Safetensors
English
bert
hgbert-small / trainer_state.json
davidmezzetti's picture
Add model
117b80c
Raw
History Blame Contribute Delete
235 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.0,
"eval_steps": 500,
"global_step": 632616,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.003161475523856494,
"grad_norm": 1.8537416458129883,
"learning_rate": 1.9960000000000002e-05,
"loss": 9.1530712890625,
"step": 500
},
{
"epoch": 0.006322951047712988,
"grad_norm": 0.54729163646698,
"learning_rate": 3.9960000000000004e-05,
"loss": 7.00430859375,
"step": 1000
},
{
"epoch": 0.009484426571569483,
"grad_norm": 0.6920751929283142,
"learning_rate": 5.996e-05,
"loss": 6.60928759765625,
"step": 1500
},
{
"epoch": 0.012645902095425977,
"grad_norm": 0.7651448845863342,
"learning_rate": 7.996e-05,
"loss": 6.4331806640625,
"step": 2000
},
{
"epoch": 0.01580737761928247,
"grad_norm": 0.9238298535346985,
"learning_rate": 9.996000000000001e-05,
"loss": 6.3314375,
"step": 2500
},
{
"epoch": 0.018968853143138967,
"grad_norm": 0.9256702661514282,
"learning_rate": 9.992080823213505e-05,
"loss": 6.24858642578125,
"step": 3000
},
{
"epoch": 0.02213032866699546,
"grad_norm": 1.0555678606033325,
"learning_rate": 9.984145776333247e-05,
"loss": 6.17907470703125,
"step": 3500
},
{
"epoch": 0.025291804190851953,
"grad_norm": 0.8636297583580017,
"learning_rate": 9.976210729452991e-05,
"loss": 6.14103759765625,
"step": 4000
},
{
"epoch": 0.02845327971470845,
"grad_norm": 0.8230151534080505,
"learning_rate": 9.968275682572733e-05,
"loss": 6.1013662109375,
"step": 4500
},
{
"epoch": 0.03161475523856494,
"grad_norm": 0.7868441343307495,
"learning_rate": 9.960340635692477e-05,
"loss": 6.07035888671875,
"step": 5000
},
{
"epoch": 0.03477623076242144,
"grad_norm": 0.8165586590766907,
"learning_rate": 9.952405588812219e-05,
"loss": 6.04968603515625,
"step": 5500
},
{
"epoch": 0.037937706286277934,
"grad_norm": 0.9183712005615234,
"learning_rate": 9.944470541931962e-05,
"loss": 6.03665771484375,
"step": 6000
},
{
"epoch": 0.04109918181013442,
"grad_norm": 1.149253249168396,
"learning_rate": 9.936535495051705e-05,
"loss": 6.01305224609375,
"step": 6500
},
{
"epoch": 0.04426065733399092,
"grad_norm": 1.086187481880188,
"learning_rate": 9.928600448171448e-05,
"loss": 6.004318359375,
"step": 7000
},
{
"epoch": 0.04742213285784742,
"grad_norm": 0.8483502864837646,
"learning_rate": 9.920665401291191e-05,
"loss": 5.9925888671875,
"step": 7500
},
{
"epoch": 0.05058360838170391,
"grad_norm": 1.0872936248779297,
"learning_rate": 9.912730354410934e-05,
"loss": 5.98390478515625,
"step": 8000
},
{
"epoch": 0.053745083905560403,
"grad_norm": 1.1740615367889404,
"learning_rate": 9.904795307530677e-05,
"loss": 5.9712822265625,
"step": 8500
},
{
"epoch": 0.0569065594294169,
"grad_norm": 1.114344835281372,
"learning_rate": 9.89686026065042e-05,
"loss": 5.96275439453125,
"step": 9000
},
{
"epoch": 0.06006803495327339,
"grad_norm": 0.9349523186683655,
"learning_rate": 9.888925213770164e-05,
"loss": 5.95448583984375,
"step": 9500
},
{
"epoch": 0.06322951047712988,
"grad_norm": 1.0393985509872437,
"learning_rate": 9.880990166889906e-05,
"loss": 5.95046044921875,
"step": 10000
},
{
"epoch": 0.06639098600098638,
"grad_norm": 1.039152979850769,
"learning_rate": 9.87305512000965e-05,
"loss": 5.94594189453125,
"step": 10500
},
{
"epoch": 0.06955246152484287,
"grad_norm": 1.0210031270980835,
"learning_rate": 9.865120073129392e-05,
"loss": 5.939291015625,
"step": 11000
},
{
"epoch": 0.07271393704869937,
"grad_norm": 1.0634866952896118,
"learning_rate": 9.857185026249136e-05,
"loss": 5.93764794921875,
"step": 11500
},
{
"epoch": 0.07587541257255587,
"grad_norm": 1.246435523033142,
"learning_rate": 9.849249979368878e-05,
"loss": 5.92595458984375,
"step": 12000
},
{
"epoch": 0.07903688809641236,
"grad_norm": 1.058523416519165,
"learning_rate": 9.841314932488622e-05,
"loss": 5.91917236328125,
"step": 12500
},
{
"epoch": 0.08219836362026885,
"grad_norm": 1.0644090175628662,
"learning_rate": 9.833379885608364e-05,
"loss": 5.91900244140625,
"step": 13000
},
{
"epoch": 0.08535983914412534,
"grad_norm": 1.0447696447372437,
"learning_rate": 9.825444838728108e-05,
"loss": 5.9045927734375,
"step": 13500
},
{
"epoch": 0.08852131466798184,
"grad_norm": 1.029859185218811,
"learning_rate": 9.81750979184785e-05,
"loss": 5.9073916015625,
"step": 14000
},
{
"epoch": 0.09168279019183834,
"grad_norm": 1.0749905109405518,
"learning_rate": 9.809574744967594e-05,
"loss": 5.90510595703125,
"step": 14500
},
{
"epoch": 0.09484426571569483,
"grad_norm": 1.0689607858657837,
"learning_rate": 9.801639698087338e-05,
"loss": 5.89893310546875,
"step": 15000
},
{
"epoch": 0.09800574123955132,
"grad_norm": 1.0687636137008667,
"learning_rate": 9.79370465120708e-05,
"loss": 5.89199267578125,
"step": 15500
},
{
"epoch": 0.10116721676340781,
"grad_norm": 1.2068299055099487,
"learning_rate": 9.785769604326824e-05,
"loss": 5.8961171875,
"step": 16000
},
{
"epoch": 0.10432869228726431,
"grad_norm": 1.14532470703125,
"learning_rate": 9.777834557446566e-05,
"loss": 5.891658203125,
"step": 16500
},
{
"epoch": 0.10749016781112081,
"grad_norm": 1.0716787576675415,
"learning_rate": 9.76989951056631e-05,
"loss": 5.8866650390625,
"step": 17000
},
{
"epoch": 0.1106516433349773,
"grad_norm": 0.9980911612510681,
"learning_rate": 9.761964463686052e-05,
"loss": 5.88161669921875,
"step": 17500
},
{
"epoch": 0.1138131188588338,
"grad_norm": 1.249799132347107,
"learning_rate": 9.754029416805795e-05,
"loss": 5.88307275390625,
"step": 18000
},
{
"epoch": 0.11697459438269028,
"grad_norm": 1.1821871995925903,
"learning_rate": 9.746094369925538e-05,
"loss": 5.87755859375,
"step": 18500
},
{
"epoch": 0.12013606990654678,
"grad_norm": 1.237335443496704,
"learning_rate": 9.73815932304528e-05,
"loss": 5.87326708984375,
"step": 19000
},
{
"epoch": 0.12329754543040328,
"grad_norm": 1.3327748775482178,
"learning_rate": 9.730224276165024e-05,
"loss": 5.86807421875,
"step": 19500
},
{
"epoch": 0.12645902095425976,
"grad_norm": 1.145980715751648,
"learning_rate": 9.722289229284767e-05,
"loss": 5.8705751953125,
"step": 20000
},
{
"epoch": 0.12962049647811627,
"grad_norm": 1.3894362449645996,
"learning_rate": 9.71435418240451e-05,
"loss": 5.8654111328125,
"step": 20500
},
{
"epoch": 0.13278197200197275,
"grad_norm": 1.4751064777374268,
"learning_rate": 9.706419135524253e-05,
"loss": 5.86437841796875,
"step": 21000
},
{
"epoch": 0.13594344752582926,
"grad_norm": 1.2660388946533203,
"learning_rate": 9.698484088643997e-05,
"loss": 5.86302978515625,
"step": 21500
},
{
"epoch": 0.13910492304968575,
"grad_norm": 1.2101713418960571,
"learning_rate": 9.690549041763739e-05,
"loss": 5.86247314453125,
"step": 22000
},
{
"epoch": 0.14226639857354223,
"grad_norm": 1.0292099714279175,
"learning_rate": 9.682613994883483e-05,
"loss": 5.85936279296875,
"step": 22500
},
{
"epoch": 0.14542787409739874,
"grad_norm": 1.0616521835327148,
"learning_rate": 9.674678948003225e-05,
"loss": 5.85700341796875,
"step": 23000
},
{
"epoch": 0.14858934962125522,
"grad_norm": 1.286318302154541,
"learning_rate": 9.666743901122969e-05,
"loss": 5.85010888671875,
"step": 23500
},
{
"epoch": 0.15175082514511173,
"grad_norm": 1.1704819202423096,
"learning_rate": 9.658808854242711e-05,
"loss": 5.84995166015625,
"step": 24000
},
{
"epoch": 0.15491230066896822,
"grad_norm": 1.3493387699127197,
"learning_rate": 9.650873807362453e-05,
"loss": 5.851232421875,
"step": 24500
},
{
"epoch": 0.15807377619282473,
"grad_norm": 1.234583854675293,
"learning_rate": 9.642938760482197e-05,
"loss": 5.8467060546875,
"step": 25000
},
{
"epoch": 0.1612352517166812,
"grad_norm": 1.0729691982269287,
"learning_rate": 9.63500371360194e-05,
"loss": 5.8481142578125,
"step": 25500
},
{
"epoch": 0.1643967272405377,
"grad_norm": 1.217193365097046,
"learning_rate": 9.627068666721683e-05,
"loss": 5.8437177734375,
"step": 26000
},
{
"epoch": 0.1675582027643942,
"grad_norm": 1.2050867080688477,
"learning_rate": 9.619133619841426e-05,
"loss": 5.83731787109375,
"step": 26500
},
{
"epoch": 0.1707196782882507,
"grad_norm": 1.091496229171753,
"learning_rate": 9.61119857296117e-05,
"loss": 5.84143798828125,
"step": 27000
},
{
"epoch": 0.1738811538121072,
"grad_norm": 1.3209389448165894,
"learning_rate": 9.603263526080912e-05,
"loss": 5.83608935546875,
"step": 27500
},
{
"epoch": 0.17704262933596368,
"grad_norm": 1.1381261348724365,
"learning_rate": 9.595328479200655e-05,
"loss": 5.8344013671875,
"step": 28000
},
{
"epoch": 0.18020410485982016,
"grad_norm": 1.147188663482666,
"learning_rate": 9.587393432320398e-05,
"loss": 5.82754833984375,
"step": 28500
},
{
"epoch": 0.18336558038367667,
"grad_norm": 1.1861779689788818,
"learning_rate": 9.579458385440142e-05,
"loss": 5.83367724609375,
"step": 29000
},
{
"epoch": 0.18652705590753316,
"grad_norm": 1.1430096626281738,
"learning_rate": 9.571523338559885e-05,
"loss": 5.83239013671875,
"step": 29500
},
{
"epoch": 0.18968853143138967,
"grad_norm": 1.1952162981033325,
"learning_rate": 9.563588291679628e-05,
"loss": 5.83822705078125,
"step": 30000
},
{
"epoch": 0.19285000695524615,
"grad_norm": 1.2185235023498535,
"learning_rate": 9.555653244799371e-05,
"loss": 5.8323056640625,
"step": 30500
},
{
"epoch": 0.19601148247910263,
"grad_norm": 1.0978896617889404,
"learning_rate": 9.547718197919114e-05,
"loss": 5.82861962890625,
"step": 31000
},
{
"epoch": 0.19917295800295914,
"grad_norm": 1.1465359926223755,
"learning_rate": 9.539783151038857e-05,
"loss": 5.82329345703125,
"step": 31500
},
{
"epoch": 0.20233443352681563,
"grad_norm": 1.2042582035064697,
"learning_rate": 9.5318481041586e-05,
"loss": 5.82667724609375,
"step": 32000
},
{
"epoch": 0.20549590905067214,
"grad_norm": 1.1498785018920898,
"learning_rate": 9.523913057278343e-05,
"loss": 5.82408740234375,
"step": 32500
},
{
"epoch": 0.20865738457452862,
"grad_norm": 1.2287983894348145,
"learning_rate": 9.515978010398086e-05,
"loss": 5.82123583984375,
"step": 33000
},
{
"epoch": 0.21181886009838513,
"grad_norm": 1.2773981094360352,
"learning_rate": 9.50804296351783e-05,
"loss": 5.82193798828125,
"step": 33500
},
{
"epoch": 0.21498033562224161,
"grad_norm": 1.1635549068450928,
"learning_rate": 9.500107916637572e-05,
"loss": 5.8212890625,
"step": 34000
},
{
"epoch": 0.2181418111460981,
