7626914978027344,
"step": 3642
},
{
"epoch": 0.7603547209181012,
"grad_norm": 0.1968459039926529,
"learning_rate": 0.00028651510432886954,
"loss": 1.5119953155517578,
"step": 3644
},
{
"epoch": 0.7607720396452791,
"grad_norm": 0.2195347398519516,
"learning_rate": 0.0002860168171909063,
"loss": 1.4143831729888916,
"step": 3646
},
{
"epoch": 0.7611893583724569,
"grad_norm": 0.1913500279188156,
"learning_rate": 0.000285518530052943,
"loss": 1.7684820890426636,
"step": 3648
},
{
"epoch": 0.7616066770996348,
"grad_norm": 0.1927497684955597,
"learning_rate": 0.0002850202429149798,
"loss": 1.6670191287994385,
"step": 3650
},
{
"epoch": 0.7620239958268127,
"grad_norm": 0.216668963432312,
"learning_rate": 0.0002845219557770165,
"loss": 1.626707911491394,
"step": 3652
},
{
"epoch": 0.7624413145539906,
"grad_norm": 0.1823483556509018,
"learning_rate": 0.00028402366863905325,
"loss": 1.5955305099487305,
"step": 3654
},
{
"epoch": 0.7628586332811685,
"grad_norm": 0.19510966539382935,
"learning_rate": 0.00028352538150109,
"loss": 1.6058166027069092,
"step": 3656
},
{
"epoch": 0.7632759520083464,
"grad_norm": 0.18889285624027252,
"learning_rate": 0.0002830270943631268,
"loss": 1.780603051185608,
"step": 3658
},
{
"epoch": 0.7636932707355243,
"grad_norm": 0.20094002783298492,
"learning_rate": 0.0002825288072251635,
"loss": 1.6105626821517944,
"step": 3660
},
{
"epoch": 0.7641105894627022,
"grad_norm": 0.18301644921302795,
"learning_rate": 0.00028203052008720026,
"loss": 1.7110977172851562,
"step": 3662
},
{
"epoch": 0.76452790818988,
"grad_norm": 0.17168305814266205,
"learning_rate": 0.000281532232949237,
"loss": 1.5489096641540527,
"step": 3664
},
{
"epoch": 0.7649452269170579,
"grad_norm": 0.19458550214767456,
"learning_rate": 0.00028103394581127373,
"loss": 1.6385643482208252,
"step": 3666
},
{
"epoch": 0.7653625456442358,
"grad_norm": 0.2028268724679947,
"learning_rate": 0.0002805356586733105,
"loss": 1.686463713645935,
"step": 3668
},
{
"epoch": 0.7657798643714137,
"grad_norm": 0.1836620718240738,
"learning_rate": 0.00028003737153534726,
"loss": 1.5797909498214722,
"step": 3670
},
{
"epoch": 0.7661971830985915,
"grad_norm": 0.21502596139907837,
"learning_rate": 0.00027953908439738403,
"loss": 1.589252233505249,
"step": 3672
},
{
"epoch": 0.7666145018257694,
"grad_norm": 0.1878790408372879,
"learning_rate": 0.00027904079725942074,
"loss": 1.561044692993164,
"step": 3674
},
{
"epoch": 0.7670318205529473,
"grad_norm": 0.21280565857887268,
"learning_rate": 0.0002785425101214575,
"loss": 1.4992973804473877,
"step": 3676
},
{
"epoch": 0.7674491392801251,
"grad_norm": 0.17974157631397247,
"learning_rate": 0.00027804422298349427,
"loss": 1.6194812059402466,
"step": 3678
},
{
"epoch": 0.7678664580073031,
"grad_norm": 0.18471801280975342,
"learning_rate": 0.00027754593584553103,
"loss": 1.6645277738571167,
"step": 3680
},
{
"epoch": 0.768283776734481,
"grad_norm": 0.2178938388824463,
"learning_rate": 0.00027704764870756774,
"loss": 1.4975954294204712,
"step": 3682
},
{
"epoch": 0.7687010954616589,
"grad_norm": 0.1764337122440338,
"learning_rate": 0.0002765493615696045,
"loss": 1.6571986675262451,
"step": 3684
},
{
"epoch": 0.7691184141888368,
"grad_norm": 0.18027573823928833,
"learning_rate": 0.0002760510744316413,
"loss": 1.6677141189575195,
"step": 3686
},
{
"epoch": 0.7695357329160146,
"grad_norm": 0.19011811912059784,
"learning_rate": 0.00027555278729367804,
"loss": 1.532283902168274,
"step": 3688
},
{
"epoch": 0.7699530516431925,
"grad_norm": 0.17116892337799072,
"learning_rate": 0.00027505450015571475,
"loss": 1.801459550857544,
"step": 3690
},
{
"epoch": 0.7703703703703704,
"grad_norm": 0.1825629025697708,
"learning_rate": 0.00027455621301775146,
"loss": 1.5290942192077637,
"step": 3692
},
{
"epoch": 0.7707876890975482,
"grad_norm": 0.16886523365974426,
"learning_rate": 0.0002740579258797882,
"loss": 1.6664427518844604,
"step": 3694
},
{
"epoch": 0.7712050078247261,
"grad_norm": 0.2470306009054184,
"learning_rate": 0.000273559638741825,
"loss": 1.59676194190979,
"step": 3696
},
{
"epoch": 0.771622326551904,
"grad_norm": 0.22003215551376343,
"learning_rate": 0.0002730613516038617,
"loss": 1.8252915143966675,
"step": 3698
},
{
"epoch": 0.7720396452790819,
"grad_norm": 0.1497248411178589,
"learning_rate": 0.00027256306446589847,
"loss": 1.63771653175354,
"step": 3700
},
{
"epoch": 0.7724569640062597,
"grad_norm": 0.16998142004013062,
"learning_rate": 0.00027206477732793523,
"loss": 1.7147274017333984,
"step": 3702
},
{
"epoch": 0.7728742827334376,
"grad_norm": 0.2003924697637558,
"learning_rate": 0.000271566490189972,
"loss": 1.5882560014724731,
"step": 3704
},
{
"epoch": 0.7732916014606156,
"grad_norm": 0.23485775291919708,
"learning_rate": 0.0002710682030520087,
"loss": 1.4673707485198975,
"step": 3706
},
{
"epoch": 0.7737089201877935,
"grad_norm": 0.18744097650051117,
"learning_rate": 0.00027056991591404547,
"loss": 1.6588497161865234,
"step": 3708
},
{
"epoch": 0.7741262389149713,
"grad_norm": 0.1956653594970703,
"learning_rate": 0.00027007162877608224,
"loss": 1.624337911605835,
"step": 3710
},
{
"epoch": 0.7745435576421492,
"grad_norm": 0.1610584706068039,
"learning_rate": 0.000269573341638119,
"loss": 1.5354138612747192,
"step": 3712
},
{
"epoch": 0.7749608763693271,
"grad_norm": 0.20504875481128693,
"learning_rate": 0.0002690750545001557,
"loss": 1.706824779510498,
"step": 3714
},
{
"epoch": 0.775378195096505,
"grad_norm": 0.21478965878486633,
"learning_rate": 0.0002685767673621925,
"loss": 1.512431025505066,
"step": 3716
},
{
"epoch": 0.7757955138236828,
"grad_norm": 0.1747409999370575,
"learning_rate": 0.00026807848022422924,
"loss": 1.6951833963394165,
"step": 3718
},
{
"epoch": 0.7762128325508607,
"grad_norm": 0.19729028642177582,
"learning_rate": 0.000267580193086266,
"loss": 1.4797358512878418,
"step": 3720
},
{
"epoch": 0.7766301512780386,
"grad_norm": 0.1793026328086853,
"learning_rate": 0.0002670819059483027,
"loss": 1.699170708656311,
"step": 3722
},
{
"epoch": 0.7770474700052165,
"grad_norm": 0.19655486941337585,
"learning_rate": 0.0002665836188103395,
"loss": 1.6706569194793701,
"step": 3724
},
{
"epoch": 0.7774647887323943,
"grad_norm": 0.1673012524843216,
"learning_rate": 0.00026608533167237625,
"loss": 1.5427839756011963,
"step": 3726
},
{
"epoch": 0.7778821074595722,
"grad_norm": 0.17791549861431122,
"learning_rate": 0.000265587044534413,
"loss": 1.6230547428131104,
"step": 3728
},
{
"epoch": 0.7782994261867501,
"grad_norm": 0.19570524990558624,
"learning_rate": 0.0002650887573964497,
"loss": 1.4675053358078003,
"step": 3730
},
{
"epoch": 0.7787167449139281,
"grad_norm": 0.1805286705493927,
"learning_rate": 0.00026459047025848643,
"loss": 1.5337735414505005,
"step": 3732
},
{
"epoch": 0.7791340636411059,
"grad_norm": 0.20290841162204742,
"learning_rate": 0.0002640921831205232,
"loss": 1.637925386428833,
"step": 3734
},
{
"epoch": 0.7795513823682838,
"grad_norm": 0.19543811678886414,
"learning_rate": 0.00026359389598255996,
"loss": 1.6386189460754395,
"step": 3736
},
{
"epoch": 0.7799687010954617,
"grad_norm": 0.19317297637462616,
"learning_rate": 0.0002630956088445967,
"loss": 1.5798938274383545,
"step": 3738
},
{
"epoch": 0.7803860198226396,
"grad_norm": 0.1734863519668579,
"learning_rate": 0.00026259732170663344,
"loss": 1.6295630931854248,
"step": 3740
},
{
"epoch": 0.7808033385498174,
"grad_norm": 0.19331972301006317,
"learning_rate": 0.0002620990345686702,
"loss": 1.8345309495925903,
"step": 3742
},
{
"epoch": 0.7812206572769953,
"grad_norm": 0.20073075592517853,
"learning_rate": 0.00026160074743070697,
"loss": 1.593034029006958,
"step": 3744
},
{
"epoch": 0.7816379760041732,
"grad_norm": 0.18429666757583618,
"learning_rate": 0.0002611024602927437,
"loss": 1.6535524129867554,
"step": 3746
},
{
"epoch": 0.782055294731351,
"grad_norm": 0.1882796436548233,
"learning_rate": 0.00026060417315478045,
"loss": 1.7575433254241943,
"step": 3748
},
{
"epoch": 0.7824726134585289,
"grad_norm": 0.19296231865882874,
"learning_rate": 0.0002601058860168172,
"loss": 1.5442777872085571,
"step": 3750
},
{
"epoch": 0.7828899321857068,
"grad_norm": 0.18059085309505463,
"learning_rate": 0.000259607598878854,
"loss": 1.6075869798660278,
"step": 3752
},
{
"epoch": 0.7833072509128847,
"grad_norm": 0.1694050133228302,
"learning_rate": 0.0002591093117408907,
"loss": 1.725911021232605,
"step": 3754
},
{
"epoch": 0.7837245696400627,
"grad_norm": 0.1803259253501892,
"learning_rate": 0.00025861102460292745,
"loss": 1.6701686382293701,
"step": 3756
},
{
"epoch": 0.7841418883672405,
"grad_norm": 0.19327299296855927,
"learning_rate": 0.0002581127374649642,
"loss": 1.383103609085083,
"step": 3758
},
{
"epoch": 0.7845592070944184,
"grad_norm": 0.1650385558605194,
"learning_rate": 0.000257614450327001,
"loss": 1.525803565979004,
"step": 3760
},
{
"epoch": 0.7849765258215963,
"grad_norm": 0.18854647874832153,
"learning_rate": 0.0002571161631890377,
"loss": 1.6652625799179077,
"step": 3762
},
{
"epoch": 0.7853938445487741,
"grad_norm": 0.20049698650836945,
"learning_rate": 0.00025661787605107446,
"loss": 1.543817162513733,
"step": 3764
},
{
"epoch": 0.785811163275952,
"grad_norm": 0.16368618607521057,
"learning_rate": 0.0002561195889131112,
"loss": 1.6001052856445312,
"step": 3766
},
{
"epoch": 0.7862284820031299,
"grad_norm": 0.18294206261634827,
"learning_rate": 0.00025562130177514793,
"loss": 1.633233666419983,
"step": 3768
},
{
"epoch": 0.7866458007303078,
"grad_norm": 0.19872161746025085,
"learning_rate": 0.0002551230146371847,
"loss": 1.701698899269104,
"step": 3770
},
{
"epoch": 0.7870631194574856,
"grad_norm": 0.20103883743286133,
"learning_rate": 0.00025462472749922146,
"loss": 1.7317423820495605,
"step": 3772
},
{
"epoch": 0.7874804381846635,
"grad_norm": 0.18021374940872192,
"learning_rate": 0.00025412644036125817,
"loss": 1.6171764135360718,
"step": 3774
},
{
"epoch": 0.7878977569118414,
"grad_norm": 0.20081184804439545,
"learning_rate": 0.00025362815322329494,
"loss": 1.6483221054077148,
"step": 3776
},
{
"epoch": 0.7883150756390193,
"grad_norm": 0.19757476449012756,
"learning_rate": 0.00025312986608533165,
"loss": 1.701400637626648,
"step": 3778
},
{
"epoch": 0.7887323943661971,
"grad_norm": 0.18681901693344116,
"learning_rate": 0.0002526315789473684,
"loss": 1.61826491355896,
"step": 3780
},
{
"epoch": 0.7891497130933751,
"grad_norm": 0.17457766830921173,
"learning_rate": 0.0002521332918094052,
"loss": 1.7168126106262207,
"step": 3782
},
{
"epoch": 0.789567031820553,
"grad_norm": 0.20330972969532013,
"learning_rate": 0.00025163500467144194,
"loss": 1.6683218479156494,
"step": 3784
},
{
"epoch": 0.7899843505477309,
"grad_norm": 0.18788500130176544,
"learning_rate": 0.00025113671753347865,
"loss": 1.5661441087722778,
"step": 3786
},
{
"epoch": 0.7904016692749087,