"grad_norm": 1.0567307472229004,
"learning_rate": 9.492172869757316e-05,
"loss": 5.81920849609375,
"step": 34500
},
{
"epoch": 0.2213032866699546,
"grad_norm": 1.167801856994629,
"learning_rate": 9.484237822877058e-05,
"loss": 5.81664501953125,
"step": 35000
},
{
"epoch": 0.2244647621938111,
"grad_norm": 1.2146143913269043,
"learning_rate": 9.476302775996802e-05,
"loss": 5.81528125,
"step": 35500
},
{
"epoch": 0.2276262377176676,
"grad_norm": 1.1864373683929443,
"learning_rate": 9.468367729116544e-05,
"loss": 5.8171875,
"step": 36000
},
{
"epoch": 0.23078771324152408,
"grad_norm": 1.4294815063476562,
"learning_rate": 9.460432682236288e-05,
"loss": 5.815041015625,
"step": 36500
},
{
"epoch": 0.23394918876538057,
"grad_norm": 1.1029082536697388,
"learning_rate": 9.45249763535603e-05,
"loss": 5.8178857421875,
"step": 37000
},
{
"epoch": 0.23711066428923708,
"grad_norm": 1.212059736251831,
"learning_rate": 9.444562588475773e-05,
"loss": 5.80740771484375,
"step": 37500
},
{
"epoch": 0.24027213981309356,
"grad_norm": 1.1588537693023682,
"learning_rate": 9.436627541595516e-05,
"loss": 5.80828271484375,
"step": 38000
},
{
"epoch": 0.24343361533695007,
"grad_norm": 1.3309358358383179,
"learning_rate": 9.428692494715259e-05,
"loss": 5.809466796875,
"step": 38500
},
{
"epoch": 0.24659509086080655,
"grad_norm": 1.2989411354064941,
"learning_rate": 9.420757447835002e-05,
"loss": 5.80472607421875,
"step": 39000
},
{
"epoch": 0.24975656638466304,
"grad_norm": 1.3860821723937988,
"learning_rate": 9.412822400954745e-05,
"loss": 5.81056591796875,
"step": 39500
},
{
"epoch": 0.2529180419085195,
"grad_norm": 1.1357406377792358,
"learning_rate": 9.404887354074488e-05,
"loss": 5.8078837890625,
"step": 40000
},
{
"epoch": 0.25607951743237606,
"grad_norm": 1.1922794580459595,
"learning_rate": 9.396952307194231e-05,
"loss": 5.80509521484375,
"step": 40500
},
{
"epoch": 0.25924099295623254,
"grad_norm": 1.3267223834991455,
"learning_rate": 9.389017260313975e-05,
"loss": 5.80380859375,
"step": 41000
},
{
"epoch": 0.262402468480089,
"grad_norm": 1.0675228834152222,
"learning_rate": 9.381082213433717e-05,
"loss": 5.8023525390625,
"step": 41500
},
{
"epoch": 0.2655639440039455,
"grad_norm": 1.1232720613479614,
"learning_rate": 9.37314716655346e-05,
"loss": 5.802138671875,
"step": 42000
},
{
"epoch": 0.268725419527802,
"grad_norm": 1.2875728607177734,
"learning_rate": 9.365212119673203e-05,
"loss": 5.79818896484375,
"step": 42500
},
{
"epoch": 0.27188689505165853,
"grad_norm": 1.2377989292144775,
"learning_rate": 9.357277072792945e-05,
"loss": 5.79720556640625,
"step": 43000
},
{
"epoch": 0.275048370575515,
"grad_norm": 1.3427847623825073,
"learning_rate": 9.349342025912689e-05,
"loss": 5.804609375,
"step": 43500
},
{
"epoch": 0.2782098460993715,
"grad_norm": 1.2763382196426392,
"learning_rate": 9.341406979032433e-05,
"loss": 5.8009287109375,
"step": 44000
},
{
"epoch": 0.281371321623228,
"grad_norm": 1.4750946760177612,
"learning_rate": 9.333471932152175e-05,
"loss": 5.7976279296875,
"step": 44500
},
{
"epoch": 0.28453279714708446,
"grad_norm": 1.181031346321106,
"learning_rate": 9.325536885271919e-05,
"loss": 5.79604296875,
"step": 45000
},
{
"epoch": 0.287694272670941,
"grad_norm": 1.2075753211975098,
"learning_rate": 9.317601838391661e-05,
"loss": 5.7934462890625,
"step": 45500
},
{
"epoch": 0.2908557481947975,
"grad_norm": 1.2902746200561523,
"learning_rate": 9.309666791511405e-05,
"loss": 5.7988681640625,
"step": 46000
},
{
"epoch": 0.29401722371865396,
"grad_norm": 1.1361738443374634,
"learning_rate": 9.301731744631149e-05,
"loss": 5.7902841796875,
"step": 46500
},
{
"epoch": 0.29717869924251045,
"grad_norm": 1.3249701261520386,
"learning_rate": 9.293796697750891e-05,
"loss": 5.79316845703125,
"step": 47000
},
{
"epoch": 0.300340174766367,
"grad_norm": 1.3028849363327026,
"learning_rate": 9.285861650870635e-05,
"loss": 5.7922802734375,
"step": 47500
},
{
"epoch": 0.30350165029022347,
"grad_norm": 1.2621126174926758,
"learning_rate": 9.277926603990377e-05,
"loss": 5.7919111328125,
"step": 48000
},
{
"epoch": 0.30666312581407995,
"grad_norm": 1.2226518392562866,
"learning_rate": 9.269991557110121e-05,
"loss": 5.79368310546875,
"step": 48500
},
{
"epoch": 0.30982460133793643,
"grad_norm": 1.246754765510559,
"learning_rate": 9.262056510229863e-05,
"loss": 5.78959814453125,
"step": 49000
},
{
"epoch": 0.3129860768617929,
"grad_norm": 1.2730836868286133,
"learning_rate": 9.254121463349606e-05,
"loss": 5.789521484375,
"step": 49500
},
{
"epoch": 0.31614755238564946,
"grad_norm": 1.1967133283615112,
"learning_rate": 9.246186416469349e-05,
"loss": 5.7874521484375,
"step": 50000
},
{
"epoch": 0.31930902790950594,
"grad_norm": 1.2050294876098633,
"learning_rate": 9.238251369589092e-05,
"loss": 5.78395458984375,
"step": 50500
},
{
"epoch": 0.3224705034333624,
"grad_norm": 1.2638012170791626,
"learning_rate": 9.230316322708835e-05,
"loss": 5.788017578125,
"step": 51000
},
{
"epoch": 0.3256319789572189,
"grad_norm": 1.363849401473999,
"learning_rate": 9.222381275828578e-05,
"loss": 5.784681640625,
"step": 51500
},
{
"epoch": 0.3287934544810754,
"grad_norm": 1.2883983850479126,
"learning_rate": 9.214446228948321e-05,
"loss": 5.78190576171875,
"step": 52000
},
{
"epoch": 0.3319549300049319,
"grad_norm": 1.2847943305969238,
"learning_rate": 9.206511182068064e-05,
"loss": 5.78243359375,
"step": 52500
},
{
"epoch": 0.3351164055287884,
"grad_norm": 1.1165565252304077,
"learning_rate": 9.198576135187808e-05,
"loss": 5.778548828125,
"step": 53000
},
{
"epoch": 0.3382778810526449,
"grad_norm": 1.1766024827957153,
"learning_rate": 9.19064108830755e-05,
"loss": 5.7818505859375,
"step": 53500
},
{
"epoch": 0.3414393565765014,
"grad_norm": 1.330378770828247,
"learning_rate": 9.182706041427294e-05,
"loss": 5.78132470703125,
"step": 54000
},
{
"epoch": 0.34460083210035786,
"grad_norm": 1.1629866361618042,
"learning_rate": 9.174770994547036e-05,
"loss": 5.7811533203125,
"step": 54500
},
{
"epoch": 0.3477623076242144,
"grad_norm": 1.1440378427505493,
"learning_rate": 9.16683594766678e-05,
"loss": 5.775611328125,
"step": 55000
},
{
"epoch": 0.3509237831480709,
"grad_norm": 1.2620115280151367,
"learning_rate": 9.158900900786522e-05,
"loss": 5.78110205078125,
"step": 55500
},
{
"epoch": 0.35408525867192736,
"grad_norm": 1.2070493698120117,
"learning_rate": 9.150965853906264e-05,
"loss": 5.77914501953125,
"step": 56000
},
{
"epoch": 0.35724673419578384,
"grad_norm": 1.132567286491394,
"learning_rate": 9.143030807026008e-05,
"loss": 5.77395751953125,
"step": 56500
},
{
"epoch": 0.3604082097196403,
"grad_norm": 1.1880218982696533,
"learning_rate": 9.13509576014575e-05,
"loss": 5.77958642578125,
"step": 57000
},
{
"epoch": 0.36356968524349687,
"grad_norm": 1.2243353128433228,
"learning_rate": 9.127160713265494e-05,
"loss": 5.774345703125,
"step": 57500
},
{
"epoch": 0.36673116076735335,
"grad_norm": 1.2004894018173218,
"learning_rate": 9.119225666385237e-05,
"loss": 5.7721181640625,
"step": 58000
},
{
"epoch": 0.36989263629120983,
"grad_norm": 1.1901127099990845,
"learning_rate": 9.11129061950498e-05,
"loss": 5.7730234375,
"step": 58500
},
{
"epoch": 0.3730541118150663,
"grad_norm": 1.278987169265747,
"learning_rate": 9.103355572624723e-05,
"loss": 5.7787724609375,
"step": 59000
},
{
"epoch": 0.3762155873389228,
"grad_norm": 1.4656606912612915,
"learning_rate": 9.095420525744466e-05,
"loss": 5.77456005859375,
"step": 59500
},
{
"epoch": 0.37937706286277934,
"grad_norm": 1.321642279624939,
"learning_rate": 9.087485478864209e-05,
"loss": 5.7724130859375,
"step": 60000
},
{
"epoch": 0.3825385383866358,
"grad_norm": 1.3238755464553833,
"learning_rate": 9.079550431983952e-05,
"loss": 5.771455078125,
"step": 60500
},
{
"epoch": 0.3857000139104923,
"grad_norm": 1.2099475860595703,
"learning_rate": 9.071615385103696e-05,
"loss": 5.77299609375,
"step": 61000
},
{
"epoch": 0.3888614894343488,
"grad_norm": 1.1627109050750732,
"learning_rate": 9.063680338223439e-05,
"loss": 5.76878466796875,
"step": 61500
},
{
"epoch": 0.39202296495820527,
"grad_norm": 1.376214623451233,
"learning_rate": 9.055745291343182e-05,
"loss": 5.77015625,
"step": 62000
},
{
"epoch": 0.3951844404820618,
"grad_norm": 1.1330876350402832,
"learning_rate": 9.047810244462925e-05,
"loss": 5.76786279296875,
"step": 62500
},
{
"epoch": 0.3983459160059183,
"grad_norm": 1.198279857635498,
"learning_rate": 9.039875197582668e-05,
"loss": 5.769275390625,
"step": 63000
},
{
"epoch": 0.40150739152977477,
"grad_norm": 1.2219595909118652,
"learning_rate": 9.031940150702411e-05,
"loss": 5.77046728515625,
"step": 63500
},
{
"epoch": 0.40466886705363125,
"grad_norm": 1.1769604682922363,
"learning_rate": 9.024005103822154e-05,
"loss": 5.76344482421875,
"step": 64000
},
{
"epoch": 0.4078303425774878,
"grad_norm": 1.142532467842102,
"learning_rate": 9.016070056941897e-05,
"loss": 5.76774951171875,
"step": 64500
},
{
"epoch": 0.4109918181013443,
"grad_norm": 1.2945693731307983,
"learning_rate": 9.00813501006164e-05,
"loss": 5.76451708984375,
"step": 65000
},
{
"epoch": 0.41415329362520076,
"grad_norm": 1.1853876113891602,
"learning_rate": 9.000199963181383e-05,
"loss": 5.76903173828125,
"step": 65500
},
{
"epoch": 0.41731476914905724,
"grad_norm": 1.2533129453659058,
"learning_rate": 8.992264916301127e-05,
"loss": 5.7662158203125,
"step": 66000
},
{
"epoch": 0.4204762446729137,
"grad_norm": 1.251891851425171,
"learning_rate": 8.984329869420869e-05,
"loss": 5.7630029296875,
"step": 66500
},
{
"epoch": 0.42363772019677026,
"grad_norm": 1.2563832998275757,
"learning_rate": 8.976394822540613e-05,
"loss": 5.76238818359375,
"step": 67000
},
{
"epoch": 0.42679919572062675,
"grad_norm": 1.2545146942138672,
"learning_rate": 8.968459775660355e-05,
"loss": 5.7638017578125,
"step": 67500
},
{
"epoch": 0.42996067124448323,
"grad_norm": 1.4037646055221558,
"learning_rate": 8.960524728780097e-05,
"loss": 5.76652783203125,
"step": 68000
},
{
"epoch": 0.4331221467683397,
"grad_norm": 1.215163230895996,
"learning_rate": 8.952589681899841e-05,
"loss": 5.766744140625,
"step": 68500
},
{
"epoch": 0.4362836222921962,
"grad_norm": 1.3993630409240723,
"learning_rate": 8.944654635019584e-05,
"loss": 5.76658984375,
"step": 69000
},
{
"epoch": 0.43944509781605273,
"grad_norm": 1.3046010732650757,