"grad_norm": 0.16008377075195312,
"learning_rate": 0.0002506384303955154,
"loss": 1.6120243072509766,
"step": 3788
},
{
"epoch": 0.7908189880020866,
"grad_norm": 0.1842314451932907,
"learning_rate": 0.0002501401432575522,
"loss": 1.6992385387420654,
"step": 3790
},
{
"epoch": 0.7912363067292645,
"grad_norm": 0.1744423359632492,
"learning_rate": 0.00024964185611958895,
"loss": 1.6774944067001343,
"step": 3792
},
{
"epoch": 0.7916536254564424,
"grad_norm": 0.20514267683029175,
"learning_rate": 0.00024914356898162566,
"loss": 1.631840705871582,
"step": 3794
},
{
"epoch": 0.7920709441836202,
"grad_norm": 0.16552075743675232,
"learning_rate": 0.0002486452818436624,
"loss": 1.558586597442627,
"step": 3796
},
{
"epoch": 0.7924882629107981,
"grad_norm": 0.17318940162658691,
"learning_rate": 0.0002481469947056992,
"loss": 1.6639153957366943,
"step": 3798
},
{
"epoch": 0.792905581637976,
"grad_norm": 0.16856642067432404,
"learning_rate": 0.0002476487075677359,
"loss": 1.6824519634246826,
"step": 3800
},
{
"epoch": 0.7933229003651538,
"grad_norm": 0.1910776048898697,
"learning_rate": 0.00024715042042977266,
"loss": 1.548015832901001,
"step": 3802
},
{
"epoch": 0.7937402190923317,
"grad_norm": 0.18034255504608154,
"learning_rate": 0.00024665213329180943,
"loss": 1.6619542837142944,
"step": 3804
},
{
"epoch": 0.7941575378195096,
"grad_norm": 0.2144138365983963,
"learning_rate": 0.0002461538461538462,
"loss": 1.540140151977539,
"step": 3806
},
{
"epoch": 0.7945748565466876,
"grad_norm": 0.1667701005935669,
"learning_rate": 0.0002456555590158829,
"loss": 1.7628003358840942,
"step": 3808
},
{
"epoch": 0.7949921752738655,
"grad_norm": 0.17015314102172852,
"learning_rate": 0.00024515727187791967,
"loss": 1.5367592573165894,
"step": 3810
},
{
"epoch": 0.7954094940010433,
"grad_norm": 0.18134693801403046,
"learning_rate": 0.00024465898473995644,
"loss": 1.5571942329406738,
"step": 3812
},
{
"epoch": 0.7958268127282212,
"grad_norm": 0.1884811818599701,
"learning_rate": 0.0002441606976019932,
"loss": 1.6780465841293335,
"step": 3814
},
{
"epoch": 0.7962441314553991,
"grad_norm": 0.20955635607242584,
"learning_rate": 0.00024366241046402988,
"loss": 1.5253673791885376,
"step": 3816
},
{
"epoch": 0.796661450182577,
"grad_norm": 0.1925605833530426,
"learning_rate": 0.00024316412332606665,
"loss": 1.6923273801803589,
"step": 3818
},
{
"epoch": 0.7970787689097548,
"grad_norm": 0.17855115234851837,
"learning_rate": 0.0002426658361881034,
"loss": 1.6492364406585693,
"step": 3820
},
{
"epoch": 0.7974960876369327,
"grad_norm": 0.1645374745130539,
"learning_rate": 0.00024216754905014015,
"loss": 1.8101212978363037,
"step": 3822
},
{
"epoch": 0.7979134063641106,
"grad_norm": 0.19214420020580292,
"learning_rate": 0.0002416692619121769,
"loss": 1.6729837656021118,
"step": 3824
},
{
"epoch": 0.7983307250912884,
"grad_norm": 0.1854393035173416,
"learning_rate": 0.00024117097477421365,
"loss": 1.7643306255340576,
"step": 3826
},
{
"epoch": 0.7987480438184663,
"grad_norm": 0.16908498108386993,
"learning_rate": 0.0002406726876362504,
"loss": 1.594125509262085,
"step": 3828
},
{
"epoch": 0.7991653625456442,
"grad_norm": 0.1607009917497635,
"learning_rate": 0.00024017440049828716,
"loss": 1.6117491722106934,
"step": 3830
},
{
"epoch": 0.7995826812728222,
"grad_norm": 0.19979916512966156,
"learning_rate": 0.0002396761133603239,
"loss": 1.6051199436187744,
"step": 3832
},
{
"epoch": 0.8,
"grad_norm": 0.1697085052728653,
"learning_rate": 0.00023917782622236063,
"loss": 1.6251349449157715,
"step": 3834
},
{
"epoch": 0.8004173187271779,
"grad_norm": 0.182024747133255,
"learning_rate": 0.0002386795390843974,
"loss": 1.6238594055175781,
"step": 3836
},
{
"epoch": 0.8004173187271779,
"eval_loss": 1.6281813383102417,
"eval_runtime": 25.5965,
"eval_samples_per_second": 60.555,
"eval_steps_per_second": 7.579,
"step": 3836
},
{
"epoch": 0.8008346374543558,
"grad_norm": 0.16110607981681824,
"learning_rate": 0.00023818125194643414,
"loss": 1.5853067636489868,
"step": 3838
},
{
"epoch": 0.8012519561815337,
"grad_norm": 0.1834481954574585,
"learning_rate": 0.0002376829648084709,
"loss": 1.4189400672912598,
"step": 3840
},
{
"epoch": 0.8016692749087115,
"grad_norm": 0.21303535997867584,
"learning_rate": 0.00023718467767050764,
"loss": 1.576297640800476,
"step": 3842
},
{
"epoch": 0.8020865936358894,
"grad_norm": 0.20752784609794617,
"learning_rate": 0.0002366863905325444,
"loss": 1.6360690593719482,
"step": 3844
},
{
"epoch": 0.8025039123630673,
"grad_norm": 0.16906622052192688,
"learning_rate": 0.00023618810339458114,
"loss": 1.7308926582336426,
"step": 3846
},
{
"epoch": 0.8029212310902452,
"grad_norm": 0.16772274672985077,
"learning_rate": 0.0002356898162566179,
"loss": 1.6283552646636963,
"step": 3848
},
{
"epoch": 0.803338549817423,
"grad_norm": 0.20605945587158203,
"learning_rate": 0.00023519152911865464,
"loss": 1.7920221090316772,
"step": 3850
},
{
"epoch": 0.8037558685446009,
"grad_norm": 0.19974687695503235,
"learning_rate": 0.0002346932419806914,
"loss": 1.6810439825057983,
"step": 3852
},
{
"epoch": 0.8041731872717788,
"grad_norm": 0.1850401908159256,
"learning_rate": 0.00023419495484272815,
"loss": 1.764756679534912,
"step": 3854
},
{
"epoch": 0.8045905059989567,
"grad_norm": 0.20353806018829346,
"learning_rate": 0.00023369666770476486,
"loss": 1.595638394355774,
"step": 3856
},
{
"epoch": 0.8050078247261346,
"grad_norm": 0.1756061315536499,
"learning_rate": 0.00023319838056680162,
"loss": 1.5819897651672363,
"step": 3858
},
{
"epoch": 0.8054251434533125,
"grad_norm": 0.1912943422794342,
"learning_rate": 0.00023270009342883836,
"loss": 1.6187857389450073,
"step": 3860
},
{
"epoch": 0.8058424621804904,
"grad_norm": 0.19449324905872345,
"learning_rate": 0.00023220180629087513,
"loss": 1.7449872493743896,
"step": 3862
},
{
"epoch": 0.8062597809076683,
"grad_norm": 0.1860467940568924,
"learning_rate": 0.00023170351915291186,
"loss": 1.6441023349761963,
"step": 3864
},
{
"epoch": 0.8066770996348461,
"grad_norm": 0.19042663276195526,
"learning_rate": 0.0002312052320149486,
"loss": 1.6760553121566772,
"step": 3866
},
{
"epoch": 0.807094418362024,
"grad_norm": 0.18120992183685303,
"learning_rate": 0.00023070694487698537,
"loss": 1.6071486473083496,
"step": 3868
},
{
"epoch": 0.8075117370892019,
"grad_norm": 0.17611190676689148,
"learning_rate": 0.0002302086577390221,
"loss": 1.721142292022705,
"step": 3870
},
{
"epoch": 0.8079290558163797,
"grad_norm": 0.17643216252326965,
"learning_rate": 0.00022971037060105887,
"loss": 1.6106164455413818,
"step": 3872
},
{
"epoch": 0.8083463745435576,
"grad_norm": 0.18115536868572235,
"learning_rate": 0.0002292120834630956,
"loss": 1.6975512504577637,
"step": 3874
},
{
"epoch": 0.8087636932707355,
"grad_norm": 0.1759224385023117,
"learning_rate": 0.00022871379632513237,
"loss": 1.686223030090332,
"step": 3876
},
{
"epoch": 0.8091810119979134,
"grad_norm": 0.18049465119838715,
"learning_rate": 0.0002282155091871691,
"loss": 1.5944403409957886,
"step": 3878
},
{
"epoch": 0.8095983307250912,
"grad_norm": 0.17533127963542938,
"learning_rate": 0.00022771722204920587,
"loss": 1.5372076034545898,
"step": 3880
},
{
"epoch": 0.8100156494522691,
"grad_norm": 0.2384546995162964,
"learning_rate": 0.0002272189349112426,
"loss": 1.6710647344589233,
"step": 3882
},
{
"epoch": 0.8104329681794471,
"grad_norm": 0.15719540417194366,
"learning_rate": 0.00022672064777327938,
"loss": 1.618177056312561,
"step": 3884
},
{
"epoch": 0.810850286906625,
"grad_norm": 0.1845632940530777,
"learning_rate": 0.00022622236063531611,
"loss": 1.7172913551330566,
"step": 3886
},
{
"epoch": 0.8112676056338028,
"grad_norm": 0.23309579491615295,
"learning_rate": 0.00022572407349735288,
"loss": 1.574898362159729,
"step": 3888
},
{
"epoch": 0.8116849243609807,
"grad_norm": 0.17386038601398468,
"learning_rate": 0.00022522578635938962,
"loss": 1.7423713207244873,
"step": 3890
},
{
"epoch": 0.8121022430881586,
"grad_norm": 0.19860124588012695,
"learning_rate": 0.00022472749922142638,
"loss": 1.6700553894042969,
"step": 3892
},
{
"epoch": 0.8125195618153365,
"grad_norm": 0.15150749683380127,
"learning_rate": 0.00022422921208346312,
"loss": 1.6848392486572266,
"step": 3894
},
{
"epoch": 0.8129368805425143,
"grad_norm": 0.1860383003950119,
"learning_rate": 0.00022373092494549989,
"loss": 1.468409538269043,
"step": 3896
},
{
"epoch": 0.8133541992696922,
"grad_norm": 0.16297782957553864,
"learning_rate": 0.00022323263780753657,
"loss": 1.722561240196228,
"step": 3898
},
{
"epoch": 0.8137715179968701,
"grad_norm": 0.15188246965408325,
"learning_rate": 0.00022273435066957333,
"loss": 1.6671578884124756,
"step": 3900
},
{
"epoch": 0.814188836724048,
"grad_norm": 0.1724192351102829,
"learning_rate": 0.00022223606353161007,
"loss": 1.5962269306182861,
"step": 3902
},
{
"epoch": 0.8146061554512258,
"grad_norm": 0.18045011162757874,
"learning_rate": 0.00022173777639364684,
"loss": 1.5897287130355835,
"step": 3904
},
{
"epoch": 0.8150234741784037,
"grad_norm": 0.1753378063440323,
"learning_rate": 0.00022123948925568357,
"loss": 1.6983929872512817,
"step": 3906
},
{
"epoch": 0.8154407929055817,
"grad_norm": 0.1929623782634735,
"learning_rate": 0.00022074120211772034,
"loss": 1.6955832242965698,
"step": 3908
},
{
"epoch": 0.8158581116327596,
"grad_norm": 0.1988271176815033,
"learning_rate": 0.00022024291497975708,
"loss": 1.6405982971191406,
"step": 3910
},
{
"epoch": 0.8162754303599374,
"grad_norm": 0.1987425833940506,
"learning_rate": 0.00021974462784179384,
"loss": 1.6183756589889526,
"step": 3912
},
{
"epoch": 0.8166927490871153,
"grad_norm": 0.2016177624464035,
"learning_rate": 0.00021924634070383058,
"loss": 1.5909420251846313,
"step": 3914
},
{
"epoch": 0.8171100678142932,
"grad_norm": 0.1823321431875229,
"learning_rate": 0.00021874805356586735,
"loss": 1.6325428485870361,
"step": 3916
},
{
"epoch": 0.8175273865414711,
"grad_norm": 0.1803046017885208,
"learning_rate": 0.00021824976642790408,
"loss": 1.6903064250946045,
"step": 3918
},
{
"epoch": 0.8179447052686489,
"grad_norm": 0.15474292635917664,
"learning_rate": 0.00021775147928994085,
"loss": 1.6624375581741333,
"step": 3920
},
{
"epoch": 0.8183620239958268,
"grad_norm": 0.18099120259284973,
"learning_rate": 0.00021725319215197759,
"loss": 1.6347200870513916,
"step": 3922
},
{
"epoch": 0.8187793427230047,
"grad_norm": 0.15678411722183228,
"learning_rate": 0.00021675490501401435,
"loss": 1.616406798362732,
"step": 3924
},
{
"epoch": 0.8191966614501826,
"grad_norm": 0.21508070826530457,
"learning_rate": 0.0002162566178760511,
"loss": 1.5474953651428223,
"step": 3926
},
{
"epoch": 0.8196139801773604,
"grad_norm": 0.19049592316150665,
"learning_rate": 0.00021575833073808785,
"loss": 1.576249599456787,
"step": 3928
},
{
"epoch": 0.8200312989045383,
"grad_norm": 0.1880476325750351,