"learning_rate": 8.936719588139327e-05,
"loss": 5.758736328125,
"step": 69500
},
{
"epoch": 0.4426065733399092,
"grad_norm": 1.3707127571105957,
"learning_rate": 8.92878454125907e-05,
"loss": 5.762345703125,
"step": 70000
},
{
"epoch": 0.4457680488637657,
"grad_norm": 1.2552387714385986,
"learning_rate": 8.920849494378813e-05,
"loss": 5.76258935546875,
"step": 70500
},
{
"epoch": 0.4489295243876222,
"grad_norm": 1.4611188173294067,
"learning_rate": 8.912914447498556e-05,
"loss": 5.76005908203125,
"step": 71000
},
{
"epoch": 0.45209099991147866,
"grad_norm": 1.4065369367599487,
"learning_rate": 8.9049794006183e-05,
"loss": 5.7294384765625,
"step": 71500
},
{
"epoch": 0.4552524754353352,
"grad_norm": 1.3063828945159912,
"learning_rate": 8.897044353738042e-05,
"loss": 5.70132763671875,
"step": 72000
},
{
"epoch": 0.4584139509591917,
"grad_norm": 1.488344669342041,
"learning_rate": 8.889109306857786e-05,
"loss": 5.6699873046875,
"step": 72500
},
{
"epoch": 0.46157542648304817,
"grad_norm": 1.5163249969482422,
"learning_rate": 8.881174259977528e-05,
"loss": 5.6535791015625,
"step": 73000
},
{
"epoch": 0.46473690200690465,
"grad_norm": 1.3599097728729248,
"learning_rate": 8.873239213097272e-05,
"loss": 5.62909814453125,
"step": 73500
},
{
"epoch": 0.46789837753076113,
"grad_norm": 1.7950043678283691,
"learning_rate": 8.865304166217014e-05,
"loss": 5.61224169921875,
"step": 74000
},
{
"epoch": 0.47105985305461767,
"grad_norm": 1.717045545578003,
"learning_rate": 8.857369119336756e-05,
"loss": 5.59769775390625,
"step": 74500
},
{
"epoch": 0.47422132857847415,
"grad_norm": 1.681662917137146,
"learning_rate": 8.8494340724565e-05,
"loss": 5.5796005859375,
"step": 75000
},
{
"epoch": 0.47738280410233064,
"grad_norm": 1.547600507736206,
"learning_rate": 8.841499025576244e-05,
"loss": 5.5633076171875,
"step": 75500
},
{
"epoch": 0.4805442796261871,
"grad_norm": 1.5499380826950073,
"learning_rate": 8.833563978695986e-05,
"loss": 5.54577880859375,
"step": 76000
},
{
"epoch": 0.4837057551500436,
"grad_norm": 1.9412866830825806,
"learning_rate": 8.82562893181573e-05,
"loss": 5.500171875,
"step": 76500
},
{
"epoch": 0.48686723067390014,
"grad_norm": 2.2403793334960938,
"learning_rate": 8.817693884935472e-05,
"loss": 5.379625,
"step": 77000
},
{
"epoch": 0.4900287061977566,
"grad_norm": 2.0986599922180176,
"learning_rate": 8.809758838055216e-05,
"loss": 5.26443017578125,
"step": 77500
},
{
"epoch": 0.4931901817216131,
"grad_norm": 2.176771879196167,
"learning_rate": 8.80182379117496e-05,
"loss": 5.1494853515625,
"step": 78000
},
{
"epoch": 0.4963516572454696,
"grad_norm": 2.7446465492248535,
"learning_rate": 8.793888744294702e-05,
"loss": 5.06522021484375,
"step": 78500
},
{
"epoch": 0.4995131327693261,
"grad_norm": 2.6227922439575195,
"learning_rate": 8.785953697414446e-05,
"loss": 4.98740478515625,
"step": 79000
},
{
"epoch": 0.5026746082931826,
"grad_norm": 2.5257158279418945,
"learning_rate": 8.778018650534188e-05,
"loss": 4.90416748046875,
"step": 79500
},
{
"epoch": 0.505836083817039,
"grad_norm": 2.6340887546539307,
"learning_rate": 8.770083603653932e-05,
"loss": 4.783,
"step": 80000
},
{
"epoch": 0.5089975593408956,
"grad_norm": 2.3913183212280273,
"learning_rate": 8.762148556773674e-05,
"loss": 4.62919970703125,
"step": 80500
},
{
"epoch": 0.5121590348647521,
"grad_norm": 2.3178250789642334,
"learning_rate": 8.754213509893417e-05,
"loss": 4.4761005859375,
"step": 81000
},
{
"epoch": 0.5153205103886086,
"grad_norm": 2.169550895690918,
"learning_rate": 8.74627846301316e-05,
"loss": 4.355857421875,
"step": 81500
},
{
"epoch": 0.5184819859124651,
"grad_norm": 2.4557347297668457,
"learning_rate": 8.738343416132903e-05,
"loss": 4.26971728515625,
"step": 82000
},
{
"epoch": 0.5216434614363216,
"grad_norm": 2.2696731090545654,
"learning_rate": 8.730408369252646e-05,
"loss": 4.192244140625,
"step": 82500
},
{
"epoch": 0.524804936960178,
"grad_norm": 2.192823886871338,
"learning_rate": 8.722473322372389e-05,
"loss": 4.128521484375,
"step": 83000
},
{
"epoch": 0.5279664124840345,
"grad_norm": 2.2004024982452393,
"learning_rate": 8.714538275492132e-05,
"loss": 4.0751279296875,
"step": 83500
},
{
"epoch": 0.531127888007891,
"grad_norm": 2.2106776237487793,
"learning_rate": 8.706603228611875e-05,
"loss": 4.02268701171875,
"step": 84000
},
{
"epoch": 0.5342893635317475,
"grad_norm": 2.1030519008636475,
"learning_rate": 8.698668181731619e-05,
"loss": 3.982795654296875,
"step": 84500
},
{
"epoch": 0.537450839055604,
"grad_norm": 2.1196064949035645,
"learning_rate": 8.690733134851361e-05,
"loss": 3.9394892578125,
"step": 85000
},
{
"epoch": 0.5406123145794606,
"grad_norm": 2.170078754425049,
"learning_rate": 8.682798087971105e-05,
"loss": 3.90093798828125,
"step": 85500
},
{
"epoch": 0.5437737901033171,
"grad_norm": 2.063465118408203,
"learning_rate": 8.674863041090847e-05,
"loss": 3.86781787109375,
"step": 86000
},
{
"epoch": 0.5469352656271735,
"grad_norm": 1.927087426185608,
"learning_rate": 8.666927994210591e-05,
"loss": 3.8325546875,
"step": 86500
},
{
"epoch": 0.55009674115103,
"grad_norm": 2.041510820388794,
"learning_rate": 8.658992947330333e-05,
"loss": 3.802294921875,
"step": 87000
},
{
"epoch": 0.5532582166748865,
"grad_norm": 2.022496461868286,
"learning_rate": 8.651057900450075e-05,
"loss": 3.769738525390625,
"step": 87500
},
{
"epoch": 0.556419692198743,
"grad_norm": 1.9612098932266235,
"learning_rate": 8.643122853569819e-05,
"loss": 3.74056591796875,
"step": 88000
},
{
"epoch": 0.5595811677225995,
"grad_norm": 2.006304979324341,
"learning_rate": 8.635187806689562e-05,
"loss": 3.70625927734375,
"step": 88500
},
{
"epoch": 0.562742643246456,
"grad_norm": 2.100412368774414,
"learning_rate": 8.627252759809305e-05,
"loss": 3.6867265625,
"step": 89000
},
{
"epoch": 0.5659041187703124,
"grad_norm": 1.9054338932037354,
"learning_rate": 8.619317712929048e-05,
"loss": 3.653140380859375,
"step": 89500
},
{
"epoch": 0.5690655942941689,
"grad_norm": 1.9641392230987549,
"learning_rate": 8.611382666048791e-05,
"loss": 3.621235107421875,
"step": 90000
},
{
"epoch": 0.5722270698180255,
"grad_norm": 1.927490472793579,
"learning_rate": 8.603447619168534e-05,
"loss": 3.597233642578125,
"step": 90500
},
{
"epoch": 0.575388545341882,
"grad_norm": 1.9694675207138062,
"learning_rate": 8.595512572288277e-05,
"loss": 3.577526611328125,
"step": 91000
},
{
"epoch": 0.5785500208657385,
"grad_norm": 1.8693190813064575,
"learning_rate": 8.58757752540802e-05,
"loss": 3.548358154296875,
"step": 91500
},
{
"epoch": 0.581711496389595,
"grad_norm": 1.9727332592010498,
"learning_rate": 8.579642478527763e-05,
"loss": 3.536473388671875,
"step": 92000
},
{
"epoch": 0.5848729719134514,
"grad_norm": 1.9331039190292358,
"learning_rate": 8.571707431647507e-05,
"loss": 3.511009033203125,
"step": 92500
},
{
"epoch": 0.5880344474373079,
"grad_norm": 1.8771967887878418,
"learning_rate": 8.56377238476725e-05,
"loss": 3.494967529296875,
"step": 93000
},
{
"epoch": 0.5911959229611644,
"grad_norm": 1.8703397512435913,
"learning_rate": 8.555837337886993e-05,
"loss": 3.47105908203125,
"step": 93500
},
{
"epoch": 0.5943573984850209,
"grad_norm": 1.6760011911392212,
"learning_rate": 8.547902291006736e-05,
"loss": 3.459435546875,
"step": 94000
},
{
"epoch": 0.5975188740088774,
"grad_norm": 1.8288774490356445,
"learning_rate": 8.53996724412648e-05,
"loss": 3.4371181640625,
"step": 94500
},
{
"epoch": 0.600680349532734,
"grad_norm": 1.920793056488037,
"learning_rate": 8.532032197246222e-05,
"loss": 3.423636474609375,
"step": 95000
},
{
"epoch": 0.6038418250565905,
"grad_norm": 1.8292603492736816,
"learning_rate": 8.524097150365965e-05,
"loss": 3.407457763671875,
"step": 95500
},
{
"epoch": 0.6070033005804469,
"grad_norm": 1.8346236944198608,
"learning_rate": 8.516162103485708e-05,
"loss": 3.401252197265625,
"step": 96000
},
{
"epoch": 0.6101647761043034,
"grad_norm": 1.7894023656845093,
"learning_rate": 8.508227056605452e-05,
"loss": 3.3778291015625,
"step": 96500
},
{
"epoch": 0.6133262516281599,
"grad_norm": 1.8625959157943726,
"learning_rate": 8.500292009725194e-05,
"loss": 3.376712890625,
"step": 97000
},
{
"epoch": 0.6164877271520164,
"grad_norm": 1.769983172416687,
"learning_rate": 8.492356962844938e-05,
"loss": 3.3507890625,
"step": 97500
},
{
"epoch": 0.6196492026758729,
"grad_norm": 1.714897871017456,
"learning_rate": 8.48442191596468e-05,
"loss": 3.34123828125,
"step": 98000
},
{
"epoch": 0.6228106781997294,
"grad_norm": 1.7576982975006104,
"learning_rate": 8.476486869084424e-05,
"loss": 3.3274443359375,
"step": 98500
},
{
"epoch": 0.6259721537235858,
"grad_norm": 1.671154499053955,
"learning_rate": 8.468551822204166e-05,
"loss": 3.318343994140625,
"step": 99000
},
{
"epoch": 0.6291336292474423,
"grad_norm": 1.7648520469665527,
"learning_rate": 8.460616775323908e-05,
"loss": 3.30464697265625,
"step": 99500
},
{
"epoch": 0.6322951047712989,
"grad_norm": 1.8779264688491821,
"learning_rate": 8.452681728443652e-05,
"loss": 3.29561572265625,
"step": 100000
},
{
"epoch": 0.6354565802951554,
"grad_norm": 1.7380156517028809,
"learning_rate": 8.444746681563395e-05,
"loss": 3.289083984375,
"step": 100500
},
{
"epoch": 0.6386180558190119,
"grad_norm": 1.7497286796569824,
"learning_rate": 8.436811634683138e-05,
"loss": 3.27760498046875,
"step": 101000
},
{
"epoch": 0.6417795313428684,
"grad_norm": 1.8647540807724,
"learning_rate": 8.42887658780288e-05,
"loss": 3.26778662109375,
"step": 101500
},
{
"epoch": 0.6449410068667248,
"grad_norm": 1.8389524221420288,
"learning_rate": 8.420941540922624e-05,
"loss": 3.255142333984375,
"step": 102000
},
{
"epoch": 0.6481024823905813,
"grad_norm": 1.7231682538986206,
"learning_rate": 8.413006494042367e-05,
"loss": 3.243255126953125,
"step": 102500
},
{
"epoch": 0.6512639579144378,
"grad_norm": 1.677039623260498,
"learning_rate": 8.40507144716211e-05,
"loss": 3.236385009765625,
"step": 103000
},
{
"epoch": 0.6544254334382943,
"grad_norm": 1.7377673387527466,
"learning_rate": 8.397136400281853e-05,
"loss": 3.229406005859375,
"step": 103500
},
{
"epoch": 0.6575869089621508,
"grad_norm": 1.73897385597229,
"learning_rate": 8.389201353401596e-05,
"loss": 3.214864013671875,
"step": 104000
},
{
"epoch": 0.6607483844860073,
"grad_norm": 1.7922133207321167,
"learning_rate": 8.381266306521339e-05,
"loss": 3.20723486328125,
"step": 104500
},
{
"epoch": 0.6639098600098639,
"grad_norm": 1.6385589838027954,
"learning_rate": 8.373331259641083e-05,
"loss": 3.196197265625,
"step": 105000
},
{
"epoch": 0.6670713355337203,
"grad_norm": 1.7266284227371216,
"learning_rate": 8.365396212760825e-05,
"loss": 3.19570947265625,
"step": 105500
},
{
"epoch": 0.6702328110575768,
"grad_norm": 1.7090206146240234,
"learning_rate": 8.357461165880567e-05,
"loss": 3.176626220703125,
"step": 106000
},
{
"epoch": 0.6733942865814333,
"grad_norm": 1.7661921977996826,
"learning_rate": 8.349526119000311e-05,
"loss": 3.171093994140625,
"step": 106500
},
{
"epoch": 0.6765557621052898,
"grad_norm": 1.666761875152588,
"learning_rate": 8.341591072120055e-05,
"loss": 3.165853515625,
"step": 107000
},
{
"epoch": 0.6797172376291463,
"grad_norm": 1.6946364641189575,
"learning_rate": 8.333656025239797e-05,
"loss": 3.162961181640625,
"step": 107500
},
{
"epoch": 0.6828787131530027,
"grad_norm": 1.7139852046966553,
"learning_rate": 8.325720978359541e-05,
"loss": 3.15451025390625,
"step": 108000
},
{
"epoch": 0.6860401886768592,
"grad_norm": 1.599587082862854,
"learning_rate": 8.317785931479283e-05,
"loss": 3.150181396484375,
"step": 108500
},
{
"epoch": 0.6892016642007157,
"grad_norm": 1.8003439903259277,
"learning_rate": 8.309850884599027e-05,
"loss": 3.140494140625,
"step": 109000
},
{
"epoch": 0.6923631397245723,
"grad_norm": 1.6548261642456055,
"learning_rate": 8.30191583771877e-05,
"loss": 3.1321083984375,
"step": 109500
},
{
"epoch": 0.6955246152484288,
"grad_norm": 1.6758365631103516,
"learning_rate": 8.293980790838513e-05,
"loss": 3.125663818359375,
"step": 110000
},
{
"epoch": 0.6986860907722853,
"grad_norm": 1.6611418724060059,
"learning_rate": 8.286045743958257e-05,
"loss": 3.11963720703125,
"step": 110500
},
{
"epoch": 0.7018475662961418,
"grad_norm": 1.6213195323944092,
"learning_rate": 8.278110697077999e-05,
"loss": 3.1164208984375,
"step": 111000
},
{
"epoch": 0.7050090418199982,
"grad_norm": 1.628892183303833,
"learning_rate": 8.270175650197743e-05,
"loss": 3.107272705078125,
"step": 111500
},
{
"epoch": 0.7081705173438547,
"grad_norm": 1.570810317993164,
"learning_rate": 8.262240603317485e-05,
"loss": 3.104776123046875,
"step": 112000
},
{
"epoch": 0.7113319928677112,
"grad_norm": 1.733945608139038,
"learning_rate": 8.254305556437228e-05,
"loss": 3.0981181640625,
"step": 112500
},
{
"epoch": 0.7144934683915677,
"grad_norm": 1.6770610809326172,
"learning_rate": 8.246370509556971e-05,
"loss": 3.09365673828125,
"step": 113000
},
{
"epoch": 0.7176549439154242,
"grad_norm": 1.7015069723129272,
"learning_rate": 8.238435462676714e-05,
"loss": 3.087583740234375,
"step": 113500
},
{
"epoch": 0.7208164194392807,
"grad_norm": 1.6894475221633911,
"learning_rate": 8.230500415796457e-05,
"loss": 3.0811455078125,
"step": 114000
},
{
"epoch": 0.7239778949631372,
"grad_norm": 1.720460295677185,
"learning_rate": 8.2225653689162e-05,
"loss": 3.080820556640625,
"step": 114500
},
{
"epoch": 0.7271393704869937,
"grad_norm": 1.6984832286834717,
"learning_rate": 8.214630322035943e-05,
"loss": 3.07356982421875,
"step": 115000
},
{
"epoch": 0.7303008460108502,
"grad_norm": 1.732249140739441,
"learning_rate": 8.206695275155686e-05,
"loss": 3.066628173828125,
"step": 115500
},
{
"epoch": 0.7334623215347067,
"grad_norm": 1.6200437545776367,
"learning_rate": 8.19876022827543e-05,
"loss": 3.059205078125,
"step": 116000
},
{
"epoch": 0.7366237970585632,
"grad_norm": 1.6846116781234741,
"learning_rate": 8.190825181395172e-05,
"loss": 3.05877197265625,
"step": 116500
},
{
"epoch": 0.7397852725824197,
"grad_norm": 1.6158053874969482,
"learning_rate": 8.182890134514916e-05,
"loss": 3.04815234375,
"step": 117000
},
{
"epoch": 0.7429467481062761,
"grad_norm": 1.5606156587600708,
"learning_rate": 8.174955087634658e-05,
"loss": 3.04610791015625,
"step": 117500
},
{
"epoch": 0.7461082236301326,
"grad_norm": 1.6943882703781128,
"learning_rate": 8.1670200407544e-05,
"loss": 3.04420703125,
"step": 118000
},
{
"epoch": 0.7492696991539891,
"grad_norm": 1.7836681604385376,
"learning_rate": 8.159084993874144e-05,
"loss": 3.03944775390625,
"step": 118500
},
{
"epoch": 0.7524311746778456,
"grad_norm": 1.7005430459976196,
"learning_rate": 8.151149946993886e-05,
"loss": 3.03258203125,
"step": 119000
},
{
"epoch": 0.7555926502017022,
"grad_norm": 1.6048686504364014,
"learning_rate": 8.14321490011363e-05,
"loss": 3.03843212890625,
"step": 119500
},
{
"epoch": 0.7587541257255587,
"grad_norm": 1.6519755125045776,
"learning_rate": 8.135279853233372e-05,
"loss": 3.025102783203125,
"step": 120000
},
{
"epoch": 0.7619156012494152,
"grad_norm": 1.6674240827560425,
"learning_rate": 8.127344806353116e-05,
"loss": 3.0205283203125,
"step": 120500
},
{
"epoch": 0.7650770767732716,
"grad_norm": 1.706898808479309,
"learning_rate": 8.119409759472859e-05,
"loss": 3.018805419921875,
"step": 121000
},
{
"epoch": 0.7682385522971281,
"grad_norm": 1.6843581199645996,
"learning_rate": 8.111474712592602e-05,
"loss": 3.00777490234375,
"step": 121500
},
{
"epoch": 0.7714000278209846,
"grad_norm": 1.6738733053207397,
"learning_rate": 8.103539665712345e-05,
"loss": 3.009791259765625,
"step": 122000
},
{
"epoch": 0.7745615033448411,
"grad_norm": 1.611879587173462,
"learning_rate": 8.095604618832088e-05,
"loss": 3.002276123046875,
"step": 122500
},
{
"epoch": 0.7777229788686976,
"grad_norm": 1.64024019241333,
"learning_rate": 8.087669571951831e-05,
"loss": 3.000127685546875,
"step": 123000
},
{
"epoch": 0.780884454392554,
"grad_norm": 1.6195515394210815,
"learning_rate": 8.079734525071574e-05,
"loss": 2.9970771484375,
"step": 123500
},
{
"epoch": 0.7840459299164105,
"grad_norm": 1.706299901008606,
"learning_rate": 8.071799478191318e-05,
"loss": 2.99021435546875,
"step": 124000
},
{
"epoch": 0.7872074054402671,
"grad_norm": 1.6817374229431152,
"learning_rate": 8.06386443131106e-05,
"loss": 2.991185302734375,
"step": 124500
},
{
"epoch": 0.7903688809641236,
"grad_norm": 1.7697136402130127,
"learning_rate": 8.055929384430804e-05,
"loss": 2.992920166015625,
"step": 125000
},
{
"epoch": 0.7935303564879801,
"grad_norm": 1.7556347846984863,
"learning_rate": 8.047994337550547e-05,
"loss": 2.98538037109375,
"step": 125500
},
{
"epoch": 0.7966918320118366,
"grad_norm": 1.5977143049240112,
"learning_rate": 8.04005929067029e-05,
"loss": 2.979695556640625,
"step": 126000
},
{
"epoch": 0.7998533075356931,
"grad_norm": 1.7634971141815186,
"learning_rate": 8.032124243790033e-05,
"loss": 2.972553955078125,
"step": 126500
},
{
"epoch": 0.8030147830595495,
"grad_norm": 1.6287751197814941,
"learning_rate": 8.024189196909776e-05,
"loss": 2.9744697265625,
"step": 127000
},
{
"epoch": 0.806176258583406,
"grad_norm": 1.7679568529129028,
"learning_rate": 8.016254150029519e-05,
"loss": 2.96809033203125,
"step": 127500
},
{
"epoch": 0.8093377341072625,
"grad_norm": 1.7397295236587524,
"learning_rate": 8.008319103149263e-05,
"loss": 2.9700439453125,
"step": 128000
},
{
"epoch": 0.812499209631119,
"grad_norm": 1.6410273313522339,
"learning_rate": 8.000384056269005e-05,
"loss": 2.960937255859375,
"step": 128500
},
{
"epoch": 0.8156606851549756,
"grad_norm": 1.6685179471969604,
"learning_rate": 7.992449009388749e-05,
"loss": 2.96211767578125,
"step": 129000
},
{
"epoch": 0.8188221606788321,
"grad_norm": 1.5752959251403809,
"learning_rate": 7.984513962508491e-05,
"loss": 2.955095458984375,
"step": 129500
},
{
"epoch": 0.8219836362026886,
"grad_norm": 1.7220090627670288,
"learning_rate": 7.976578915628235e-05,
"loss": 2.9514921875,
"step": 130000
},
{
"epoch": 0.825145111726545,
"grad_norm": 1.5628429651260376,
"learning_rate": 7.968643868747977e-05,
"loss": 2.950453369140625,
"step": 130500
},
{
"epoch": 0.8283065872504015,
"grad_norm": 1.6883184909820557,
"learning_rate": 7.96070882186772e-05,
"loss": 2.952093505859375,
"step": 131000
},
{
"epoch": 0.831468062774258,
"grad_norm": 1.6512421369552612,
"learning_rate": 7.952773774987463e-05,
"loss": 2.9493134765625,
"step": 131500
},
{
"epoch": 0.8346295382981145,
"grad_norm": 1.6790326833724976,
"learning_rate": 7.944838728107206e-05,
"loss": 2.935697998046875,
"step": 132000
},
{
"epoch": 0.837791013821971,
"grad_norm": 1.6734904050827026,
"learning_rate": 7.936903681226949e-05,
"loss": 2.934459716796875,
"step": 132500
},
{
"epoch": 0.8409524893458274,
"grad_norm": 1.6342508792877197,
"learning_rate": 7.928968634346692e-05,
"loss": 2.9415400390625,
"step": 133000
},
{
"epoch": 0.8441139648696839,
"grad_norm": 1.6908241510391235,
"learning_rate": 7.921033587466435e-05,
"loss": 2.93002294921875,
"step": 133500
},
{
"epoch": 0.8472754403935405,
"grad_norm": 1.6352535486221313,
"learning_rate": 7.913098540586178e-05,
"loss": 2.925991455078125,
"step": 134000
},
{
"epoch": 0.850436915917397,
"grad_norm": 1.7154276371002197,
"learning_rate": 7.905163493705921e-05,
"loss": 2.926026123046875,
"step": 134500
},
{
"epoch": 0.8535983914412535,
"grad_norm": 1.6526648998260498,
"learning_rate": 7.897228446825664e-05,
"loss": 2.922724365234375,
"step": 135000
},
{
"epoch": 0.85675986696511,
"grad_norm": 1.6378566026687622,
"learning_rate": 7.889293399945407e-05,
"loss": 2.922158935546875,
"step": 135500
},
{
"epoch": 0.8599213424889665,
"grad_norm": 1.5609475374221802,
"learning_rate": 7.88135835306515e-05,
"loss": 2.9226767578125,
"step": 136000
},
{
"epoch": 0.8630828180128229,
"grad_norm": 1.7239793539047241,
"learning_rate": 7.873423306184892e-05,
"loss": 2.914996826171875,
"step": 136500
},
{
"epoch": 0.8662442935366794,
"grad_norm": 1.6344202756881714,
"learning_rate": 7.865488259304636e-05,
"loss": 2.911488525390625,
"step": 137000
},
{
"epoch": 0.8694057690605359,
"grad_norm": 1.6337497234344482,
"learning_rate": 7.857553212424378e-05,
"loss": 2.91001708984375,
"step": 137500
},
{
"epoch": 0.8725672445843924,
"grad_norm": 1.5935488939285278,
"learning_rate": 7.849618165544122e-05,
"loss": 2.908462646484375,
"step": 138000
},
{
"epoch": 0.8757287201082489,
"grad_norm": 1.6326372623443604,
"learning_rate": 7.841683118663866e-05,
"loss": 2.900247802734375,
"step": 138500
},
{
"epoch": 0.8788901956321055,
"grad_norm": 1.6682499647140503,
"learning_rate": 7.833748071783608e-05,
"loss": 2.9056259765625,
"step": 139000
},
{