"learning_rate": 0.0002152600436001246,
"loss": 1.64571213722229,
"step": 3930
},
{
"epoch": 0.8204486176317162,
"grad_norm": 0.2017848640680313,
"learning_rate": 0.00021476175646216136,
"loss": 1.5606279373168945,
"step": 3932
},
{
"epoch": 0.8208659363588942,
"grad_norm": 0.1754579246044159,
"learning_rate": 0.0002142634693241981,
"loss": 1.6150054931640625,
"step": 3934
},
{
"epoch": 0.821283255086072,
"grad_norm": 0.17297819256782532,
"learning_rate": 0.00021376518218623486,
"loss": 1.6455872058868408,
"step": 3936
},
{
"epoch": 0.8217005738132499,
"grad_norm": 0.17700959742069244,
"learning_rate": 0.0002132668950482716,
"loss": 1.6134843826293945,
"step": 3938
},
{
"epoch": 0.8221178925404278,
"grad_norm": 0.19243700802326202,
"learning_rate": 0.0002127686079103083,
"loss": 1.6845083236694336,
"step": 3940
},
{
"epoch": 0.8225352112676056,
"grad_norm": 0.1604607105255127,
"learning_rate": 0.00021227032077234505,
"loss": 1.5880975723266602,
"step": 3942
},
{
"epoch": 0.8229525299947835,
"grad_norm": 0.1929977387189865,
"learning_rate": 0.0002117720336343818,
"loss": 1.4770071506500244,
"step": 3944
},
{
"epoch": 0.8233698487219614,
"grad_norm": 0.1966904252767563,
"learning_rate": 0.00021127374649641855,
"loss": 1.651863932609558,
"step": 3946
},
{
"epoch": 0.8237871674491393,
"grad_norm": 0.17531289160251617,
"learning_rate": 0.0002107754593584553,
"loss": 1.5507947206497192,
"step": 3948
},
{
"epoch": 0.8242044861763171,
"grad_norm": 0.18605005741119385,
"learning_rate": 0.00021027717222049205,
"loss": 1.7106924057006836,
"step": 3950
},
{
"epoch": 0.824621804903495,
"grad_norm": 0.21515291929244995,
"learning_rate": 0.00020977888508252882,
"loss": 1.7130653858184814,
"step": 3952
},
{
"epoch": 0.8250391236306729,
"grad_norm": 0.20646436512470245,
"learning_rate": 0.00020928059794456555,
"loss": 1.656790018081665,
"step": 3954
},
{
"epoch": 0.8254564423578508,
"grad_norm": 0.15887944400310516,
"learning_rate": 0.00020878231080660232,
"loss": 1.6606398820877075,
"step": 3956
},
{
"epoch": 0.8258737610850287,
"grad_norm": 0.17161159217357635,
"learning_rate": 0.00020828402366863906,
"loss": 1.6557297706604004,
"step": 3958
},
{
"epoch": 0.8262910798122066,
"grad_norm": 0.19186924397945404,
"learning_rate": 0.00020778573653067582,
"loss": 1.6486374139785767,
"step": 3960
},
{
"epoch": 0.8267083985393845,
"grad_norm": 0.16262342035770416,
"learning_rate": 0.00020728744939271256,
"loss": 1.688956379890442,
"step": 3962
},
{
"epoch": 0.8271257172665624,
"grad_norm": 0.15369129180908203,
"learning_rate": 0.00020678916225474932,
"loss": 1.68675696849823,
"step": 3964
},
{
"epoch": 0.8275430359937402,
"grad_norm": 0.18608498573303223,
"learning_rate": 0.00020629087511678606,
"loss": 1.6111623048782349,
"step": 3966
},
{
"epoch": 0.8279603547209181,
"grad_norm": 0.19858677685260773,
"learning_rate": 0.00020579258797882283,
"loss": 1.5222647190093994,
"step": 3968
},
{
"epoch": 0.828377673448096,
"grad_norm": 0.18191881477832794,
"learning_rate": 0.00020529430084085957,
"loss": 1.5683807134628296,
"step": 3970
},
{
"epoch": 0.8287949921752739,
"grad_norm": 0.16898715496063232,
"learning_rate": 0.00020479601370289633,
"loss": 1.5900148153305054,
"step": 3972
},
{
"epoch": 0.8292123109024517,
"grad_norm": 0.2008662223815918,
"learning_rate": 0.00020429772656493307,
"loss": 1.6750458478927612,
"step": 3974
},
{
"epoch": 0.8296296296296296,
"grad_norm": 0.1838344931602478,
"learning_rate": 0.0002037994394269698,
"loss": 1.597100019454956,
"step": 3976
},
{
"epoch": 0.8300469483568075,
"grad_norm": 0.19462299346923828,
"learning_rate": 0.00020330115228900657,
"loss": 1.6593374013900757,
"step": 3978
},
{
"epoch": 0.8304642670839854,
"grad_norm": 0.17652329802513123,
"learning_rate": 0.0002028028651510433,
"loss": 1.6299662590026855,
"step": 3980
},
{
"epoch": 0.8308815858111632,
"grad_norm": 0.18837544322013855,
"learning_rate": 0.00020230457801308002,
"loss": 1.6133947372436523,
"step": 3982
},
{
"epoch": 0.8312989045383412,
"grad_norm": 0.21258091926574707,
"learning_rate": 0.00020180629087511678,
"loss": 1.5445491075515747,
"step": 3984
},
{
"epoch": 0.8317162232655191,
"grad_norm": 0.17253796756267548,
"learning_rate": 0.00020130800373715352,
"loss": 1.5718294382095337,
"step": 3986
},
{
"epoch": 0.832133541992697,
"grad_norm": 0.1892494112253189,
"learning_rate": 0.0002008097165991903,
"loss": 1.5831724405288696,
"step": 3988
},
{
"epoch": 0.8325508607198748,
"grad_norm": 0.18101245164871216,
"learning_rate": 0.00020031142946122702,
"loss": 1.6219923496246338,
"step": 3990
},
{
"epoch": 0.8329681794470527,
"grad_norm": 0.2411927729845047,
"learning_rate": 0.0001998131423232638,
"loss": 1.6351206302642822,
"step": 3992
},
{
"epoch": 0.8333854981742306,
"grad_norm": 0.16940459609031677,
"learning_rate": 0.00019931485518530053,
"loss": 1.620373010635376,
"step": 3994
},
{
"epoch": 0.8338028169014085,
"grad_norm": 0.1804088056087494,
"learning_rate": 0.0001988165680473373,
"loss": 1.6183083057403564,
"step": 3996
},
{
"epoch": 0.8342201356285863,
"grad_norm": 0.17482876777648926,
"learning_rate": 0.00019831828090937403,
"loss": 1.73622727394104,
"step": 3998
},
{
"epoch": 0.8346374543557642,
"grad_norm": 0.17766039073467255,
"learning_rate": 0.0001978199937714108,
"loss": 1.7408409118652344,
"step": 4000
},
{
"epoch": 0.8350547730829421,
"grad_norm": 0.17571066319942474,
"learning_rate": 0.00019732170663344753,
"loss": 1.6120573282241821,
"step": 4002
},
{
"epoch": 0.8354720918101199,
"grad_norm": 0.18707862496376038,
"learning_rate": 0.0001968234194954843,
"loss": 1.7047288417816162,
"step": 4004
},
{
"epoch": 0.8358894105372978,
"grad_norm": 0.1814654916524887,
"learning_rate": 0.00019632513235752104,
"loss": 1.6476895809173584,
"step": 4006
},
{
"epoch": 0.8363067292644757,
"grad_norm": 0.2454279214143753,
"learning_rate": 0.00019582684521955777,
"loss": 1.6494747400283813,
"step": 4008
},
{
"epoch": 0.8367240479916537,
"grad_norm": 0.18015429377555847,
"learning_rate": 0.00019532855808159454,
"loss": 1.6449940204620361,
"step": 4010
},
{
"epoch": 0.8371413667188315,
"grad_norm": 0.19813057780265808,
"learning_rate": 0.00019483027094363128,
"loss": 1.461126685142517,
"step": 4012
},
{
"epoch": 0.8375586854460094,
"grad_norm": 0.1895369291305542,
"learning_rate": 0.00019433198380566801,
"loss": 1.6304774284362793,
"step": 4014
},
{
"epoch": 0.8379760041731873,
"grad_norm": 0.22598986327648163,
"learning_rate": 0.00019383369666770478,
"loss": 1.5876799821853638,
"step": 4016
},
{
"epoch": 0.8383933229003652,
"grad_norm": 0.16353179514408112,
"learning_rate": 0.00019333540952974152,
"loss": 1.670882225036621,
"step": 4018
},
{
"epoch": 0.838810641627543,
"grad_norm": 0.16480985283851624,
"learning_rate": 0.00019283712239177828,
"loss": 1.6846975088119507,
"step": 4020
},
{
"epoch": 0.8392279603547209,
"grad_norm": 0.1873014122247696,
"learning_rate": 0.00019233883525381502,
"loss": 1.690040946006775,
"step": 4022
},
{
"epoch": 0.8396452790818988,
"grad_norm": 0.17535239458084106,
"learning_rate": 0.00019184054811585176,
"loss": 1.6181294918060303,
"step": 4024
},
{
"epoch": 0.8400625978090767,
"grad_norm": 0.23253878951072693,
"learning_rate": 0.00019134226097788852,
"loss": 1.5434880256652832,
"step": 4026
},
{
"epoch": 0.8404799165362545,
"grad_norm": 0.16973750293254852,
"learning_rate": 0.00019084397383992526,
"loss": 1.6462557315826416,
"step": 4028
},
{
"epoch": 0.8408972352634324,
"grad_norm": 0.19195866584777832,
"learning_rate": 0.00019034568670196203,
"loss": 1.6567882299423218,
"step": 4030
},
{
"epoch": 0.8413145539906103,
"grad_norm": 0.19295237958431244,
"learning_rate": 0.00018984739956399876,
"loss": 1.635372281074524,
"step": 4032
},
{
"epoch": 0.8417318727177883,
"grad_norm": 0.195025235414505,
"learning_rate": 0.0001893491124260355,
"loss": 1.7076184749603271,
"step": 4034
},
{
"epoch": 0.8421491914449661,
"grad_norm": 0.22713610529899597,
"learning_rate": 0.00018885082528807227,
"loss": 1.5234016180038452,
"step": 4036
},
{
"epoch": 0.842566510172144,
"grad_norm": 0.1975746899843216,
"learning_rate": 0.000188352538150109,
"loss": 1.711593747138977,
"step": 4038
},
{
"epoch": 0.8429838288993219,
"grad_norm": 0.19030791521072388,
"learning_rate": 0.00018785425101214574,
"loss": 1.6802881956100464,
"step": 4040
},
{
"epoch": 0.8434011476264998,
"grad_norm": 0.2094624787569046,
"learning_rate": 0.0001873559638741825,
"loss": 1.6112802028656006,
"step": 4042
},
{
"epoch": 0.8438184663536776,
"grad_norm": 0.18288825452327728,
"learning_rate": 0.00018685767673621924,
"loss": 1.5856302976608276,
"step": 4044
},
{
"epoch": 0.8442357850808555,
"grad_norm": 0.21274502575397491,
"learning_rate": 0.000186359389598256,
"loss": 1.674270749092102,
"step": 4046
},
{
"epoch": 0.8446531038080334,
"grad_norm": 0.19420906901359558,
"learning_rate": 0.00018586110246029275,
"loss": 1.6582943201065063,
"step": 4048
},
{
"epoch": 0.8450704225352113,
"grad_norm": 0.18451790511608124,
"learning_rate": 0.0001853628153223295,
"loss": 1.6256777048110962,
"step": 4050
},
{
"epoch": 0.8454877412623891,
"grad_norm": 0.16896241903305054,
"learning_rate": 0.00018486452818436625,
"loss": 1.7003672122955322,
"step": 4052
},
{
"epoch": 0.845905059989567,
"grad_norm": 0.18207846581935883,
"learning_rate": 0.000184366241046403,
"loss": 1.6286587715148926,
"step": 4054
},
{
"epoch": 0.8463223787167449,
"grad_norm": 0.22023826837539673,
"learning_rate": 0.00018386795390843973,
"loss": 1.5539515018463135,
"step": 4056
},
{
"epoch": 0.8467396974439227,
"grad_norm": 0.16868507862091064,
"learning_rate": 0.0001833696667704765,
"loss": 1.6270174980163574,
"step": 4058
},
{
"epoch": 0.8471570161711007,
"grad_norm": 0.18117032945156097,
"learning_rate": 0.00018287137963251323,
"loss": 1.5214279890060425,
"step": 4060
},
{
"epoch": 0.8475743348982786,
"grad_norm": 0.15181727707386017,
"learning_rate": 0.00018237309249455,
"loss": 1.6494383811950684,
"step": 4062
},
{
"epoch": 0.8479916536254565,
"grad_norm": 0.23103384673595428,
"learning_rate": 0.00018187480535658673,
"loss": 1.5475603342056274,
"step": 4064
},
{
"epoch": 0.8484089723526343,
"grad_norm": 0.1830928772687912,
"learning_rate": 0.0001813765182186235,
"loss": 1.7238051891326904,
"step": 4066
},
{
"epoch": 0.8488262910798122,
"grad_norm": 0.1808210015296936,
"learning_rate": 0.00018087823108066023,
"loss": 1.6027417182922363,
"step": 4068
},
{
"epoch": 0.8492436098069901,
"grad_norm": 0.17436878383159637,
"learning_rate": 0.000180379943942697,
"loss": 1.5270894765853882,
"step": 4070
},
{
"epoch": 0.849660928534168,
"grad_norm": 0.2353087067604065,
"learning_rate": 0.00017988165680473374,
"loss": 1.660226821899414,
"step": 4072
},
{
"epoch": 0.8500782472613458,
"grad_norm": 0.18193411827087402,
"learning_rate": 0.0001793833696667705,