"epoch": 0.882051671155962,
"grad_norm": 1.7714394330978394,
"learning_rate": 7.825813024903352e-05,
"loss": 2.902502197265625,
"step": 139500
},
{
"epoch": 0.8852131466798184,
"grad_norm": 1.6796106100082397,
"learning_rate": 7.817877978023094e-05,
"loss": 2.904083984375,
"step": 140000
},
{
"epoch": 0.8883746222036749,
"grad_norm": 1.5651307106018066,
"learning_rate": 7.809942931142838e-05,
"loss": 2.9000126953125,
"step": 140500
},
{
"epoch": 0.8915360977275314,
"grad_norm": 1.6732896566390991,
"learning_rate": 7.802007884262582e-05,
"loss": 2.89006787109375,
"step": 141000
},
{
"epoch": 0.8946975732513879,
"grad_norm": 1.6779829263687134,
"learning_rate": 7.794072837382324e-05,
"loss": 2.88901904296875,
"step": 141500
},
{
"epoch": 0.8978590487752444,
"grad_norm": 1.6832624673843384,
"learning_rate": 7.786137790502068e-05,
"loss": 2.88577783203125,
"step": 142000
},
{
"epoch": 0.9010205242991008,
"grad_norm": 1.6220341920852661,
"learning_rate": 7.77820274362181e-05,
"loss": 2.88572509765625,
"step": 142500
},
{
"epoch": 0.9041819998229573,
"grad_norm": 1.6310882568359375,
"learning_rate": 7.770267696741552e-05,
"loss": 2.877633544921875,
"step": 143000
},
{
"epoch": 0.9073434753468138,
"grad_norm": 1.6173427104949951,
"learning_rate": 7.762332649861296e-05,
"loss": 2.876183349609375,
"step": 143500
},
{
"epoch": 0.9105049508706704,
"grad_norm": 1.6362117528915405,
"learning_rate": 7.754397602981039e-05,
"loss": 2.878767333984375,
"step": 144000
},
{
"epoch": 0.9136664263945269,
"grad_norm": 1.6228299140930176,
"learning_rate": 7.746462556100782e-05,
"loss": 2.873588623046875,
"step": 144500
},
{
"epoch": 0.9168279019183834,
"grad_norm": 1.6470800638198853,
"learning_rate": 7.738527509220525e-05,
"loss": 2.873649658203125,
"step": 145000
},
{
"epoch": 0.9199893774422399,
"grad_norm": 1.6289581060409546,
"learning_rate": 7.730592462340268e-05,
"loss": 2.86982763671875,
"step": 145500
},
{
"epoch": 0.9231508529660963,
"grad_norm": 1.6964225769042969,
"learning_rate": 7.722657415460011e-05,
"loss": 2.871602783203125,
"step": 146000
},
{
"epoch": 0.9263123284899528,
"grad_norm": 1.5833568572998047,
"learning_rate": 7.714722368579754e-05,
"loss": 2.863375732421875,
"step": 146500
},
{
"epoch": 0.9294738040138093,
"grad_norm": 1.5952502489089966,
"learning_rate": 7.706787321699497e-05,
"loss": 2.857998779296875,
"step": 147000
},
{
"epoch": 0.9326352795376658,
"grad_norm": 1.672086238861084,
"learning_rate": 7.69885227481924e-05,
"loss": 2.85889111328125,
"step": 147500
},
{
"epoch": 0.9357967550615223,
"grad_norm": 1.7695456743240356,
"learning_rate": 7.690917227938983e-05,
"loss": 2.85682763671875,
"step": 148000
},
{
"epoch": 0.9389582305853789,
"grad_norm": 1.6341122388839722,
"learning_rate": 7.682982181058727e-05,
"loss": 2.859611083984375,
"step": 148500
},
{
"epoch": 0.9421197061092353,
"grad_norm": 1.5835165977478027,
"learning_rate": 7.675047134178469e-05,
"loss": 2.85981201171875,
"step": 149000
},
{
"epoch": 0.9452811816330918,
"grad_norm": 1.6251165866851807,
"learning_rate": 7.667112087298211e-05,
"loss": 2.856054931640625,
"step": 149500
},
{
"epoch": 0.9484426571569483,
"grad_norm": 1.6627188920974731,
"learning_rate": 7.659177040417955e-05,
"loss": 2.847407958984375,
"step": 150000
},
{
"epoch": 0.9516041326808048,
"grad_norm": 1.5853898525238037,
"learning_rate": 7.651241993537697e-05,
"loss": 2.848890380859375,
"step": 150500
},
{
"epoch": 0.9547656082046613,
"grad_norm": 1.5742989778518677,
"learning_rate": 7.643306946657441e-05,
"loss": 2.84681591796875,
"step": 151000
},
{
"epoch": 0.9579270837285178,
"grad_norm": 1.736249327659607,
"learning_rate": 7.635371899777183e-05,
"loss": 2.847203369140625,
"step": 151500
},
{
"epoch": 0.9610885592523742,
"grad_norm": 1.5661486387252808,
"learning_rate": 7.627436852896927e-05,
"loss": 2.847230712890625,
"step": 152000
},
{
"epoch": 0.9642500347762307,
"grad_norm": 1.6370384693145752,
"learning_rate": 7.61950180601667e-05,
"loss": 2.836507080078125,
"step": 152500
},
{
"epoch": 0.9674115103000872,
"grad_norm": 1.6324483156204224,
"learning_rate": 7.611566759136413e-05,
"loss": 2.838198974609375,
"step": 153000
},
{
"epoch": 0.9705729858239438,
"grad_norm": 1.65053129196167,
"learning_rate": 7.603631712256156e-05,
"loss": 2.837197021484375,
"step": 153500
},
{
"epoch": 0.9737344613478003,
"grad_norm": 1.8417246341705322,
"learning_rate": 7.5956966653759e-05,
"loss": 2.834068115234375,
"step": 154000
},
{
"epoch": 0.9768959368716568,
"grad_norm": 1.5988879203796387,
"learning_rate": 7.587761618495642e-05,
"loss": 2.8373505859375,
"step": 154500
},
{
"epoch": 0.9800574123955132,
"grad_norm": 1.6057460308074951,
"learning_rate": 7.579826571615385e-05,
"loss": 2.82822265625,
"step": 155000
},
{
"epoch": 0.9832188879193697,
"grad_norm": 1.6583877801895142,
"learning_rate": 7.571891524735129e-05,
"loss": 2.833552734375,
"step": 155500
},
{
"epoch": 0.9863803634432262,
"grad_norm": 1.6536940336227417,
"learning_rate": 7.563956477854872e-05,
"loss": 2.82963134765625,
"step": 156000
},
{
"epoch": 0.9895418389670827,
"grad_norm": 1.600208044052124,
"learning_rate": 7.556021430974615e-05,
"loss": 2.82328076171875,
"step": 156500
},
{
"epoch": 0.9927033144909392,
"grad_norm": 1.6163676977157593,
"learning_rate": 7.548086384094358e-05,
"loss": 2.82417919921875,
"step": 157000
},
{
"epoch": 0.9958647900147957,
"grad_norm": 1.6232937574386597,
"learning_rate": 7.540151337214101e-05,
"loss": 2.819564697265625,
"step": 157500
},
{
"epoch": 0.9990262655386521,
"grad_norm": 1.691518783569336,
"learning_rate": 7.532216290333844e-05,
"loss": 2.826250244140625,
"step": 158000
},
{
"epoch": 1.0021877410625086,
"grad_norm": 1.5797455310821533,
"learning_rate": 7.524281243453587e-05,
"loss": 2.821036376953125,
"step": 158500
},
{
"epoch": 1.0053492165863651,
"grad_norm": 1.6649246215820312,
"learning_rate": 7.51634619657333e-05,
"loss": 2.815264892578125,
"step": 159000
},
{
"epoch": 1.0085106921102216,
"grad_norm": 1.6174602508544922,
"learning_rate": 7.508411149693074e-05,
"loss": 2.817345458984375,
"step": 159500
},
{
"epoch": 1.011672167634078,
"grad_norm": 1.6105290651321411,
"learning_rate": 7.500476102812816e-05,
"loss": 2.812448974609375,
"step": 160000
},
{
"epoch": 1.0148336431579346,
"grad_norm": 1.750746488571167,
"learning_rate": 7.49254105593256e-05,
"loss": 2.806902099609375,
"step": 160500
},
{
"epoch": 1.0179951186817913,
"grad_norm": 1.670493721961975,
"learning_rate": 7.484606009052302e-05,
"loss": 2.811728271484375,
"step": 161000
},
{
"epoch": 1.0211565942056477,
"grad_norm": 1.5906498432159424,
"learning_rate": 7.476670962172044e-05,
"loss": 2.812373046875,
"step": 161500
},
{
"epoch": 1.0243180697295042,
"grad_norm": 1.6185046434402466,
"learning_rate": 7.468735915291788e-05,
"loss": 2.808435791015625,
"step": 162000
},
{
"epoch": 1.0274795452533607,
"grad_norm": 1.7215126752853394,
"learning_rate": 7.46080086841153e-05,
"loss": 2.80660888671875,
"step": 162500
},
{
"epoch": 1.0306410207772172,
"grad_norm": 1.6222294569015503,
"learning_rate": 7.452865821531274e-05,
"loss": 2.8032041015625,
"step": 163000
},
{
"epoch": 1.0338024963010737,
"grad_norm": 1.6534178256988525,
"learning_rate": 7.444930774651017e-05,
"loss": 2.804851318359375,
"step": 163500
},
{
"epoch": 1.0369639718249302,
"grad_norm": 1.6033555269241333,
"learning_rate": 7.43699572777076e-05,
"loss": 2.8048232421875,
"step": 164000
},
{
"epoch": 1.0401254473487866,
"grad_norm": 1.6493546962738037,
"learning_rate": 7.429060680890503e-05,
"loss": 2.800197998046875,
"step": 164500
},
{
"epoch": 1.0432869228726431,
"grad_norm": 1.722131609916687,
"learning_rate": 7.421125634010246e-05,
"loss": 2.792795654296875,
"step": 165000
},
{
"epoch": 1.0464483983964996,
"grad_norm": 1.6339683532714844,
"learning_rate": 7.413190587129989e-05,
"loss": 2.793512451171875,
"step": 165500
},
{
"epoch": 1.049609873920356,
"grad_norm": 1.6593602895736694,
"learning_rate": 7.405255540249732e-05,
"loss": 2.791615478515625,
"step": 166000
},
{
"epoch": 1.0527713494442126,
"grad_norm": 1.605626106262207,
"learning_rate": 7.397320493369475e-05,
"loss": 2.7911318359375,
"step": 166500
},
{
"epoch": 1.055932824968069,
"grad_norm": 1.7574527263641357,
"learning_rate": 7.389385446489218e-05,
"loss": 2.790322998046875,
"step": 167000
},
{
"epoch": 1.0590943004919255,
"grad_norm": 1.640408992767334,
"learning_rate": 7.381450399608961e-05,
"loss": 2.782034912109375,
"step": 167500
},
{
"epoch": 1.062255776015782,
"grad_norm": 1.635996699333191,
"learning_rate": 7.373515352728703e-05,
"loss": 2.789693359375,
"step": 168000
},
{
"epoch": 1.0654172515396385,
"grad_norm": 1.7289674282073975,
"learning_rate": 7.365580305848447e-05,
"loss": 2.782711669921875,
"step": 168500
},
{
"epoch": 1.068578727063495,
"grad_norm": 1.5934735536575317,
"learning_rate": 7.357645258968189e-05,
"loss": 2.780425048828125,
"step": 169000
},
{
"epoch": 1.0717402025873515,
"grad_norm": 1.5848383903503418,
"learning_rate": 7.349710212087933e-05,
"loss": 2.785382080078125,
"step": 169500
},
{
"epoch": 1.0749016781112082,
"grad_norm": 1.704858660697937,
"learning_rate": 7.341775165207675e-05,
"loss": 2.776736083984375,
"step": 170000
},
{
"epoch": 1.0780631536350644,
"grad_norm": 1.6746150255203247,
"learning_rate": 7.333840118327419e-05,
"loss": 2.779871337890625,
"step": 170500
},
{
"epoch": 1.0812246291589211,
"grad_norm": 1.6855316162109375,
"learning_rate": 7.325905071447163e-05,
"loss": 2.779400634765625,
"step": 171000
},
{
"epoch": 1.0843861046827776,
"grad_norm": 1.6358593702316284,
"learning_rate": 7.317970024566905e-05,
"loss": 2.7807392578125,
"step": 171500
},
{
"epoch": 1.0875475802066341,
"grad_norm": 1.6463135480880737,
"learning_rate": 7.310034977686649e-05,
"loss": 2.77948828125,
"step": 172000
},
{
"epoch": 1.0907090557304906,
"grad_norm": 1.612248182296753,
"learning_rate": 7.302099930806393e-05,
"loss": 2.773500244140625,
"step": 172500
},
{
"epoch": 1.093870531254347,
"grad_norm": 1.646470308303833,
"learning_rate": 7.294164883926135e-05,
"loss": 2.773683349609375,
"step": 173000
},
{
"epoch": 1.0970320067782036,
"grad_norm": 1.6269583702087402,
"learning_rate": 7.286229837045879e-05,
"loss": 2.7780439453125,
"step": 173500
},
{
"epoch": 1.10019348230206,
"grad_norm": 1.6568814516067505,
"learning_rate": 7.278294790165621e-05,
"loss": 2.772577880859375,
"step": 174000
},
{
"epoch": 1.1033549578259165,
"grad_norm": 1.7530194520950317,
"learning_rate": 7.270359743285363e-05,
"loss": 2.769607666015625,
"step": 174500
},
{
"epoch": 1.106516433349773,
"grad_norm": 1.7004567384719849,
"learning_rate": 7.262424696405107e-05,
"loss": 2.76552392578125,
"step": 175000
},
{
"epoch": 1.1096779088736295,
"grad_norm": 1.6303637027740479,
"learning_rate": 7.25448964952485e-05,
"loss": 2.767409912109375,
"step": 175500
},
{
"epoch": 1.112839384397486,
"grad_norm": 1.6466504335403442,
"learning_rate": 7.246554602644593e-05,
"loss": 2.7694228515625,
"step": 176000
},
{
"epoch": 1.1160008599213425,
"grad_norm": 1.6781409978866577,
"learning_rate": 7.238619555764336e-05,
"loss": 2.76787646484375,
"step": 176500
},
{
"epoch": 1.119162335445199,
"grad_norm": 1.7949708700180054,
"learning_rate": 7.23068450888408e-05,
"loss": 2.75872021484375,
"step": 177000
},
{
"epoch": 1.1223238109690554,
"grad_norm": 1.625771164894104,
"learning_rate": 7.222749462003822e-05,
"loss": 2.766646484375,
"step": 177500
},
{
"epoch": 1.125485286492912,
"grad_norm": 1.6540682315826416,
"learning_rate": 7.214814415123565e-05,
"loss": 2.761159423828125,
"step": 178000
},
{
"epoch": 1.1286467620167684,
"grad_norm": 1.6319142580032349,
"learning_rate": 7.206879368243308e-05,
"loss": 2.75873046875,
"step": 178500
},
{
"epoch": 1.1318082375406249,
"grad_norm": 1.6187270879745483,
"learning_rate": 7.198944321363051e-05,
"loss": 2.763001953125,
"step": 179000
},
{
"epoch": 1.1349697130644816,
"grad_norm": 1.6837782859802246,
"learning_rate": 7.191009274482794e-05,
"loss": 2.755226806640625,
"step": 179500
},
{
"epoch": 1.1381311885883378,
"grad_norm": 1.6209704875946045,
"learning_rate": 7.183074227602538e-05,
"loss": 2.758800537109375,
"step": 180000
},
{
"epoch": 1.1412926641121945,
"grad_norm": 1.6386946439743042,
"learning_rate": 7.17513918072228e-05,
"loss": 2.75598974609375,
"step": 180500
},
{
"epoch": 1.144454139636051,
"grad_norm": 1.6710797548294067,
"learning_rate": 7.167204133842022e-05,
"loss": 2.754275634765625,
"step": 181000
},
{
"epoch": 1.1476156151599075,
"grad_norm": 1.6988400220870972,
"learning_rate": 7.159269086961766e-05,
"loss": 2.747623779296875,
"step": 181500
},
{
"epoch": 1.150777090683764,
"grad_norm": 1.6538398265838623,
"learning_rate": 7.151334040081508e-05,
"loss": 2.7496640625,
"step": 182000
},
{
"epoch": 1.1539385662076205,
"grad_norm": 1.6186549663543701,
"learning_rate": 7.143398993201252e-05,
"loss": 2.74736474609375,
"step": 182500
},
{
"epoch": 1.157100041731477,
"grad_norm": 1.5646486282348633,
"learning_rate": 7.135463946320994e-05,
"loss": 2.740775146484375,
"step": 183000
},
{
"epoch": 1.1602615172553334,
"grad_norm": 1.5905307531356812,
"learning_rate": 7.127528899440738e-05,
"loss": 2.750447509765625,
"step": 183500
},
{
"epoch": 1.16342299277919,
"grad_norm": 1.620119571685791,
"learning_rate": 7.11959385256048e-05,
"loss": 2.744733642578125,
"step": 184000
},
{
"epoch": 1.1665844683030464,
"grad_norm": 1.6881380081176758,
"learning_rate": 7.111658805680224e-05,
"loss": 2.74177734375,
"step": 184500
},
{
"epoch": 1.169745943826903,
"grad_norm": 1.6603327989578247,
"learning_rate": 7.103723758799967e-05,
"loss": 2.745770751953125,
"step": 185000
},
{
"epoch": 1.1729074193507594,
"grad_norm": 1.6077136993408203,
"learning_rate": 7.09578871191971e-05,
"loss": 2.74176904296875,
"step": 185500
},
{
"epoch": 1.1760688948746159,
"grad_norm": 1.6371090412139893,
"learning_rate": 7.087853665039453e-05,
"loss": 2.737481201171875,
"step": 186000
},
{
"epoch": 1.1792303703984723,
"grad_norm": 1.6856745481491089,
"learning_rate": 7.079918618159196e-05,
"loss": 2.742541259765625,
"step": 186500
},
{
"epoch": 1.1823918459223288,
"grad_norm": 1.5946813821792603,
"learning_rate": 7.07198357127894e-05,
"loss": 2.7407490234375,
"step": 187000
},
{
"epoch": 1.1855533214461853,
"grad_norm": 1.6449916362762451,
"learning_rate": 7.064048524398683e-05,
"loss": 2.735882568359375,
"step": 187500
},
{
"epoch": 1.1887147969700418,
"grad_norm": 1.704580307006836,
"learning_rate": 7.056113477518426e-05,
"loss": 2.73212548828125,
"step": 188000
},
{
"epoch": 1.1918762724938983,
"grad_norm": 1.6035581827163696,
"learning_rate": 7.048178430638169e-05,
"loss": 2.735862548828125,
"step": 188500
},
{
"epoch": 1.1950377480177548,
"grad_norm": 1.6927502155303955,
"learning_rate": 7.040243383757912e-05,
"loss": 2.740005859375,
"step": 189000
},
{
"epoch": 1.1981992235416112,
"grad_norm": 1.5842722654342651,
"learning_rate": 7.032308336877655e-05,
"loss": 2.73451318359375,
"step": 189500
},
{
"epoch": 1.201360699065468,
"grad_norm": 1.6199719905853271,
"learning_rate": 7.024373289997398e-05,
"loss": 2.730331298828125,
"step": 190000
},
{
"epoch": 1.2045221745893244,
"grad_norm": 1.6090415716171265,
"learning_rate": 7.016438243117141e-05,
"loss": 2.72792138671875,
"step": 190500
},
{
"epoch": 1.207683650113181,
"grad_norm": 1.6963419914245605,
"learning_rate": 7.008503196236885e-05,
"loss": 2.7249306640625,
"step": 191000
},
{
"epoch": 1.2108451256370374,
"grad_norm": 1.6181541681289673,
"learning_rate": 7.000568149356627e-05,
"loss": 2.72613330078125,
"step": 191500
},
{
"epoch": 1.2140066011608939,
"grad_norm": 1.6325851678848267,
"learning_rate": 6.99263310247637e-05,
"loss": 2.725619140625,
"step": 192000
},
{
"epoch": 1.2171680766847504,
"grad_norm": 1.7136417627334595,
"learning_rate": 6.984698055596113e-05,
"loss": 2.7276318359375,
"step": 192500
},
{
"epoch": 1.2203295522086068,
"grad_norm": 1.6793721914291382,
"learning_rate": 6.976763008715855e-05,
"loss": 2.723571044921875,
"step": 193000
},
{
"epoch": 1.2234910277324633,
"grad_norm": 1.6753904819488525,
"learning_rate": 6.968827961835599e-05,
"loss": 2.72617333984375,
"step": 193500
},
{
"epoch": 1.2266525032563198,
"grad_norm": 1.764931559562683,
"learning_rate": 6.960892914955341e-05,
"loss": 2.721798583984375,
"step": 194000
},
{
"epoch": 1.2298139787801763,
"grad_norm": 1.7031928300857544,
"learning_rate": 6.952957868075085e-05,
"loss": 2.725414794921875,
"step": 194500
},
{
"epoch": 1.2329754543040328,
"grad_norm": 1.6813712120056152,
"learning_rate": 6.945022821194827e-05,
"loss": 2.723485107421875,
"step": 195000
},
{
"epoch": 1.2361369298278893,
"grad_norm": 1.6624706983566284,
"learning_rate": 6.937087774314571e-05,
"loss": 2.7216982421875,
"step": 195500
},
{
"epoch": 1.2392984053517457,
"grad_norm": 1.6521817445755005,
"learning_rate": 6.929152727434314e-05,
"loss": 2.714099365234375,
"step": 196000
},
{
"epoch": 1.2424598808756022,
"grad_norm": 1.6477317810058594,
"learning_rate": 6.921217680554057e-05,
"loss": 2.709621337890625,
"step": 196500
},
{
"epoch": 1.2456213563994587,
"grad_norm": 1.6554585695266724,
"learning_rate": 6.9132826336738e-05,
"loss": 2.713253662109375,
"step": 197000
},
{
"epoch": 1.2487828319233152,
"grad_norm": 1.7003973722457886,
"learning_rate": 6.905347586793543e-05,
"loss": 2.71363818359375,
"step": 197500
},
{
"epoch": 1.2519443074471717,
"grad_norm": 1.6144161224365234,
"learning_rate": 6.897412539913286e-05,
"loss": 2.712405029296875,
"step": 198000
},
{
"epoch": 1.2551057829710284,
"grad_norm": 1.6349759101867676,
"learning_rate": 6.88947749303303e-05,
"loss": 2.71151708984375,
"step": 198500
},
{
"epoch": 1.2582672584948846,
"grad_norm": 1.599913239479065,
"learning_rate": 6.881542446152772e-05,
"loss": 2.7117060546875,
"step": 199000
},
{
"epoch": 1.2614287340187413,
"grad_norm": 1.6667516231536865,
"learning_rate": 6.873607399272514e-05,
"loss": 2.711345458984375,
"step": 199500
},
{
"epoch": 1.2645902095425976,
"grad_norm": 1.5705628395080566,
"learning_rate": 6.865672352392258e-05,
"loss": 2.716991455078125,
"step": 200000
},
{
"epoch": 1.2677516850664543,
"grad_norm": 1.7059932947158813,
"learning_rate": 6.857737305512e-05,
"loss": 2.707239013671875,
"step": 200500
},
{
"epoch": 1.2709131605903108,
"grad_norm": 1.6093007326126099,
"learning_rate": 6.849802258631744e-05,
"loss": 2.704814453125,
"step": 201000
},
{
"epoch": 1.2740746361141673,
"grad_norm": 1.6355576515197754,
"learning_rate": 6.841867211751486e-05,
"loss": 2.70478076171875,
"step": 201500
},
{
"epoch": 1.2772361116380238,
"grad_norm": 1.6828811168670654,
"learning_rate": 6.83393216487123e-05,
"loss": 2.7034560546875,
"step": 202000
},
{
"epoch": 1.2803975871618802,
"grad_norm": 1.6052861213684082,
"learning_rate": 6.825997117990974e-05,
"loss": 2.706068115234375,
"step": 202500
},
{
"epoch": 1.2835590626857367,
"grad_norm": 1.6489814519882202,
"learning_rate": 6.818062071110716e-05,
"loss": 2.700542724609375,
"step": 203000
},
{
"epoch": 1.2867205382095932,
"grad_norm": 1.6461740732192993,
"learning_rate": 6.81012702423046e-05,
"loss": 2.702365478515625,
"step": 203500
},
{
"epoch": 1.2898820137334497,
"grad_norm": 1.6999510526657104,
"learning_rate": 6.802191977350204e-05,
"loss": 2.7069765625,
"step": 204000
},
{
"epoch": 1.2930434892573062,
"grad_norm": 1.7012547254562378,
"learning_rate": 6.794256930469946e-05,
"loss": 2.7000263671875,
"step": 204500
},
{
"epoch": 1.2962049647811626,
"grad_norm": 1.6145321130752563,
"learning_rate": 6.78632188358969e-05,
"loss": 2.700908935546875,
"step": 205000
},
{
"epoch": 1.2993664403050191,
"grad_norm": 1.6511955261230469,
"learning_rate": 6.778386836709432e-05,
"loss": 2.7005595703125,
"step": 205500
},
{
"epoch": 1.3025279158288756,
"grad_norm": 1.7109709978103638,
"learning_rate": 6.770451789829174e-05,
"loss": 2.69959814453125,
"step": 206000
},
{
"epoch": 1.305689391352732,
"grad_norm": 1.6310850381851196,
"learning_rate": 6.762516742948918e-05,
"loss": 2.698029052734375,
"step": 206500
},
{
"epoch": 1.3088508668765886,
"grad_norm": 1.5690122842788696,
"learning_rate": 6.75458169606866e-05,
"loss": 2.698597412109375,
"step": 207000
},
{
"epoch": 1.312012342400445,
"grad_norm": 1.673174500465393,
"learning_rate": 6.746646649188404e-05,
"loss": 2.6977373046875,
"step": 207500
},
{
"epoch": 1.3151738179243015,
"grad_norm": 1.5447022914886475,
"learning_rate": 6.738711602308147e-05,
"loss": 2.69799609375,
"step": 208000
},
{
"epoch": 1.318335293448158,
"grad_norm": 1.622251272201538,
"learning_rate": 6.73077655542789e-05,
"loss": 2.69298291015625,