"loss": 1.6704041957855225,
"step": 4074
},
{
"epoch": 0.8504955659885237,
"grad_norm": 0.157639741897583,
"learning_rate": 0.0001788850825288072,
"loss": 1.721926212310791,
"step": 4076
},
{
"epoch": 0.8509128847157016,
"grad_norm": 0.1731785237789154,
"learning_rate": 0.00017838679539084398,
"loss": 1.6603143215179443,
"step": 4078
},
{
"epoch": 0.8513302034428795,
"grad_norm": 0.16457518935203552,
"learning_rate": 0.00017788850825288072,
"loss": 1.5349699258804321,
"step": 4080
},
{
"epoch": 0.8517475221700573,
"grad_norm": 0.18919974565505981,
"learning_rate": 0.00017739022111491748,
"loss": 1.5536229610443115,
"step": 4082
},
{
"epoch": 0.8521648408972352,
"grad_norm": 0.18905764818191528,
"learning_rate": 0.00017689193397695422,
"loss": 1.5567257404327393,
"step": 4084
},
{
"epoch": 0.8525821596244132,
"grad_norm": 0.18444348871707916,
"learning_rate": 0.00017639364683899098,
"loss": 1.528328776359558,
"step": 4086
},
{
"epoch": 0.8529994783515911,
"grad_norm": 0.18102359771728516,
"learning_rate": 0.00017589535970102772,
"loss": 1.6205244064331055,
"step": 4088
},
{
"epoch": 0.8534167970787689,
"grad_norm": 0.19649460911750793,
"learning_rate": 0.00017539707256306449,
"loss": 1.5504631996154785,
"step": 4090
},
{
"epoch": 0.8538341158059468,
"grad_norm": 0.17354848980903625,
"learning_rate": 0.00017489878542510122,
"loss": 1.652526617050171,
"step": 4092
},
{
"epoch": 0.8542514345331247,
"grad_norm": 0.19733025133609772,
"learning_rate": 0.000174400498287138,
"loss": 1.679112434387207,
"step": 4094
},
{
"epoch": 0.8546687532603026,
"grad_norm": 0.20596131682395935,
"learning_rate": 0.0001739022111491747,
"loss": 1.4270410537719727,
"step": 4096
},
{
"epoch": 0.8550860719874804,
"grad_norm": 0.18710580468177795,
"learning_rate": 0.00017340392401121146,
"loss": 1.5961430072784424,
"step": 4098
},
{
"epoch": 0.8555033907146583,
"grad_norm": 0.18958649039268494,
"learning_rate": 0.0001729056368732482,
"loss": 1.593994379043579,
"step": 4100
},
{
"epoch": 0.8559207094418362,
"grad_norm": 0.1896360069513321,
"learning_rate": 0.00017240734973528497,
"loss": 1.6488350629806519,
"step": 4102
},
{
"epoch": 0.856338028169014,
"grad_norm": 0.19518868625164032,
"learning_rate": 0.0001719090625973217,
"loss": 1.6326905488967896,
"step": 4104
},
{
"epoch": 0.8567553468961919,
"grad_norm": 0.18605078756809235,
"learning_rate": 0.00017141077545935847,
"loss": 1.6978832483291626,
"step": 4106
},
{
"epoch": 0.8571726656233698,
"grad_norm": 0.19510576128959656,
"learning_rate": 0.0001709124883213952,
"loss": 1.722651720046997,
"step": 4108
},
{
"epoch": 0.8575899843505478,
"grad_norm": 0.16520658135414124,
"learning_rate": 0.00017041420118343197,
"loss": 1.6256784200668335,
"step": 4110
},
{
"epoch": 0.8580073030777257,
"grad_norm": 0.1725723147392273,
"learning_rate": 0.0001699159140454687,
"loss": 1.6118927001953125,
"step": 4112
},
{
"epoch": 0.8584246218049035,
"grad_norm": 0.17283189296722412,
"learning_rate": 0.00016941762690750548,
"loss": 1.6664702892303467,
"step": 4114
},
{
"epoch": 0.8588419405320814,
"grad_norm": 0.1931280642747879,
"learning_rate": 0.0001689193397695422,
"loss": 1.6569066047668457,
"step": 4116
},
{
"epoch": 0.8592592592592593,
"grad_norm": 0.1914769858121872,
"learning_rate": 0.00016842105263157895,
"loss": 1.6330338716506958,
"step": 4118
},
{
"epoch": 0.8596765779864372,
"grad_norm": 0.19188779592514038,
"learning_rate": 0.0001679227654936157,
"loss": 1.6686639785766602,
"step": 4120
},
{
"epoch": 0.860093896713615,
"grad_norm": 0.20317953824996948,
"learning_rate": 0.00016742447835565245,
"loss": 1.661203384399414,
"step": 4122
},
{
"epoch": 0.8605112154407929,
"grad_norm": 0.1870126873254776,
"learning_rate": 0.0001669261912176892,
"loss": 1.6325947046279907,
"step": 4124
},
{
"epoch": 0.8609285341679708,
"grad_norm": 0.1979297399520874,
"learning_rate": 0.00016642790407972596,
"loss": 1.6142420768737793,
"step": 4126
},
{
"epoch": 0.8613458528951486,
"grad_norm": 0.17889894545078278,
"learning_rate": 0.0001659296169417627,
"loss": 1.7480411529541016,
"step": 4128
},
{
"epoch": 0.8617631716223265,
"grad_norm": 0.1875065267086029,
"learning_rate": 0.00016543132980379946,
"loss": 1.6685492992401123,
"step": 4130
},
{
"epoch": 0.8621804903495044,
"grad_norm": 0.1852869987487793,
"learning_rate": 0.0001649330426658362,
"loss": 1.5089213848114014,
"step": 4132
},
{
"epoch": 0.8625978090766823,
"grad_norm": 0.21147358417510986,
"learning_rate": 0.00016443475552787296,
"loss": 1.6361830234527588,
"step": 4134
},
{
"epoch": 0.8630151278038602,
"grad_norm": 0.19058911502361298,
"learning_rate": 0.0001639364683899097,
"loss": 1.6463992595672607,
"step": 4136
},
{
"epoch": 0.8634324465310381,
"grad_norm": 0.2115446925163269,
"learning_rate": 0.00016343818125194644,
"loss": 1.5269619226455688,
"step": 4138
},
{
"epoch": 0.863849765258216,
"grad_norm": 0.15730158984661102,
"learning_rate": 0.00016293989411398318,
"loss": 1.6363976001739502,
"step": 4140
},
{
"epoch": 0.8642670839853939,
"grad_norm": 0.16966697573661804,
"learning_rate": 0.00016244160697601994,
"loss": 1.6779712438583374,
"step": 4142
},
{
"epoch": 0.8646844027125717,
"grad_norm": 0.20783443748950958,
"learning_rate": 0.00016194331983805668,
"loss": 1.5522081851959229,
"step": 4144
},
{
"epoch": 0.8651017214397496,
"grad_norm": 0.19765686988830566,
"learning_rate": 0.00016144503270009344,
"loss": 1.671380639076233,
"step": 4146
},
{
"epoch": 0.8655190401669275,
"grad_norm": 0.20410001277923584,
"learning_rate": 0.00016094674556213018,
"loss": 1.6107127666473389,
"step": 4148
},
{
"epoch": 0.8659363588941054,
"grad_norm": 0.190677672624588,
"learning_rate": 0.00016044845842416695,
"loss": 1.5826854705810547,
"step": 4150
},
{
"epoch": 0.8663536776212832,
"grad_norm": 0.2105415016412735,
"learning_rate": 0.00015995017128620368,
"loss": 1.5512585639953613,
"step": 4152
},
{
"epoch": 0.8667709963484611,
"grad_norm": 0.19135600328445435,
"learning_rate": 0.00015945188414824045,
"loss": 1.6250895261764526,
"step": 4154
},
{
"epoch": 0.867188315075639,
"grad_norm": 0.18524521589279175,
"learning_rate": 0.0001589535970102772,
"loss": 1.5601701736450195,
"step": 4156
},
{
"epoch": 0.8676056338028169,
"grad_norm": 0.18301470577716827,
"learning_rate": 0.00015845530987231395,
"loss": 1.6799218654632568,
"step": 4158
},
{
"epoch": 0.8680229525299947,
"grad_norm": 0.17439837753772736,
"learning_rate": 0.00015795702273435066,
"loss": 1.6412632465362549,
"step": 4160
},
{
"epoch": 0.8684402712571727,
"grad_norm": 0.1827184110879898,
"learning_rate": 0.00015745873559638743,
"loss": 1.7115901708602905,
"step": 4162
},
{
"epoch": 0.8688575899843506,
"grad_norm": 0.17392058670520782,
"learning_rate": 0.00015696044845842417,
"loss": 1.626024603843689,
"step": 4164
},
{
"epoch": 0.8692749087115285,
"grad_norm": 0.17182692885398865,
"learning_rate": 0.00015646216132046093,
"loss": 1.748465657234192,
"step": 4166
},
{
"epoch": 0.8696922274387063,
"grad_norm": 0.22162099182605743,
"learning_rate": 0.00015596387418249767,
"loss": 1.7301583290100098,
"step": 4168
},
{
"epoch": 0.8701095461658842,
"grad_norm": 0.16363243758678436,
"learning_rate": 0.00015546558704453443,
"loss": 1.6637986898422241,
"step": 4170
},
{
"epoch": 0.8705268648930621,
"grad_norm": 0.1959371715784073,
"learning_rate": 0.00015496729990657117,
"loss": 1.6402276754379272,
"step": 4172
},
{
"epoch": 0.87094418362024,
"grad_norm": 0.16994540393352509,
"learning_rate": 0.00015446901276860794,
"loss": 1.6850452423095703,
"step": 4174
},
{
"epoch": 0.8713615023474178,
"grad_norm": 0.18915590643882751,
"learning_rate": 0.00015397072563064467,
"loss": 1.7614344358444214,
"step": 4176
},
{
"epoch": 0.8717788210745957,
"grad_norm": 0.21564990282058716,
"learning_rate": 0.00015347243849268144,
"loss": 1.740942358970642,
"step": 4178
},
{
"epoch": 0.8721961398017736,
"grad_norm": 0.17307381331920624,
"learning_rate": 0.00015297415135471815,
"loss": 1.6896008253097534,
"step": 4180
},
{
"epoch": 0.8726134585289514,
"grad_norm": 0.17976069450378418,
"learning_rate": 0.00015247586421675491,
"loss": 1.5140951871871948,
"step": 4182
},
{
"epoch": 0.8730307772561293,
"grad_norm": 0.18065252900123596,
"learning_rate": 0.00015197757707879165,
"loss": 1.6537516117095947,
"step": 4184
},
{
"epoch": 0.8734480959833073,
"grad_norm": 0.17097200453281403,
"learning_rate": 0.00015147928994082842,
"loss": 1.5193525552749634,
"step": 4186
},
{
"epoch": 0.8738654147104852,
"grad_norm": 0.20268303155899048,
"learning_rate": 0.00015098100280286516,
"loss": 1.6746410131454468,
"step": 4188
},
{
"epoch": 0.874282733437663,
"grad_norm": 0.1816016286611557,
"learning_rate": 0.00015048271566490192,
"loss": 1.5819029808044434,
"step": 4190
},
{
"epoch": 0.8747000521648409,
"grad_norm": 0.175307035446167,
"learning_rate": 0.00014998442852693866,
"loss": 1.6841760873794556,
"step": 4192
},
{
"epoch": 0.8751173708920188,
"grad_norm": 0.20195969939231873,
"learning_rate": 0.00014948614138897542,
"loss": 1.6861398220062256,
"step": 4194
},
{
"epoch": 0.8755346896191967,
"grad_norm": 0.18869267404079437,
"learning_rate": 0.00014898785425101216,
"loss": 1.6720318794250488,
"step": 4196
},
{
"epoch": 0.8759520083463745,
"grad_norm": 0.16363096237182617,
"learning_rate": 0.0001484895671130489,
"loss": 1.781600832939148,
"step": 4198
},
{
"epoch": 0.8763693270735524,
"grad_norm": 0.17589423060417175,
"learning_rate": 0.00014799127997508564,
"loss": 1.5781477689743042,
"step": 4200
},
{
"epoch": 0.8767866458007303,
"grad_norm": 0.19791170954704285,
"learning_rate": 0.0001474929928371224,
"loss": 1.5444563627243042,
"step": 4202
},
{
"epoch": 0.8772039645279082,
"grad_norm": 0.24233515560626984,
"learning_rate": 0.00014699470569915914,
"loss": 1.7058964967727661,
"step": 4204
},
{
"epoch": 0.877621283255086,
"grad_norm": 0.1882489025592804,
"learning_rate": 0.0001464964185611959,
"loss": 1.5815865993499756,
"step": 4206
},
{
"epoch": 0.8780386019822639,
"grad_norm": 0.17342013120651245,
"learning_rate": 0.00014599813142323264,
"loss": 1.646026372909546,
"step": 4208
},
{
"epoch": 0.8784559207094418,
"grad_norm": 0.21900935471057892,
"learning_rate": 0.0001454998442852694,
"loss": 1.6741538047790527,
"step": 4210
},
{
"epoch": 0.8788732394366198,
"grad_norm": 0.18033616244792938,
"learning_rate": 0.00014500155714730614,
"loss": 1.6662341356277466,
"step": 4212
},
{
"epoch": 0.8792905581637976,
"grad_norm": 0.17369306087493896,
"learning_rate": 0.00014450327000934288,
"loss": 1.6094465255737305,
"step": 4214
},
{
"epoch": 0.8797078768909755,
"grad_norm": 0.19994284212589264,
"learning_rate": 0.00014400498287137965,
"loss": 1.698074460029602,
"step": 4216
},
{
"epoch": 0.8801251956181534,
"grad_norm": 0.18714231252670288,
"learning_rate": 0.00014350669573341639,
"loss": 1.6944200992584229,