"step": 208500
},
{
"epoch": 1.3214967689720147,
"grad_norm": 1.6124510765075684,
"learning_rate": 6.722841508547633e-05,
"loss": 2.69695947265625,
"step": 209000
},
{
"epoch": 1.324658244495871,
"grad_norm": 1.6567950248718262,
"learning_rate": 6.714906461667376e-05,
"loss": 2.69071630859375,
"step": 209500
},
{
"epoch": 1.3278197200197277,
"grad_norm": 1.6351819038391113,
"learning_rate": 6.706971414787119e-05,
"loss": 2.695449951171875,
"step": 210000
},
{
"epoch": 1.3309811955435842,
"grad_norm": 1.6336193084716797,
"learning_rate": 6.699036367906862e-05,
"loss": 2.69038330078125,
"step": 210500
},
{
"epoch": 1.3341426710674407,
"grad_norm": 1.7055922746658325,
"learning_rate": 6.691101321026605e-05,
"loss": 2.684313232421875,
"step": 211000
},
{
"epoch": 1.3373041465912971,
"grad_norm": 1.7058066129684448,
"learning_rate": 6.683166274146347e-05,
"loss": 2.688022705078125,
"step": 211500
},
{
"epoch": 1.3404656221151536,
"grad_norm": 1.6944793462753296,
"learning_rate": 6.675231227266091e-05,
"loss": 2.689552978515625,
"step": 212000
},
{
"epoch": 1.3436270976390101,
"grad_norm": 1.6105682849884033,
"learning_rate": 6.667296180385833e-05,
"loss": 2.68797509765625,
"step": 212500
},
{
"epoch": 1.3467885731628666,
"grad_norm": 1.6687462329864502,
"learning_rate": 6.659361133505577e-05,
"loss": 2.6857646484375,
"step": 213000
},
{
"epoch": 1.349950048686723,
"grad_norm": 1.6710156202316284,
"learning_rate": 6.65142608662532e-05,
"loss": 2.683174560546875,
"step": 213500
},
{
"epoch": 1.3531115242105796,
"grad_norm": 1.657759189605713,
"learning_rate": 6.643491039745063e-05,
"loss": 2.681852294921875,
"step": 214000
},
{
"epoch": 1.356272999734436,
"grad_norm": 1.7139235734939575,
"learning_rate": 6.635555992864805e-05,
"loss": 2.684544921875,
"step": 214500
},
{
"epoch": 1.3594344752582925,
"grad_norm": 1.596709966659546,
"learning_rate": 6.627620945984549e-05,
"loss": 2.68274072265625,
"step": 215000
},
{
"epoch": 1.362595950782149,
"grad_norm": 1.6329495906829834,
"learning_rate": 6.619685899104292e-05,
"loss": 2.684390625,
"step": 215500
},
{
"epoch": 1.3657574263060055,
"grad_norm": 1.6994291543960571,
"learning_rate": 6.611750852224035e-05,
"loss": 2.684050537109375,
"step": 216000
},
{
"epoch": 1.368918901829862,
"grad_norm": 1.6723417043685913,
"learning_rate": 6.603815805343778e-05,
"loss": 2.679808349609375,
"step": 216500
},
{
"epoch": 1.3720803773537185,
"grad_norm": 1.6547695398330688,
"learning_rate": 6.595880758463521e-05,
"loss": 2.67865625,
"step": 217000
},
{
"epoch": 1.375241852877575,
"grad_norm": 1.6354767084121704,
"learning_rate": 6.587945711583264e-05,
"loss": 2.685214599609375,
"step": 217500
},
{
"epoch": 1.3784033284014314,
"grad_norm": 1.667192816734314,
"learning_rate": 6.580010664703007e-05,
"loss": 2.677499755859375,
"step": 218000
},
{
"epoch": 1.3815648039252881,
"grad_norm": 1.625460147857666,
"learning_rate": 6.572075617822751e-05,
"loss": 2.674024169921875,
"step": 218500
},
{
"epoch": 1.3847262794491444,
"grad_norm": 1.788996696472168,
"learning_rate": 6.564140570942494e-05,
"loss": 2.675288818359375,
"step": 219000
},
{
"epoch": 1.387887754973001,
"grad_norm": 1.6711689233779907,
"learning_rate": 6.556205524062237e-05,
"loss": 2.679442138671875,
"step": 219500
},
{
"epoch": 1.3910492304968574,
"grad_norm": 1.6211497783660889,
"learning_rate": 6.54827047718198e-05,
"loss": 2.67129541015625,
"step": 220000
},
{
"epoch": 1.394210706020714,
"grad_norm": 1.6549731492996216,
"learning_rate": 6.540335430301723e-05,
"loss": 2.674689453125,
"step": 220500
},
{
"epoch": 1.3973721815445705,
"grad_norm": 1.619506597518921,
"learning_rate": 6.532400383421466e-05,
"loss": 2.674619140625,
"step": 221000
},
{
"epoch": 1.400533657068427,
"grad_norm": 1.6018877029418945,
"learning_rate": 6.52446533654121e-05,
"loss": 2.67360693359375,
"step": 221500
},
{
"epoch": 1.4036951325922835,
"grad_norm": 1.5963901281356812,
"learning_rate": 6.516530289660952e-05,
"loss": 2.66919287109375,
"step": 222000
},
{
"epoch": 1.40685660811614,
"grad_norm": 1.7305083274841309,
"learning_rate": 6.508595242780695e-05,
"loss": 2.67486669921875,
"step": 222500
},
{
"epoch": 1.4100180836399965,
"grad_norm": 1.7106187343597412,
"learning_rate": 6.500660195900438e-05,
"loss": 2.66819677734375,
"step": 223000
},
{
"epoch": 1.413179559163853,
"grad_norm": 1.622053861618042,
"learning_rate": 6.492725149020182e-05,
"loss": 2.663556884765625,
"step": 223500
},
{
"epoch": 1.4163410346877094,
"grad_norm": 1.638548731803894,
"learning_rate": 6.484790102139924e-05,
"loss": 2.6644658203125,
"step": 224000
},
{
"epoch": 1.419502510211566,
"grad_norm": 1.6407787799835205,
"learning_rate": 6.476855055259666e-05,
"loss": 2.664386474609375,
"step": 224500
},
{
"epoch": 1.4226639857354224,
"grad_norm": 1.6598243713378906,
"learning_rate": 6.46892000837941e-05,
"loss": 2.66158203125,
"step": 225000
},
{
"epoch": 1.425825461259279,
"grad_norm": 1.6587129831314087,
"learning_rate": 6.460984961499152e-05,
"loss": 2.666587158203125,
"step": 225500
},
{
"epoch": 1.4289869367831354,
"grad_norm": 1.712268590927124,
"learning_rate": 6.453049914618896e-05,
"loss": 2.6628369140625,
"step": 226000
},
{
"epoch": 1.4321484123069919,
"grad_norm": 1.649362564086914,
"learning_rate": 6.445114867738638e-05,
"loss": 2.664687255859375,
"step": 226500
},
{
"epoch": 1.4353098878308483,
"grad_norm": 1.6085597276687622,
"learning_rate": 6.437179820858382e-05,
"loss": 2.660015869140625,
"step": 227000
},
{
"epoch": 1.4384713633547048,
"grad_norm": 1.718865156173706,
"learning_rate": 6.429244773978125e-05,
"loss": 2.66128173828125,
"step": 227500
},
{
"epoch": 1.4416328388785615,
"grad_norm": 1.623750925064087,
"learning_rate": 6.421309727097868e-05,
"loss": 2.660835693359375,
"step": 228000
},
{
"epoch": 1.4447943144024178,
"grad_norm": 1.7748386859893799,
"learning_rate": 6.41337468021761e-05,
"loss": 2.66336474609375,
"step": 228500
},
{
"epoch": 1.4479557899262745,
"grad_norm": 1.692211389541626,
"learning_rate": 6.405439633337354e-05,
"loss": 2.656604736328125,
"step": 229000
},
{
"epoch": 1.4511172654501308,
"grad_norm": 1.728323221206665,
"learning_rate": 6.397504586457097e-05,
"loss": 2.657124755859375,
"step": 229500
},
{
"epoch": 1.4542787409739875,
"grad_norm": 1.6859859228134155,
"learning_rate": 6.38956953957684e-05,
"loss": 2.660123779296875,
"step": 230000
},
{
"epoch": 1.457440216497844,
"grad_norm": 1.7085784673690796,
"learning_rate": 6.381634492696583e-05,
"loss": 2.662109375,
"step": 230500
},
{
"epoch": 1.4606016920217004,
"grad_norm": 1.7014106512069702,
"learning_rate": 6.373699445816325e-05,
"loss": 2.653016845703125,
"step": 231000
},
{
"epoch": 1.463763167545557,
"grad_norm": 1.7352333068847656,
"learning_rate": 6.365764398936069e-05,
"loss": 2.656542236328125,
"step": 231500
},
{
"epoch": 1.4669246430694134,
"grad_norm": 1.6745645999908447,
"learning_rate": 6.357829352055811e-05,
"loss": 2.65683251953125,
"step": 232000
},
{
"epoch": 1.4700861185932699,
"grad_norm": 1.6864614486694336,
"learning_rate": 6.349894305175555e-05,
"loss": 2.658100341796875,
"step": 232500
},
{
"epoch": 1.4732475941171264,
"grad_norm": 1.6714229583740234,
"learning_rate": 6.341959258295297e-05,
"loss": 2.65063134765625,
"step": 233000
},
{
"epoch": 1.4764090696409828,
"grad_norm": 1.6586594581604004,
"learning_rate": 6.334024211415041e-05,
"loss": 2.652447509765625,
"step": 233500
},
{
"epoch": 1.4795705451648393,
"grad_norm": 1.6757060289382935,
"learning_rate": 6.326089164534785e-05,
"loss": 2.65651171875,
"step": 234000
},
{
"epoch": 1.4827320206886958,
"grad_norm": 1.7043514251708984,
"learning_rate": 6.318154117654527e-05,
"loss": 2.6468056640625,
"step": 234500
},
{
"epoch": 1.4858934962125523,
"grad_norm": 1.6739319562911987,
"learning_rate": 6.310219070774271e-05,
"loss": 2.64984033203125,
"step": 235000
},
{
"epoch": 1.4890549717364088,
"grad_norm": 1.658068060874939,
"learning_rate": 6.302284023894015e-05,
"loss": 2.6482412109375,
"step": 235500
},
{
"epoch": 1.4922164472602653,
"grad_norm": 1.7380309104919434,
"learning_rate": 6.294348977013757e-05,
"loss": 2.645617431640625,
"step": 236000
},
{
"epoch": 1.4953779227841217,
"grad_norm": 1.6781498193740845,
"learning_rate": 6.2864139301335e-05,
"loss": 2.6455517578125,
"step": 236500
},
{
"epoch": 1.4985393983079782,
"grad_norm": 1.6390162706375122,
"learning_rate": 6.278478883253243e-05,
"loss": 2.646500732421875,
"step": 237000
},
{
"epoch": 1.501700873831835,
"grad_norm": 1.7134122848510742,
"learning_rate": 6.270543836372985e-05,
"loss": 2.64468603515625,
"step": 237500
},
{
"epoch": 1.5048623493556912,
"grad_norm": 1.63341224193573,
"learning_rate": 6.262608789492729e-05,
"loss": 2.650261962890625,
"step": 238000
},
{
"epoch": 1.508023824879548,
"grad_norm": 1.625917673110962,
"learning_rate": 6.254673742612471e-05,
"loss": 2.644395263671875,
"step": 238500
},
{
"epoch": 1.5111853004034042,
"grad_norm": 1.6020703315734863,
"learning_rate": 6.246738695732215e-05,
"loss": 2.644113037109375,
"step": 239000
},
{
"epoch": 1.5143467759272609,
"grad_norm": 1.7164833545684814,
"learning_rate": 6.238803648851958e-05,
"loss": 2.64645458984375,
"step": 239500
},
{
"epoch": 1.5175082514511171,
"grad_norm": 1.6430004835128784,
"learning_rate": 6.230868601971701e-05,
"loss": 2.64694384765625,
"step": 240000
},
{
"epoch": 1.5206697269749738,
"grad_norm": 1.6472010612487793,
"learning_rate": 6.222933555091444e-05,
"loss": 2.647444091796875,
"step": 240500
},
{
"epoch": 1.5238312024988303,
"grad_norm": 1.663031816482544,
"learning_rate": 6.214998508211187e-05,
"loss": 2.64218798828125,
"step": 241000
},
{
"epoch": 1.5269926780226868,
"grad_norm": 1.711756706237793,
"learning_rate": 6.20706346133093e-05,
"loss": 2.64162109375,
"step": 241500
},
{
"epoch": 1.5301541535465433,
"grad_norm": 1.7167319059371948,
"learning_rate": 6.199128414450673e-05,
"loss": 2.63561181640625,
"step": 242000
},
{
"epoch": 1.5333156290703998,
"grad_norm": 1.8246747255325317,
"learning_rate": 6.191193367570416e-05,
"loss": 2.643242431640625,
"step": 242500
},
{
"epoch": 1.5364771045942562,
"grad_norm": 1.5985685586929321,
"learning_rate": 6.183258320690158e-05,
"loss": 2.638106689453125,
"step": 243000
},
{
"epoch": 1.5396385801181127,
"grad_norm": 1.7320302724838257,