"step": 4218
},
{
"epoch": 0.8805425143453313,
"grad_norm": 0.18917585909366608,
"learning_rate": 0.00014300840859545315,
"loss": 1.6781073808670044,
"step": 4220
},
{
"epoch": 0.8809598330725091,
"grad_norm": 0.1856914758682251,
"learning_rate": 0.0001425101214574899,
"loss": 1.825918197631836,
"step": 4222
},
{
"epoch": 0.881377151799687,
"grad_norm": 0.18396899104118347,
"learning_rate": 0.00014201183431952663,
"loss": 1.5716090202331543,
"step": 4224
},
{
"epoch": 0.8817944705268649,
"grad_norm": 0.1865542232990265,
"learning_rate": 0.0001415135471815634,
"loss": 1.6458405256271362,
"step": 4226
},
{
"epoch": 0.8822117892540428,
"grad_norm": 0.23121407628059387,
"learning_rate": 0.00014101526004360013,
"loss": 1.6912405490875244,
"step": 4228
},
{
"epoch": 0.8826291079812206,
"grad_norm": 0.24559205770492554,
"learning_rate": 0.00014051697290563687,
"loss": 1.7803088426589966,
"step": 4230
},
{
"epoch": 0.8830464267083985,
"grad_norm": 0.19092410802841187,
"learning_rate": 0.00014001868576767363,
"loss": 1.6274560689926147,
"step": 4232
},
{
"epoch": 0.8834637454355764,
"grad_norm": 0.19192136824131012,
"learning_rate": 0.00013952039862971037,
"loss": 1.5976357460021973,
"step": 4234
},
{
"epoch": 0.8838810641627542,
"grad_norm": 0.19966883957386017,
"learning_rate": 0.00013902211149174713,
"loss": 1.6499944925308228,
"step": 4236
},
{
"epoch": 0.8842983828899322,
"grad_norm": 0.18240401148796082,
"learning_rate": 0.00013852382435378387,
"loss": 1.6760348081588745,
"step": 4238
},
{
"epoch": 0.8847157016171101,
"grad_norm": 0.22000983357429504,
"learning_rate": 0.00013802553721582064,
"loss": 1.765512466430664,
"step": 4240
},
{
"epoch": 0.885133020344288,
"grad_norm": 0.17965154349803925,
"learning_rate": 0.00013752725007785737,
"loss": 1.5602142810821533,
"step": 4242
},
{
"epoch": 0.8855503390714659,
"grad_norm": 0.20459780097007751,
"learning_rate": 0.0001370289629398941,
"loss": 1.513783574104309,
"step": 4244
},
{
"epoch": 0.8859676577986437,
"grad_norm": 0.166926771402359,
"learning_rate": 0.00013653067580193085,
"loss": 1.7161343097686768,
"step": 4246
},
{
"epoch": 0.8863849765258216,
"grad_norm": 0.17636346817016602,
"learning_rate": 0.00013603238866396762,
"loss": 1.6716426610946655,
"step": 4248
},
{
"epoch": 0.8868022952529995,
"grad_norm": 0.1830795854330063,
"learning_rate": 0.00013553410152600435,
"loss": 1.5723251104354858,
"step": 4250
},
{
"epoch": 0.8872196139801773,
"grad_norm": 0.185004323720932,
"learning_rate": 0.00013503581438804112,
"loss": 1.6882978677749634,
"step": 4252
},
{
"epoch": 0.8876369327073552,
"grad_norm": 0.17411819100379944,
"learning_rate": 0.00013453752725007786,
"loss": 1.7141852378845215,
"step": 4254
},
{
"epoch": 0.8880542514345331,
"grad_norm": 0.1981218308210373,
"learning_rate": 0.00013403924011211462,
"loss": 1.641402006149292,
"step": 4256
},
{
"epoch": 0.888471570161711,
"grad_norm": 0.17821228504180908,
"learning_rate": 0.00013354095297415136,
"loss": 1.6480239629745483,
"step": 4258
},
{
"epoch": 0.8888888888888888,
"grad_norm": 0.1579664647579193,
"learning_rate": 0.00013304266583618812,
"loss": 1.712760329246521,
"step": 4260
},
{
"epoch": 0.8893062076160668,
"grad_norm": 0.19413398206233978,
"learning_rate": 0.00013254437869822486,
"loss": 1.7463946342468262,
"step": 4262
},
{
"epoch": 0.8897235263432447,
"grad_norm": 0.2169637531042099,
"learning_rate": 0.0001320460915602616,
"loss": 1.4564887285232544,
"step": 4264
},
{
"epoch": 0.8901408450704226,
"grad_norm": 0.1617715209722519,
"learning_rate": 0.00013154780442229834,
"loss": 1.6625990867614746,
"step": 4266
},
{
"epoch": 0.8905581637976004,
"grad_norm": 0.17531977593898773,
"learning_rate": 0.0001310495172843351,
"loss": 1.5706746578216553,
"step": 4268
},
{
"epoch": 0.8909754825247783,
"grad_norm": 0.18917573988437653,
"learning_rate": 0.00013055123014637184,
"loss": 1.6090871095657349,
"step": 4270
},
{
"epoch": 0.8913928012519562,
"grad_norm": 0.16727563738822937,
"learning_rate": 0.0001300529430084086,
"loss": 1.4635460376739502,
"step": 4272
},
{
"epoch": 0.8918101199791341,
"grad_norm": 0.19415424764156342,
"learning_rate": 0.00012955465587044534,
"loss": 1.6564855575561523,
"step": 4274
},
{
"epoch": 0.8922274387063119,
"grad_norm": 0.17781133949756622,
"learning_rate": 0.0001290563687324821,
"loss": 1.6546841859817505,
"step": 4276
},
{
"epoch": 0.8926447574334898,
"grad_norm": 0.17661771178245544,
"learning_rate": 0.00012855808159451885,
"loss": 1.615142583847046,
"step": 4278
},
{
"epoch": 0.8930620761606677,
"grad_norm": 0.189763605594635,
"learning_rate": 0.0001280597944565556,
"loss": 1.648653268814087,
"step": 4280
},
{
"epoch": 0.8934793948878456,
"grad_norm": 0.1889771819114685,
"learning_rate": 0.00012756150731859235,
"loss": 1.60526704788208,
"step": 4282
},
{
"epoch": 0.8938967136150234,
"grad_norm": 0.20051011443138123,
"learning_rate": 0.00012706322018062909,
"loss": 1.6328041553497314,
"step": 4284
},
{
"epoch": 0.8943140323422013,
"grad_norm": 0.24941663444042206,
"learning_rate": 0.00012656493304266582,
"loss": 1.7087427377700806,
"step": 4286
},
{
"epoch": 0.8947313510693793,
"grad_norm": 0.17458072304725647,
"learning_rate": 0.0001260666459047026,
"loss": 1.7527358531951904,
"step": 4288
},
{
"epoch": 0.8951486697965572,
"grad_norm": 0.16338342428207397,
"learning_rate": 0.00012556835876673933,
"loss": 1.6104542016983032,
"step": 4290
},
{
"epoch": 0.895565988523735,
"grad_norm": 0.1791841834783554,
"learning_rate": 0.0001250700716287761,
"loss": 1.5636260509490967,
"step": 4292
},
{
"epoch": 0.8959833072509129,
"grad_norm": 0.1994152069091797,
"learning_rate": 0.00012457178449081283,
"loss": 1.5894802808761597,
"step": 4294
},
{
"epoch": 0.8964006259780908,
"grad_norm": 0.1872299164533615,
"learning_rate": 0.0001240734973528496,
"loss": 1.6312882900238037,
"step": 4296
},
{
"epoch": 0.8968179447052687,
"grad_norm": 0.18088090419769287,
"learning_rate": 0.00012357521021488633,
"loss": 1.5215966701507568,
"step": 4298
},
{
"epoch": 0.8972352634324465,
"grad_norm": 0.17654314637184143,
"learning_rate": 0.0001230769230769231,
"loss": 1.5884132385253906,
"step": 4300
},
{
"epoch": 0.8976525821596244,
"grad_norm": 0.17969003319740295,
"learning_rate": 0.00012257863593895984,
"loss": 1.5834084749221802,
"step": 4302
},
{
"epoch": 0.8980699008868023,
"grad_norm": 0.1986774057149887,
"learning_rate": 0.0001220803488009966,
"loss": 1.6546990871429443,
"step": 4304
},
{
"epoch": 0.8984872196139801,
"grad_norm": 0.2006673812866211,
"learning_rate": 0.00012158206166303332,
"loss": 1.5141408443450928,
"step": 4306
},
{
"epoch": 0.898904538341158,
"grad_norm": 0.18943582475185394,
"learning_rate": 0.00012108377452507008,
"loss": 1.582223892211914,
"step": 4308
},
{
"epoch": 0.8993218570683359,
"grad_norm": 0.18200165033340454,
"learning_rate": 0.00012058548738710683,
"loss": 1.5982892513275146,
"step": 4310
},
{
"epoch": 0.8997391757955139,
"grad_norm": 0.2065363973379135,
"learning_rate": 0.00012008720024914358,
"loss": 1.598642349243164,
"step": 4312
},
{
"epoch": 0.9001564945226918,
"grad_norm": 0.16489578783512115,
"learning_rate": 0.00011958891311118032,
"loss": 1.6891593933105469,
"step": 4314
},
{
"epoch": 0.9005738132498696,
"grad_norm": 0.20588769018650055,
"learning_rate": 0.00011909062597321707,
"loss": 1.6844552755355835,
"step": 4316
},
{
"epoch": 0.9009911319770475,
"grad_norm": 0.22354429960250854,
"learning_rate": 0.00011859233883525382,
"loss": 1.479601263999939,
"step": 4318
},
{
"epoch": 0.9014084507042254,
"grad_norm": 0.21314990520477295,
"learning_rate": 0.00011809405169729057,
"loss": 1.5884896516799927,
"step": 4320
},
{
"epoch": 0.9018257694314032,
"grad_norm": 0.19360527396202087,
"learning_rate": 0.00011759576455932732,
"loss": 1.6685384511947632,
"step": 4322
},
{
"epoch": 0.9022430881585811,
"grad_norm": 0.1771700084209442,
"learning_rate": 0.00011709747742136407,
"loss": 1.6167325973510742,
"step": 4324
},
{
"epoch": 0.902660406885759,
"grad_norm": 0.17964008450508118,
"learning_rate": 0.00011659919028340081,
"loss": 1.8029953241348267,
"step": 4326
},
{
"epoch": 0.9030777256129369,
"grad_norm": 0.16711945831775665,
"learning_rate": 0.00011610090314543756,
"loss": 1.7099660634994507,
"step": 4328
},
{
"epoch": 0.9034950443401147,
"grad_norm": 0.17991159856319427,
"learning_rate": 0.0001156026160074743,
"loss": 1.6438169479370117,
"step": 4330
},
{
"epoch": 0.9039123630672926,
"grad_norm": 0.19515983760356903,
"learning_rate": 0.00011510432886951105,
"loss": 1.660976529121399,
"step": 4332
},
{
"epoch": 0.9043296817944705,
"grad_norm": 0.1659526377916336,
"learning_rate": 0.0001146060417315478,
"loss": 1.6571065187454224,
"step": 4334
},
{
"epoch": 0.9047470005216484,
"grad_norm": 0.189217209815979,
"learning_rate": 0.00011410775459358455,
"loss": 1.6296979188919067,
"step": 4336
},
{
"epoch": 0.9051643192488263,
"grad_norm": 0.17310501635074615,
"learning_rate": 0.0001136094674556213,
"loss": 1.6209598779678345,
"step": 4338
},
{
"epoch": 0.9055816379760042,
"grad_norm": 0.17237012088298798,
"learning_rate": 0.00011311118031765806,
"loss": 1.6465295553207397,
"step": 4340
},
{
"epoch": 0.9059989567031821,
"grad_norm": 0.1960352063179016,
"learning_rate": 0.00011261289317969481,
"loss": 1.5829805135726929,
"step": 4342
},
{
"epoch": 0.90641627543036,
"grad_norm": 0.19082073867321014,
"learning_rate": 0.00011211460604173156,
"loss": 1.6729804277420044,
"step": 4344
},
{
"epoch": 0.9068335941575378,
"grad_norm": 0.1686243861913681,
"learning_rate": 0.00011161631890376828,
"loss": 1.5603147745132446,
"step": 4346
},
{
"epoch": 0.9072509128847157,
"grad_norm": 0.20412971079349518,
"learning_rate": 0.00011111803176580504,
"loss": 1.5871728658676147,
"step": 4348
},
{
"epoch": 0.9076682316118936,
"grad_norm": 0.1717957705259323,
"learning_rate": 0.00011061974462784179,
"loss": 1.6768443584442139,
"step": 4350
},
{
"epoch": 0.9080855503390715,
"grad_norm": 0.1993514597415924,
"learning_rate": 0.00011012145748987854,
"loss": 1.724295735359192,
"step": 4352
},
{
"epoch": 0.9085028690662493,
"grad_norm": 0.1855616569519043,
"learning_rate": 0.00010962317035191529,
"loss": 1.6664015054702759,
"step": 4354
},
{
"epoch": 0.9089201877934272,
"grad_norm": 0.20700250566005707,
"learning_rate": 0.00010912488321395204,
"loss": 1.4902336597442627,
"step": 4356
},
{
"epoch": 0.9093375065206051,
"grad_norm": 0.1661575436592102,
"learning_rate": 0.00010862659607598879,
"loss": 1.6248383522033691,
"step": 4358
},
{
"epoch": 0.909754825247783,
"grad_norm": 0.183964803814888,
"learning_rate": 0.00010812830893802554,
"loss": 1.6356079578399658,
"step": 4360
},
{
"epoch": 0.9101721439749608,
"grad_norm": 0.20293103158473969,
"learning_rate": 0.0001076300218000623,
"loss": 1.6394023895263672,
"step": 4362
},
{
"epoch": 0.9105894627021388,
"grad_norm": 0.20069614052772522,
"learning_rate": 0.00010713173466209905,
"loss": 1.6961530447006226,
"step": 4364
},
{
"epoch": 0.9110067814293167,
"grad_norm": 0.18666695058345795,
"learning_rate": 0.0001066334475241358,
"loss": 1.5653066635131836,
"step": 4366
},
{
"epoch": 0.9114241001564946,
"grad_norm": 0.21359233558177948,
"learning_rate": 0.00010613516038617252,
"loss": 1.55157470703125,
"step": 4368
},
{
"epoch": 0.9118414188836724,
"grad_norm": 0.1628497689962387,
"learning_rate": 0.00010563687324820927,
"loss": 1.6790190935134888,
"step": 4370
},
{
"epoch": 0.9122587376108503,
"grad_norm": 0.191795215010643,
"learning_rate": 0.00010513858611024603,
"loss": 1.6705849170684814,
"step": 4372
},
{
"epoch": 0.9126760563380282,
"grad_norm": 0.1760699450969696,
"learning_rate": 0.00010464029897228278,
"loss": 1.6555774211883545,
"step": 4374
},
{
"epoch": 0.913093375065206,
"grad_norm": 0.21481812000274658,
"learning_rate": 0.00010414201183431953,
"loss": 1.5153712034225464,
"step": 4376
},
{
"epoch": 0.9135106937923839,
"grad_norm": 0.1772165447473526,
"learning_rate": 0.00010364372469635628,
"loss": 1.6661906242370605,
"step": 4378
},
{
"epoch": 0.9139280125195618,
"grad_norm": 0.1901162713766098,
"learning_rate": 0.00010314543755839303,
"loss": 1.5453205108642578,
"step": 4380
},
{
"epoch": 0.9143453312467397,
"grad_norm": 0.17241789400577545,
"learning_rate": 0.00010264715042042978,
"loss": 1.7289237976074219,
"step": 4382
},
{
"epoch": 0.9147626499739175,
"grad_norm": 0.16452018916606903,
"learning_rate": 0.00010214886328246653,
"loss": 1.67987060546875,
"step": 4384
},
{
"epoch": 0.9151799687010954,
"grad_norm": 0.20041242241859436,
"learning_rate": 0.00010165057614450329,
"loss": 1.6784254312515259,
"step": 4386
},
{
"epoch": 0.9155972874282734,
"grad_norm": 0.20399820804595947,
"learning_rate": 0.00010115228900654001,
"loss": 1.6442837715148926,
"step": 4388
},
{
"epoch": 0.9160146061554513,
"grad_norm": 0.21368463337421417,
"learning_rate": 0.00010065400186857676,
"loss": 1.5905625820159912,
"step": 4390
},
{
"epoch": 0.9164319248826291,
"grad_norm": 0.17233221232891083,
"learning_rate": 0.00010015571473061351,
"loss": 1.7597204446792603,
"step": 4392
},
{
"epoch": 0.916849243609807,
"grad_norm": 0.18968565762043,
"learning_rate": 9.965742759265026e-05,
"loss": 1.6928092241287231,
"step": 4394
},
{
"epoch": 0.9172665623369849,
"grad_norm": 0.1855761855840683,
"learning_rate": 9.915914045468702e-05,
"loss": 1.5505868196487427,
"step": 4396
},
{
"epoch": 0.9176838810641628,
"grad_norm": 0.20371851325035095,
"learning_rate": 9.866085331672377e-05,
"loss": 1.6179020404815674,
"step": 4398
},
{
"epoch": 0.9181011997913406,
"grad_norm": 0.20698843896389008,
"learning_rate": 9.816256617876052e-05,
"loss": 1.5716030597686768,
"step": 4400
},
{
"epoch": 0.9185185185185185,
"grad_norm": 0.16781571507453918,
"learning_rate": 9.766427904079727e-05,
"loss": 1.6481271982192993,
"step": 4402
},
{
"epoch": 0.9189358372456964,
"grad_norm": 0.16532212495803833,
"learning_rate": 9.716599190283401e-05,
"loss": 1.7493208646774292,
"step": 4404
},
{
"epoch": 0.9193531559728743,
"grad_norm": 0.19232137501239777,
"learning_rate": 9.666770476487076e-05,
"loss": 1.5340021848678589,
"step": 4406
},
{
"epoch": 0.9197704747000521,
"grad_norm": 0.2033078521490097,
"learning_rate": 9.616941762690751e-05,
"loss": 1.5652828216552734,
"step": 4408
},
{
"epoch": 0.92018779342723,
"grad_norm": 0.19920817017555237,
"learning_rate": 9.567113048894426e-05,
"loss": 1.648752212524414,
"step": 4410
},
{
"epoch": 0.9206051121544079,
"grad_norm": 0.18810296058654785,
"learning_rate": 9.517284335098101e-05,
"loss": 1.6240581274032593,
"step": 4412
},
{
"epoch": 0.9210224308815859,
"grad_norm": 0.1705373376607895,
"learning_rate": 9.467455621301775e-05,
"loss": 1.5519990921020508,
"step": 4414
},
{
"epoch": 0.9214397496087637,
"grad_norm": 0.17323507368564606,
"learning_rate": 9.41762690750545e-05,
"loss": 1.6658997535705566,
"step": 4416
},
{
"epoch": 0.9218570683359416,
"grad_norm": 0.19927287101745605,
"learning_rate": 9.367798193709125e-05,
"loss": 1.7493751049041748,
"step": 4418
},
{
"epoch": 0.9222743870631195,
"grad_norm": 0.1851804256439209,
"learning_rate": 9.3179694799128e-05,
"loss": 1.6033281087875366,
"step": 4420
},
{
"epoch": 0.9226917057902974,
"grad_norm": 0.19267021119594574,
"learning_rate": 9.268140766116476e-05,
"loss": 1.6882092952728271,
"step": 4422
},
{
"epoch": 0.9231090245174752,
"grad_norm": 0.16980218887329102,
"learning_rate": 9.21831205232015e-05,
"loss": 1.6960489749908447,
"step": 4424
},
{
"epoch": 0.9235263432446531,
"grad_norm": 0.18636591732501984,
"learning_rate": 9.168483338523825e-05,
"loss": 1.6686335802078247,
"step": 4426
},
{
"epoch": 0.923943661971831,
"grad_norm": 0.19164636731147766,
"learning_rate": 9.1186546247275e-05,
"loss": 1.5378191471099854,
"step": 4428
},
{
"epoch": 0.9243609806990088,
"grad_norm": 0.19653864204883575,
"learning_rate": 9.068825910931175e-05,
"loss": 1.6618753671646118,
"step": 4430
},
{
"epoch": 0.9247782994261867,
"grad_norm": 0.18968534469604492,
"learning_rate": 9.01899719713485e-05,
"loss": 1.6201266050338745,
"step": 4432
},
{
"epoch": 0.9251956181533646,
"grad_norm": 0.19709599018096924,
"learning_rate": 8.969168483338525e-05,
"loss": 1.664633870124817,
"step": 4434
},
{
"epoch": 0.9256129368805425,
"grad_norm": 0.18362024426460266,
"learning_rate": 8.919339769542199e-05,
"loss": 1.6451623439788818,
"step": 4436
},
{
"epoch": 0.9260302556077203,
"grad_norm": 0.2022915780544281,
"learning_rate": 8.869511055745874e-05,
"loss": 1.658370018005371,
"step": 4438
},
{
"epoch": 0.9264475743348983,
"grad_norm": 0.19555260241031647,
"learning_rate": 8.819682341949549e-05,
"loss": 1.6008943319320679,
"step": 4440
},
{
"epoch": 0.9268648930620762,
"grad_norm": 0.17300760746002197,
"learning_rate": 8.769853628153224e-05,
"loss": 1.7432292699813843,
"step": 4442
},
{
"epoch": 0.9272822117892541,
"grad_norm": 0.1831115186214447,
"learning_rate": 8.7200249143569e-05,
"loss": 1.7163116931915283,
"step": 4444
},
{
"epoch": 0.927699530516432,
"grad_norm": 0.19250436127185822,
"learning_rate": 8.670196200560573e-05,
"loss": 1.4830156564712524,
"step": 4446
},
{
"epoch": 0.9281168492436098,
"grad_norm": 0.20709781348705292,
"learning_rate": 8.620367486764248e-05,
"loss": 1.5859413146972656,
"step": 4448
},
{
"epoch": 0.9285341679707877,
"grad_norm": 0.19259729981422424,
"learning_rate": 8.570538772967923e-05,
"loss": 1.5829150676727295,
"step": 4450
},
{
"epoch": 0.9289514866979656,
"grad_norm": 0.19801819324493408,
"learning_rate": 8.520710059171599e-05,
"loss": 1.5828293561935425,
"step": 4452
},
{
"epoch": 0.9293688054251434,
"grad_norm": 0.16226257383823395,
"learning_rate": 8.470881345375274e-05,
"loss": 1.705340027809143,
"step": 4454
},
{
"epoch": 0.9297861241523213,
"grad_norm": 0.18210093677043915,
"learning_rate": 8.421052631578948e-05,
"loss": 1.704031229019165,
"step": 4456
},
{
"epoch": 0.9302034428794992,
"grad_norm": 0.19659306108951569,
"learning_rate": 8.371223917782623e-05,
"loss": 1.660591721534729,
"step": 4458
},
{
"epoch": 0.9306207616066771,
"grad_norm": 0.2188255935907364,
"learning_rate": 8.321395203986298e-05,
"loss": 1.5957276821136475,
"step": 4460
},
{
"epoch": 0.9310380803338549,
"grad_norm": 0.21026277542114258,
"learning_rate": 8.271566490189973e-05,
"loss": 1.7397994995117188,
"step": 4462
},
{
"epoch": 0.9314553990610329,
"grad_norm": 0.1685836762189865,
"learning_rate": 8.221737776393648e-05,
"loss": 1.7202115058898926,
"step": 4464
},
{
"epoch": 0.9318727177882108,
"grad_norm": 0.1884559541940689,
"learning_rate": 8.171909062597322e-05,
"loss": 1.7587333917617798,
"step": 4466
},
{
"epoch": 0.9322900365153887,
"grad_norm": 0.1983492523431778,
"learning_rate": 8.122080348800997e-05,
"loss": 1.6561076641082764,
"step": 4468
},
{
"epoch": 0.9327073552425665,
"grad_norm": 0.20395974814891815,
"learning_rate": 8.072251635004672e-05,
"loss": 1.568048357963562,
"step": 4470
},
{
"epoch": 0.9331246739697444,
"grad_norm": 0.16944561898708344,
"learning_rate": 8.022422921208347e-05,
"loss": 1.6114654541015625,
"step": 4472
},
{
"epoch": 0.9335419926969223,
"grad_norm": 0.19421088695526123,
"learning_rate": 7.972594207412022e-05,
"loss": 1.6249094009399414,
"step": 4474
},
{
"epoch": 0.9339593114241002,
"grad_norm": 0.1974789798259735,
"learning_rate": 7.922765493615698e-05,
"loss": 1.6187351942062378,
"step": 4476
},
{
"epoch": 0.934376630151278,
"grad_norm": 0.1901128739118576,
"learning_rate": 7.872936779819371e-05,
"loss": 1.5681487321853638,
"step": 4478
},
{
"epoch": 0.9347939488784559,
"grad_norm": 0.20003391802310944,
"learning_rate": 7.823108066023047e-05,
"loss": 1.5453399419784546,
"step": 4480
},
{
"epoch": 0.9352112676056338,
"grad_norm": 0.19723911583423615,
"learning_rate": 7.773279352226722e-05,
"loss": 1.5492863655090332,
"step": 4482
},
{
"epoch": 0.9356285863328117,
"grad_norm": 0.1812390238046646,
"learning_rate": 7.723450638430397e-05,
"loss": 1.804068684577942,
"step": 4484
},
{
"epoch": 0.9360459050599895,
"grad_norm": 0.20390065014362335,
"learning_rate": 7.673621924634072e-05,
"loss": 1.7171461582183838,
"step": 4486
},
{
"epoch": 0.9364632237871674,
"grad_norm": 0.1780429482460022,
"learning_rate": 7.623793210837746e-05,
"loss": 1.7452282905578613,
"step": 4488
},
{
"epoch": 0.9368805425143454,
"grad_norm": 0.2048795223236084,
"learning_rate": 7.573964497041421e-05,
"loss": 1.7104570865631104,
"step": 4490
},
{
"epoch": 0.9372978612415233,
"grad_norm": 0.2167808711528778,
"learning_rate": 7.524135783245096e-05,
"loss": 1.6328846216201782,
"step": 4492
},
{
"epoch": 0.9377151799687011,
"grad_norm": 0.19913658499717712,
"learning_rate": 7.474307069448771e-05,
"loss": 1.621012806892395,
"step": 4494
},
{
"epoch": 0.938132498695879,
"grad_norm": 0.2030884027481079,
"learning_rate": 7.424478355652445e-05,
"loss": 1.4973490238189697,
"step": 4496
},
{
"epoch": 0.9385498174230569,
"grad_norm": 0.18171103298664093,
"learning_rate": 7.37464964185612e-05,
"loss": 1.7280266284942627,
"step": 4498
},
{
"epoch": 0.9389671361502347,
"grad_norm": 0.17654871940612793,
"learning_rate": 7.324820928059795e-05,
"loss": 1.517480492591858,
"step": 4500
},
{
"epoch": 0.9393844548774126,
"grad_norm": 0.18540678918361664,
"learning_rate": 7.27499221426347e-05,
"loss": 1.6535931825637817,
"step": 4502
},
{
"epoch": 0.9398017736045905,
"grad_norm": 0.20257021486759186,
"learning_rate": 7.225163500467144e-05,
"loss": 1.5644900798797607,
"step": 4504
},
{
"epoch": 0.9402190923317684,
"grad_norm": 0.187891885638237,
"learning_rate": 7.175334786670819e-05,
"loss": 1.8327016830444336,
"step": 4506
},
{
"epoch": 0.9406364110589462,
"grad_norm": 0.16715452075004578,
"learning_rate": 7.125506072874494e-05,
"loss": 1.807386875152588,
"step": 4508
},
{
"epoch": 0.9410537297861241,
"grad_norm": 0.2130834013223648,
"learning_rate": 7.07567735907817e-05,
"loss": 1.691344976425171,
"step": 4510
},
{
"epoch": 0.941471048513302,
"grad_norm": 0.18834607303142548,
"learning_rate": 7.025848645281843e-05,
"loss": 1.7030231952667236,
"step": 4512
},
{
"epoch": 0.9418883672404799,
"grad_norm": 0.20961347222328186,
"learning_rate": 6.976019931485518e-05,
"loss": 1.5294463634490967,
"step": 4514
},
{
"epoch": 0.9423056859676578,
"grad_norm": 0.162283256649971,
"learning_rate": 6.926191217689194e-05,
"loss": 1.7005258798599243,
"step": 4516
},
{
"epoch": 0.9427230046948357,
"grad_norm": 0.17353390157222748,
"learning_rate": 6.876362503892869e-05,
"loss": 1.8133533000946045,
"step": 4518
},
{
"epoch": 0.9431403234220136,
"grad_norm": 0.2169540375471115,
"learning_rate": 6.826533790096543e-05,
"loss": 1.6217796802520752,
"step": 4520
},
{
"epoch": 0.9435576421491915,
"grad_norm": 0.1963813304901123,
"learning_rate": 6.776705076300218e-05,
"loss": 1.6406642198562622,
"step": 4522
},
{
"epoch": 0.9439749608763693,
"grad_norm": 0.18026936054229736,
"learning_rate": 6.726876362503893e-05,
"loss": 1.6873219013214111,
"step": 4524
},
{
"epoch": 0.9443922796035472,
"grad_norm": 0.18971621990203857,
"learning_rate": 6.677047648707568e-05,
"loss": 1.6877599954605103,
"step": 4526
},
{
"epoch": 0.9448095983307251,
"grad_norm": 0.16752666234970093,
"learning_rate": 6.627218934911243e-05,
"loss": 1.7026634216308594,
"step": 4528
},
{
"epoch": 0.945226917057903,
"grad_norm": 0.17078153789043427,
"learning_rate": 6.577390221114917e-05,
"loss": 1.4954590797424316,
"step": 4530
},
{
"epoch": 0.9456442357850808,
"grad_norm": 0.17435191571712494,
"learning_rate": 6.527561507318592e-05,
"loss": 1.6167274713516235,
"step": 4532
},
{
"epoch": 0.9460615545122587,
"grad_norm": 0.1811375617980957,
"learning_rate": 6.477732793522267e-05,
"loss": 1.5678499937057495,
"step": 4534
},
{
"epoch": 0.9464788732394366,
"grad_norm": 0.19363082945346832,
"learning_rate": 6.427904079725942e-05,
"loss": 1.6367244720458984,
"step": 4536
},
{
"epoch": 0.9468961919666145,
"grad_norm": 0.18442240357398987,
"learning_rate": 6.378075365929617e-05,
"loss": 1.610424518585205,
"step": 4538
},
{
"epoch": 0.9473135106937924,
"grad_norm": 0.1918598711490631,
"learning_rate": 6.328246652133291e-05,
"loss": 1.575549840927124,
"step": 4540
},
{
"epoch": 0.9477308294209703,
"grad_norm": 0.17891459167003632,
"learning_rate": 6.278417938336966e-05,
"loss": 1.686457872390747,
"step": 4542
},
{
"epoch": 0.9481481481481482,
"grad_norm": 0.19015757739543915,
"learning_rate": 6.228589224540641e-05,
"loss": 1.6366968154907227,
"step": 4544
},
{
"epoch": 0.9485654668753261,
"grad_norm": 0.1784735769033432,
"learning_rate": 6.178760510744317e-05,
"loss": 1.5920610427856445,
"step": 4546
},
{
"epoch": 0.9489827856025039,
"grad_norm": 0.1828586757183075,
"learning_rate": 6.128931796947992e-05,
"loss": 1.5715728998184204,
"step": 4548
},
{
"epoch": 0.9494001043296818,
"grad_norm": 0.2356758713722229,
"learning_rate": 6.079103083151666e-05,
"loss": 1.497330665588379,
"step": 4550
},
{
"epoch": 0.9498174230568597,
"grad_norm": 0.17772871255874634,
"learning_rate": 6.0292743693553414e-05,
"loss": 1.6327942609786987,
"step": 4552
},
{
"epoch": 0.9502347417840376,
"grad_norm": 0.17223063111305237,
"learning_rate": 5.979445655559016e-05,
"loss": 1.5001283884048462,
"step": 4554
},
{
"epoch": 0.9506520605112154,
"grad_norm": 0.1957160383462906,
"learning_rate": 5.929616941762691e-05,
"loss": 1.5489630699157715,
"step": 4556
},
{
"epoch": 0.9510693792383933,
"grad_norm": 0.1802961230278015,
"learning_rate": 5.879788227966366e-05,
"loss": 1.6028399467468262,
"step": 4558
},
{
"epoch": 0.9514866979655712,
"grad_norm": 0.2035115361213684,
"learning_rate": 5.8299595141700406e-05,
"loss": 1.7312326431274414,
"step": 4560
},
{
"epoch": 0.951904016692749,
"grad_norm": 0.19013307988643646,
"learning_rate": 5.780130800373715e-05,
"loss": 1.6920156478881836,
"step": 4562
},
{
"epoch": 0.9523213354199269,
"grad_norm": 0.1732788383960724,
"learning_rate": 5.73030208657739e-05,
"loss": 1.6895772218704224,
"step": 4564
},
{
"epoch": 0.9527386541471049,
"grad_norm": 0.16644273698329926,
"learning_rate": 5.680473372781065e-05,
"loss": 1.6125408411026,
"step": 4566
},
{
"epoch": 0.9531559728742828,
"grad_norm": 0.1578505039215088,
"learning_rate": 5.6306446589847404e-05,
"loss": 1.5833113193511963,
"step": 4568
},
{
"epoch": 0.9535732916014606,
"grad_norm": 0.22270844876766205,
"learning_rate": 5.580815945188414e-05,
"loss": 1.4422922134399414,
"step": 4570
},
{
"epoch": 0.9539906103286385,
"grad_norm": 0.1921321600675583,
"learning_rate": 5.5309872313920894e-05,
"loss": 1.6108051538467407,
"step": 4572
},
{
"epoch": 0.9544079290558164,
"grad_norm": 0.1873205453157425,
"learning_rate": 5.4811585175957645e-05,
"loss": 1.689296007156372,
"step": 4574
},
{
"epoch": 0.9548252477829943,
"grad_norm": 0.18801668286323547,
"learning_rate": 5.4313298037994396e-05,
"loss": 1.6304564476013184,
"step": 4576
},
{
"epoch": 0.9552425665101721,
"grad_norm": 0.1667712926864624,
"learning_rate": 5.381501090003115e-05,
"loss": 1.6233569383621216,
"step": 4578
},
{
"epoch": 0.95565988523735,
"grad_norm": 0.2118733823299408,
"learning_rate": 5.33167237620679e-05,
"loss": 1.6427783966064453,
"step": 4580
},
{
"epoch": 0.9560772039645279,
"grad_norm": 0.20205016434192657,
"learning_rate": 5.281843662410464e-05,
"loss": 1.6134835481643677,
"step": 4582
},
{
"epoch": 0.9564945226917058,
"grad_norm": 0.18509261310100555,
"learning_rate": 5.232014948614139e-05,
"loss": 1.6394373178482056,
"step": 4584
},
{
"epoch": 0.9569118414188836,
"grad_norm": 0.17359930276870728,
"learning_rate": 5.182186234817814e-05,
"loss": 1.6565989255905151,
"step": 4586
},
{
"epoch": 0.9573291601460615,
"grad_norm": 0.19042231142520905,
"learning_rate": 5.132357521021489e-05,
"loss": 1.7013580799102783,
"step": 4588
},
{
"epoch": 0.9577464788732394,
"grad_norm": 0.17238008975982666,
"learning_rate": 5.082528807225164e-05,
"loss": 1.6243165731430054,
"step": 4590
},
{
"epoch": 0.9581637976004174,
"grad_norm": 0.14978916943073273,
"learning_rate": 5.032700093428838e-05,
"loss": 1.7012317180633545,
"step": 4592
},
{
"epoch": 0.9585811163275952,
"grad_norm": 0.24301239848136902,
"learning_rate": 4.982871379632513e-05,
"loss": 1.553206443786621,
"step": 4594
},
{
"epoch": 0.9589984350547731,
"grad_norm": 0.20136353373527527,
"learning_rate": 4.933042665836188e-05,
"loss": 1.5516796112060547,
"step": 4596
},
{
"epoch": 0.959415753781951,
"grad_norm": 0.18118280172348022,
"learning_rate": 4.8832139520398635e-05,
"loss": 1.652338981628418,
"step": 4598
},
{
"epoch": 0.9598330725091289,
"grad_norm": 0.1630065143108368,
"learning_rate": 4.833385238243538e-05,
"loss": 1.6231677532196045,
"step": 4600
},
{
"epoch": 0.9602503912363067,
"grad_norm": 0.1724662035703659,
"learning_rate": 4.783556524447213e-05,
"loss": 1.7110240459442139,
"step": 4602
},
{
"epoch": 0.9606677099634846,
"grad_norm": 0.19082118570804596,
"learning_rate": 4.7337278106508875e-05,
"loss": 1.6536962985992432,
"step": 4604
},
{
"epoch": 0.9610850286906625,
"grad_norm": 0.18775275349617004,
"learning_rate": 4.683899096854563e-05,
"loss": 1.4913078546524048,
"step": 4606
},
{
"epoch": 0.9615023474178404,
"grad_norm": 0.18270009756088257,
"learning_rate": 4.634070383058238e-05,
"loss": 1.6964378356933594,
"step": 4608
},
{
"epoch": 0.9619196661450182,
"grad_norm": 0.20057393610477448,
"learning_rate": 4.584241669261912e-05,
"loss": 1.5497608184814453,
"step": 4610
},
{
"epoch": 0.9623369848721961,
"grad_norm": 0.18243643641471863,
"learning_rate": 4.5344129554655874e-05,
"loss": 1.6174226999282837,
"step": 4612
},
{
"epoch": 0.962754303599374,
"grad_norm": 0.18498554825782776,
"learning_rate": 4.4845842416692625e-05,
"loss": 1.7175030708312988,
"step": 4614
},
{
"epoch": 0.963171622326552,
"grad_norm": 0.19206179678440094,
"learning_rate": 4.434755527872937e-05,
"loss": 1.4646062850952148,
"step": 4616
},
{
"epoch": 0.9635889410537298,
"grad_norm": 0.17778988182544708,
"learning_rate": 4.384926814076612e-05,
"loss": 1.4982643127441406,
"step": 4618
},
{
"epoch": 0.9640062597809077,
"grad_norm": 0.1700717955827713,
"learning_rate": 4.3350981002802866e-05,
"loss": 1.5984323024749756,
"step": 4620
},
{
"epoch": 0.9644235785080856,
"grad_norm": 0.17402176558971405,
"learning_rate": 4.285269386483962e-05,
"loss": 1.5806792974472046,
"step": 4622
},
{
"epoch": 0.9648408972352635,
"grad_norm": 0.21333187818527222,
"learning_rate": 4.235440672687637e-05,
"loss": 1.7254987955093384,
"step": 4624
},
{
"epoch": 0.9652582159624413,
"grad_norm": 0.23211519420146942,
"learning_rate": 4.1856119588913113e-05,
"loss": 1.6255877017974854,
"step": 4626
},
{
"epoch": 0.9656755346896192,
"grad_norm": 0.18935810029506683,
"learning_rate": 4.1357832450949865e-05,
"loss": 1.4401967525482178,
"step": 4628
},
{
"epoch": 0.9660928534167971,
"grad_norm": 0.17174339294433594,
"learning_rate": 4.085954531298661e-05,
"loss": 1.5409501791000366,
"step": 4630
},
{
"epoch": 0.9665101721439749,
"grad_norm": 0.18103395402431488,
"learning_rate": 4.036125817502336e-05,
"loss": 1.560407280921936,
"step": 4632
},
{
"epoch": 0.9669274908711528,
"grad_norm": 0.1579323559999466,
"learning_rate": 3.986297103706011e-05,
"loss": 1.6498744487762451,
"step": 4634
},
{
"epoch": 0.9673448095983307,
"grad_norm": 0.19027456641197205,
"learning_rate": 3.936468389909686e-05,
"loss": 1.6339787244796753,
"step": 4636
},
{
"epoch": 0.9677621283255086,
"grad_norm": 0.1787186563014984,
"learning_rate": 3.886639676113361e-05,
"loss": 1.6012805700302124,
"step": 4638
},
{
"epoch": 0.9681794470526864,
"grad_norm": 0.23373079299926758,
"learning_rate": 3.836810962317036e-05,
"loss": 1.7097437381744385,
"step": 4640
},
{
"epoch": 0.9685967657798644,
"grad_norm": 0.1742262989282608,
"learning_rate": 3.7869822485207104e-05,
"loss": 1.7389062643051147,
"step": 4642
},
{
"epoch": 0.9690140845070423,
"grad_norm": 0.18238705396652222,
"learning_rate": 3.7371535347243856e-05,
"loss": 1.6251798868179321,
"step": 4644
},
{
"epoch": 0.9694314032342202,
"grad_norm": 0.17849200963974,
"learning_rate": 3.68732482092806e-05,
"loss": 1.7080751657485962,
"step": 4646
},
{
"epoch": 0.969848721961398,
"grad_norm": 0.15409721434116364,
"learning_rate": 3.637496107131735e-05,
"loss": 1.646686315536499,
"step": 4648
},
{
"epoch": 0.9702660406885759,
"grad_norm": 0.1952393651008606,
"learning_rate": 3.5876673933354096e-05,
"loss": 1.6519407033920288,
"step": 4650
},
{
"epoch": 0.9706833594157538,
"grad_norm": 0.1741565316915512,
"learning_rate": 3.537838679539085e-05,
"loss": 1.5830824375152588,
"step": 4652
},
{
"epoch": 0.9711006781429317,
"grad_norm": 0.19142566621303558,
"learning_rate": 3.488009965742759e-05,
"loss": 1.5974512100219727,
"step": 4654
},
{
"epoch": 0.9715179968701095,
"grad_norm": 0.19886387884616852,
"learning_rate": 3.4381812519464344e-05,
"loss": 1.580518364906311,
"step": 4656
},
{
"epoch": 0.9719353155972874,
"grad_norm": 0.18001171946525574,
"learning_rate": 3.388352538150109e-05,
"loss": 1.5732975006103516,
"step": 4658
},
{
"epoch": 0.9723526343244653,
"grad_norm": 0.2089742124080658,
"learning_rate": 3.338523824353784e-05,
"loss": 1.613789677619934,
"step": 4660
},
{
"epoch": 0.9727699530516432,
"grad_norm": 0.1852923035621643,
"learning_rate": 3.2886951105574584e-05,
"loss": 1.575705647468567,
"step": 4662
},
{
"epoch": 0.973187271778821,
"grad_norm": 0.18003980815410614,
"learning_rate": 3.2388663967611336e-05,
"loss": 1.641672134399414,
"step": 4664
},
{
"epoch": 0.9736045905059989,
"grad_norm": 0.17273388803005219,
"learning_rate": 3.189037682964809e-05,
"loss": 1.604931354522705,
"step": 4666
},
{
"epoch": 0.9740219092331769,
"grad_norm": 0.18566612899303436,
"learning_rate": 3.139208969168483e-05,
"loss": 1.7001930475234985,
"step": 4668
},
{
"epoch": 0.9744392279603548,
"grad_norm": 0.20328181982040405,
"learning_rate": 3.089380255372158e-05,
"loss": 1.679294466972351,
"step": 4670
},
{
"epoch": 0.9748565466875326,
"grad_norm": 0.1768922060728073,
"learning_rate": 3.039551541575833e-05,
"loss": 1.6282565593719482,
"step": 4672
},
{
"epoch": 0.9752738654147105,
"grad_norm": 0.1997314691543579,
"learning_rate": 2.989722827779508e-05,
"loss": 1.7844572067260742,
"step": 4674
},
{
"epoch": 0.9756911841418884,
"grad_norm": 0.19723384082317352,
"learning_rate": 2.939894113983183e-05,
"loss": 1.7350198030471802,
"step": 4676
},
{
"epoch": 0.9761085028690663,
"grad_norm": 0.1965782791376114,
"learning_rate": 2.8900654001868575e-05,
"loss": 1.5240877866744995,
"step": 4678
},
{
"epoch": 0.9765258215962441,
"grad_norm": 0.21728335320949554,
"learning_rate": 2.8402366863905327e-05,
"loss": 1.6430609226226807,
"step": 4680
},
{
"epoch": 0.976943140323422,
"grad_norm": 0.2064141184091568,
"learning_rate": 2.790407972594207e-05,
"loss": 1.6481841802597046,
"step": 4682
},
{
"epoch": 0.9773604590505999,
"grad_norm": 0.18274162709712982,
"learning_rate": 2.7405792587978823e-05,
"loss": 1.6097731590270996,
"step": 4684
},
{
"epoch": 0.9777777777777777,
"grad_norm": 0.19304390251636505,
"learning_rate": 2.6907505450015574e-05,
"loss": 1.636038064956665,
"step": 4686
},
{
"epoch": 0.9781950965049556,
"grad_norm": 0.278155118227005,
"learning_rate": 2.640921831205232e-05,
"loss": 1.5296932458877563,
"step": 4688
},
{
"epoch": 0.9786124152321335,
"grad_norm": 0.17809408903121948,
"learning_rate": 2.591093117408907e-05,
"loss": 1.530478596687317,
"step": 4690
},
{
"epoch": 0.9790297339593115,
"grad_norm": 0.20853684842586517,
"learning_rate": 2.541264403612582e-05,
"loss": 1.6099060773849487,
"step": 4692
},
{
"epoch": 0.9794470526864894,
"grad_norm": 0.1854981631040573,
"learning_rate": 2.4914356898162566e-05,
"loss": 1.6548269987106323,
"step": 4694
},
{
"epoch": 0.9798643714136672,
"grad_norm": 0.19578847289085388,
"learning_rate": 2.4416069760199317e-05,
"loss": 1.583350419998169,
"step": 4696
},
{
"epoch": 0.9802816901408451,
"grad_norm": 0.1766752004623413,
"learning_rate": 2.3917782622236065e-05,
"loss": 1.7396546602249146,
"step": 4698
},
{
"epoch": 0.980699008868023,
"grad_norm": 0.1721375733613968,
"learning_rate": 2.3419495484272813e-05,
"loss": 1.714516043663025,
"step": 4700
},
{
"epoch": 0.9811163275952008,
"grad_norm": 0.18396082520484924,
"learning_rate": 2.292120834630956e-05,
"loss": 1.7097505331039429,
"step": 4702
},
{
"epoch": 0.9815336463223787,
"grad_norm": 0.18421031534671783,
"learning_rate": 2.2422921208346313e-05,
"loss": 1.6922452449798584,
"step": 4704
},
{
"epoch": 0.9819509650495566,
"grad_norm": 0.19681572914123535,
"learning_rate": 2.192463407038306e-05,
"loss": 1.4659477472305298,
"step": 4706
},
{
"epoch": 0.9823682837767345,
"grad_norm": 0.1872592568397522,
"learning_rate": 2.142634693241981e-05,
"loss": 1.5210742950439453,
"step": 4708
},
{
"epoch": 0.9827856025039123,
"grad_norm": 0.17335426807403564,
"learning_rate": 2.0928059794456557e-05,
"loss": 1.6613713502883911,
"step": 4710
},
{
"epoch": 0.9832029212310902,
"grad_norm": 0.19143570959568024,
"learning_rate": 2.0429772656493305e-05,
"loss": 1.6417210102081299,
"step": 4712
},
{
"epoch": 0.9836202399582681,
"grad_norm": 0.2068614661693573,
"learning_rate": 1.9931485518530056e-05,
"loss": 1.5639116764068604,
"step": 4714
},
{
"epoch": 0.984037558685446,
"grad_norm": 0.21536195278167725,
"learning_rate": 1.9433198380566804e-05,
"loss": 1.5828635692596436,
"step": 4716
},
{
"epoch": 0.9844548774126239,
"grad_norm": 0.2028193473815918,
"learning_rate": 1.8934911242603552e-05,
"loss": 1.5901093482971191,
"step": 4718
},
{
"epoch": 0.9848721961398018,
"grad_norm": 0.1817675083875656,
"learning_rate": 1.84366241046403e-05,
"loss": 1.6806503534317017,
"step": 4720
},
{
"epoch": 0.9852895148669797,
"grad_norm": 0.2189396470785141,
"learning_rate": 1.7938336966677048e-05,
"loss": 1.6864625215530396,
"step": 4722
},
{
"epoch": 0.9857068335941576,
"grad_norm": 0.20859068632125854,
"learning_rate": 1.7440049828713796e-05,
"loss": 1.5868487358093262,
"step": 4724
},
{
"epoch": 0.9861241523213354,
"grad_norm": 0.20414243638515472,
"learning_rate": 1.6941762690750544e-05,
"loss": 1.6594513654708862,
"step": 4726
},
{
"epoch": 0.9865414710485133,
"grad_norm": 0.20474189519882202,
"learning_rate": 1.6443475552787292e-05,
"loss": 1.6990089416503906,
"step": 4728
},
{
"epoch": 0.9869587897756912,
"grad_norm": 0.21251992881298065,
"learning_rate": 1.5945188414824044e-05,
"loss": 1.6899701356887817,
"step": 4730
},
{
"epoch": 0.987376108502869,
"grad_norm": 0.1791708767414093,
"learning_rate": 1.544690127686079e-05,
"loss": 1.6678680181503296,
"step": 4732
},
{
"epoch": 0.9877934272300469,
"grad_norm": 0.1813337206840515,
"learning_rate": 1.494861413889754e-05,
"loss": 1.6770246028900146,
"step": 4734
},
{
"epoch": 0.9882107459572248,
"grad_norm": 0.17830485105514526,
"learning_rate": 1.4450327000934288e-05,
"loss": 1.5783902406692505,
"step": 4736
},
{
"epoch": 0.9886280646844027,
"grad_norm": 0.19217655062675476,
"learning_rate": 1.3952039862971036e-05,
"loss": 1.5981498956680298,
"step": 4738
},
{
"epoch": 0.9890453834115805,
"grad_norm": 0.18971745669841766,
"learning_rate": 1.3453752725007787e-05,
"loss": 1.4866902828216553,
"step": 4740
},
{
"epoch": 0.9894627021387585,
"grad_norm": 0.1601237654685974,
"learning_rate": 1.2955465587044535e-05,
"loss": 1.5577491521835327,
"step": 4742
},
{
"epoch": 0.9898800208659364,
"grad_norm": 0.18725444376468658,
"learning_rate": 1.2457178449081283e-05,
"loss": 1.666822910308838,
"step": 4744
},
{
"epoch": 0.9902973395931143,
"grad_norm": 0.20456543564796448,
"learning_rate": 1.1958891311118033e-05,
"loss": 1.5505273342132568,
"step": 4746
},
{
"epoch": 0.9907146583202922,
"grad_norm": 0.202808678150177,
"learning_rate": 1.146060417315478e-05,
"loss": 1.501298189163208,
"step": 4748
},
{
"epoch": 0.99113197704747,
"grad_norm": 0.23894049227237701,
"learning_rate": 1.096231703519153e-05,
"loss": 1.5431385040283203,
"step": 4750
},
{
"epoch": 0.9915492957746479,
"grad_norm": 0.18152473866939545,
"learning_rate": 1.0464029897228278e-05,
"loss": 1.6673369407653809,
"step": 4752
},
{
"epoch": 0.9919666145018258,
"grad_norm": 0.1959715187549591,
"learning_rate": 9.965742759265028e-06,
"loss": 1.6675604581832886,
"step": 4754
},
{
"epoch": 0.9923839332290036,
"grad_norm": 0.18566332757472992,
"learning_rate": 9.467455621301776e-06,
"loss": 1.5886743068695068,
"step": 4756
},
{
"epoch": 0.9928012519561815,
"grad_norm": 0.16858698427677155,
"learning_rate": 8.969168483338524e-06,
"loss": 1.606601357460022,
"step": 4758
},
{
"epoch": 0.9932185706833594,
"grad_norm": 0.20232905447483063,
"learning_rate": 8.470881345375272e-06,
"loss": 1.4711308479309082,
"step": 4760
},
{
"epoch": 0.9936358894105373,
"grad_norm": 0.19926717877388,
"learning_rate": 7.972594207412022e-06,
"loss": 1.5273923873901367,
"step": 4762
},
{
"epoch": 0.9940532081377151,
"grad_norm": 0.19627489149570465,
"learning_rate": 7.47430706944877e-06,
"loss": 1.7496665716171265,
"step": 4764
},
{
"epoch": 0.994470526864893,
"grad_norm": 0.2072584182024002,
"learning_rate": 6.976019931485518e-06,
"loss": 1.6775583028793335,
"step": 4766
},
{
"epoch": 0.994887845592071,
"grad_norm": 0.1914246678352356,
"learning_rate": 6.4777327935222675e-06,
"loss": 1.6985867023468018,
"step": 4768
},
{
"epoch": 0.9953051643192489,
"grad_norm": 0.21064823865890503,
"learning_rate": 5.979445655559016e-06,
"loss": 1.558905839920044,
"step": 4770
},
{
"epoch": 0.9957224830464267,
"grad_norm": 0.16335240006446838,
"learning_rate": 5.481158517595765e-06,
"loss": 1.5973092317581177,
"step": 4772
},
{
"epoch": 0.9961398017736046,
"grad_norm": 0.18516434729099274,
"learning_rate": 4.982871379632514e-06,
"loss": 1.648966908454895,
"step": 4774
},
{
"epoch": 0.9965571205007825,
"grad_norm": 0.1826256960630417,
"learning_rate": 4.484584241669262e-06,
"loss": 1.5638716220855713,
"step": 4776
},
{
"epoch": 0.9969744392279604,
"grad_norm": 0.16833841800689697,
"learning_rate": 3.986297103706011e-06,
"loss": 1.7621185779571533,
"step": 4778
},
{
"epoch": 0.9973917579551382,
"grad_norm": 0.1825326681137085,
"learning_rate": 3.488009965742759e-06,
"loss": 1.8345801830291748,
"step": 4780
},
{
"epoch": 0.9978090766823161,
"grad_norm": 0.21139578521251678,
"learning_rate": 2.989722827779508e-06,
"loss": 1.62010657787323,
"step": 4782
},
{
"epoch": 0.998226395409494,
"grad_norm": 0.1819504350423813,
"learning_rate": 2.491435689816257e-06,
"loss": 1.7124700546264648,
"step": 4784
},
{
"epoch": 0.9986437141366719,
"grad_norm": 0.1880815476179123,
"learning_rate": 1.9931485518530054e-06,
"loss": 1.5860264301300049,
"step": 4786
},
{
"epoch": 0.9990610328638497,
"grad_norm": 0.19340446591377258,
"learning_rate": 1.494861413889754e-06,
"loss": 1.6285982131958008,
"step": 4788
},
{
"epoch": 0.9994783515910276,
"grad_norm": 0.2084585726261139,
"learning_rate": 9.965742759265027e-07,
"loss": 1.490464687347412,
"step": 4790
},
{
"epoch": 0.9998956703182055,
"grad_norm": 0.23030996322631836,
"learning_rate": 4.982871379632514e-07,
"loss": 1.6368834972381592,
"step": 4792
},
{
"epoch": 1.0,
"eval_loss": 1.6273930072784424,
"eval_runtime": 25.5758,
"eval_samples_per_second": 60.604,
"eval_steps_per_second": 7.585,
"step": 4793
}
],
"logging_steps": 2,
"max_steps": 4793,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 480,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.0934343323708813e+18,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}