"learning_rate": 6.175323273809902e-05,
"loss": 2.637514892578125,
"step": 243500
},
{
"epoch": 1.5428000556419692,
"grad_norm": 1.6809501647949219,
"learning_rate": 6.167388226929644e-05,
"loss": 2.637152099609375,
"step": 244000
},
{
"epoch": 1.5459615311658257,
"grad_norm": 1.7087905406951904,
"learning_rate": 6.159453180049388e-05,
"loss": 2.63835888671875,
"step": 244500
},
{
"epoch": 1.5491230066896822,
"grad_norm": 1.7214372158050537,
"learning_rate": 6.15151813316913e-05,
"loss": 2.62838720703125,
"step": 245000
},
{
"epoch": 1.5522844822135387,
"grad_norm": 1.8188693523406982,
"learning_rate": 6.143583086288874e-05,
"loss": 2.6379287109375,
"step": 245500
},
{
"epoch": 1.5554459577373954,
"grad_norm": 1.6138770580291748,
"learning_rate": 6.135648039408616e-05,
"loss": 2.64002880859375,
"step": 246000
},
{
"epoch": 1.5586074332612516,
"grad_norm": 1.6003512144088745,
"learning_rate": 6.12771299252836e-05,
"loss": 2.6314140625,
"step": 246500
},
{
"epoch": 1.5617689087851083,
"grad_norm": 1.7507163286209106,
"learning_rate": 6.119777945648103e-05,
"loss": 2.64177587890625,
"step": 247000
},
{
"epoch": 1.5649303843089646,
"grad_norm": 1.6544255018234253,
"learning_rate": 6.111842898767846e-05,
"loss": 2.6343974609375,
"step": 247500
},
{
"epoch": 1.5680918598328213,
"grad_norm": 1.7215958833694458,
"learning_rate": 6.103907851887589e-05,
"loss": 2.6354345703125,
"step": 248000
},
{
"epoch": 1.5712533353566775,
"grad_norm": 1.6138739585876465,
"learning_rate": 6.095972805007333e-05,
"loss": 2.627140625,
"step": 248500
},
{
"epoch": 1.5744148108805343,
"grad_norm": 1.6695003509521484,
"learning_rate": 6.0880377581270754e-05,
"loss": 2.63247509765625,
"step": 249000
},
{
"epoch": 1.5775762864043905,
"grad_norm": 1.6978379487991333,
"learning_rate": 6.080102711246818e-05,
"loss": 2.628751953125,
"step": 249500
},
{
"epoch": 1.5807377619282472,
"grad_norm": 1.6494214534759521,
"learning_rate": 6.0721676643665615e-05,
"loss": 2.631275634765625,
"step": 250000
},
{
"epoch": 1.5838992374521035,
"grad_norm": 1.6484586000442505,
"learning_rate": 6.064232617486304e-05,
"loss": 2.623577392578125,
"step": 250500
},
{
"epoch": 1.5870607129759602,
"grad_norm": 1.7459800243377686,
"learning_rate": 6.0562975706060476e-05,
"loss": 2.628191162109375,
"step": 251000
},
{
"epoch": 1.5902221884998167,
"grad_norm": 1.6993423700332642,
"learning_rate": 6.04836252372579e-05,
"loss": 2.624674560546875,
"step": 251500
},
{
"epoch": 1.5933836640236732,
"grad_norm": 1.6219921112060547,
"learning_rate": 6.0404274768455336e-05,
"loss": 2.62570654296875,
"step": 252000
},
{
"epoch": 1.5965451395475296,
"grad_norm": 1.6559642553329468,
"learning_rate": 6.032492429965276e-05,
"loss": 2.62827490234375,
"step": 252500
},
{
"epoch": 1.5997066150713861,
"grad_norm": 1.7363989353179932,
"learning_rate": 6.02455738308502e-05,
"loss": 2.625711181640625,
"step": 253000
},
{
"epoch": 1.6028680905952426,
"grad_norm": 1.7029197216033936,
"learning_rate": 6.016622336204763e-05,
"loss": 2.624954345703125,
"step": 253500
},
{
"epoch": 1.606029566119099,
"grad_norm": 1.7722911834716797,
"learning_rate": 6.008687289324506e-05,
"loss": 2.62922998046875,
"step": 254000
},
{
"epoch": 1.6091910416429556,
"grad_norm": 1.6780489683151245,
"learning_rate": 6.000752242444249e-05,
"loss": 2.628587158203125,
"step": 254500
},
{
"epoch": 1.612352517166812,
"grad_norm": 1.6272549629211426,
"learning_rate": 5.992817195563991e-05,
"loss": 2.623177734375,
"step": 255000
},
{
"epoch": 1.6155139926906688,
"grad_norm": 1.634222388267517,
"learning_rate": 5.984882148683735e-05,
"loss": 2.62577001953125,
"step": 255500
},
{
"epoch": 1.618675468214525,
"grad_norm": 1.7140114307403564,
"learning_rate": 5.976947101803477e-05,
"loss": 2.625297607421875,
"step": 256000
},
{
"epoch": 1.6218369437383817,
"grad_norm": 1.6207098960876465,
"learning_rate": 5.969012054923221e-05,
"loss": 2.6181875,
"step": 256500
},
{
"epoch": 1.624998419262238,
"grad_norm": 1.7659367322921753,
"learning_rate": 5.9610770080429634e-05,
"loss": 2.617813720703125,
"step": 257000
},
{
"epoch": 1.6281598947860947,
"grad_norm": 1.6837326288223267,
"learning_rate": 5.953141961162707e-05,
"loss": 2.62165478515625,
"step": 257500
},
{
"epoch": 1.631321370309951,
"grad_norm": 1.6211479902267456,
"learning_rate": 5.9452069142824495e-05,
"loss": 2.620236572265625,
"step": 258000
},
{
"epoch": 1.6344828458338077,
"grad_norm": 1.6740574836730957,
"learning_rate": 5.937271867402193e-05,
"loss": 2.61860302734375,
"step": 258500
},
{
"epoch": 1.637644321357664,
"grad_norm": 1.7312082052230835,
"learning_rate": 5.9293368205219356e-05,
"loss": 2.62122802734375,
"step": 259000
},
{
"epoch": 1.6408057968815206,
"grad_norm": 1.709106206893921,
"learning_rate": 5.921401773641679e-05,
"loss": 2.620131591796875,
"step": 259500
},
{
"epoch": 1.6439672724053769,
"grad_norm": 1.6952303647994995,
"learning_rate": 5.9134667267614216e-05,
"loss": 2.61886181640625,
"step": 260000
},
{
"epoch": 1.6471287479292336,
"grad_norm": 1.6450752019882202,
"learning_rate": 5.9055316798811654e-05,
"loss": 2.612457275390625,
"step": 260500
},
{
"epoch": 1.65029022345309,
"grad_norm": 1.662732720375061,
"learning_rate": 5.897596633000908e-05,
"loss": 2.61746826171875,
"step": 261000
},
{
"epoch": 1.6534516989769465,
"grad_norm": 1.6304184198379517,
"learning_rate": 5.889661586120651e-05,
"loss": 2.615102783203125,
"step": 261500
},
{
"epoch": 1.656613174500803,
"grad_norm": 1.6970808506011963,
"learning_rate": 5.8817265392403945e-05,
"loss": 2.6144619140625,
"step": 262000
},
{
"epoch": 1.6597746500246595,
"grad_norm": 1.6778018474578857,
"learning_rate": 5.873791492360137e-05,
"loss": 2.6153955078125,
"step": 262500
},
{
"epoch": 1.662936125548516,
"grad_norm": 1.6311659812927246,
"learning_rate": 5.8658564454798806e-05,
"loss": 2.611220703125,
"step": 263000
},
{
"epoch": 1.6660976010723725,
"grad_norm": 1.6939971446990967,
"learning_rate": 5.857921398599623e-05,
"loss": 2.614873046875,
"step": 263500
},
{
"epoch": 1.669259076596229,
"grad_norm": 1.6402766704559326,
"learning_rate": 5.8499863517193667e-05,
"loss": 2.61980126953125,
"step": 264000
},
{
"epoch": 1.6724205521200854,
"grad_norm": 1.6008195877075195,
"learning_rate": 5.842051304839109e-05,
"loss": 2.611008056640625,
"step": 264500
},
{
"epoch": 1.675582027643942,
"grad_norm": 1.6710777282714844,
"learning_rate": 5.834116257958853e-05,
"loss": 2.61326416015625,
"step": 265000
},
{
"epoch": 1.6787435031677984,
"grad_norm": 1.6271564960479736,
"learning_rate": 5.826181211078595e-05,
"loss": 2.612386962890625,
"step": 265500
},
{
"epoch": 1.6819049786916551,
"grad_norm": 1.6416020393371582,
"learning_rate": 5.818246164198339e-05,
"loss": 2.61024365234375,
"step": 266000
},
{
"epoch": 1.6850664542155114,
"grad_norm": 1.6237388849258423,
"learning_rate": 5.810311117318081e-05,
"loss": 2.60862255859375,
"step": 266500
},
{
"epoch": 1.688227929739368,
"grad_norm": 1.7206228971481323,
"learning_rate": 5.802376070437825e-05,
"loss": 2.611340576171875,
"step": 267000
},
{
"epoch": 1.6913894052632243,
"grad_norm": 1.7325563430786133,
"learning_rate": 5.794441023557567e-05,
"loss": 2.609888671875,
"step": 267500
},
{
"epoch": 1.694550880787081,
"grad_norm": 1.728023648262024,
"learning_rate": 5.78650597667731e-05,
"loss": 2.60783837890625,
"step": 268000
},
{
"epoch": 1.6977123563109373,
"grad_norm": 1.5957330465316772,
"learning_rate": 5.7785709297970534e-05,
"loss": 2.612864013671875,
"step": 268500
},
{
"epoch": 1.700873831834794,
"grad_norm": 1.6862738132476807,
"learning_rate": 5.7706358829167964e-05,
"loss": 2.609041259765625,
"step": 269000
},
{
"epoch": 1.7040353073586503,
"grad_norm": 1.6768975257873535,
"learning_rate": 5.7627008360365394e-05,
"loss": 2.602251708984375,
"step": 269500
},
{
"epoch": 1.707196782882507,
"grad_norm": 1.6330195665359497,
"learning_rate": 5.7547657891562825e-05,
"loss": 2.60638818359375,
"step": 270000
},
{
"epoch": 1.7103582584063635,
"grad_norm": 1.6181910037994385,
"learning_rate": 5.746830742276026e-05,
"loss": 2.60128466796875,
"step": 270500
},
{
"epoch": 1.71351973393022,
"grad_norm": 1.601879358291626,
"learning_rate": 5.7388956953957686e-05,
"loss": 2.60673974609375,
"step": 271000
},
{
"epoch": 1.7166812094540764,
"grad_norm": 1.6694235801696777,
"learning_rate": 5.730960648515512e-05,
"loss": 2.60399462890625,
"step": 271500
},
{
"epoch": 1.719842684977933,
"grad_norm": 1.7662272453308105,
"learning_rate": 5.7230256016352547e-05,
"loss": 2.6027626953125,
"step": 272000
},
{
"epoch": 1.7230041605017894,
"grad_norm": 1.6605037450790405,
"learning_rate": 5.7150905547549984e-05,
"loss": 2.602885498046875,
"step": 272500
},
{
"epoch": 1.7261656360256459,
"grad_norm": 1.721268892288208,
"learning_rate": 5.707155507874741e-05,
"loss": 2.605980224609375,
"step": 273000
},
{
"epoch": 1.7293271115495024,
"grad_norm": 1.7375613451004028,
"learning_rate": 5.6992204609944845e-05,
"loss": 2.60100048828125,
"step": 273500
},
{
"epoch": 1.7324885870733588,
"grad_norm": 1.7031376361846924,
"learning_rate": 5.691285414114227e-05,
"loss": 2.60409326171875,
"step": 274000
},
{
"epoch": 1.7356500625972153,
"grad_norm": 1.7374507188796997,
"learning_rate": 5.683350367233969e-05,
"loss": 2.596071533203125,
"step": 274500
},
{
"epoch": 1.7388115381210718,
"grad_norm": 1.6787822246551514,
"learning_rate": 5.675415320353713e-05,
"loss": 2.597696044921875,
"step": 275000
},
{
"epoch": 1.7419730136449285,
"grad_norm": 1.631091833114624,
"learning_rate": 5.667480273473455e-05,
"loss": 2.59788330078125,
"step": 275500
},
{
"epoch": 1.7451344891687848,
"grad_norm": 1.7004138231277466,
"learning_rate": 5.659545226593199e-05,
"loss": 2.59458056640625,
"step": 276000
},
{
"epoch": 1.7482959646926415,
"grad_norm": 1.7346022129058838,
"learning_rate": 5.651610179712942e-05,
"loss": 2.59710595703125,
"step": 276500
},
{
"epoch": 1.7514574402164977,
"grad_norm": 1.6414403915405273,
"learning_rate": 5.643675132832685e-05,
"loss": 2.599799072265625,
"step": 277000
},
{
"epoch": 1.7546189157403544,
"grad_norm": 1.614408254623413,
"learning_rate": 5.635740085952428e-05,
"loss": 2.593843994140625,
"step": 277500
},
{
"epoch": 1.7577803912642107,
"grad_norm": 1.7073930501937866,
"learning_rate": 5.627805039072171e-05,
"loss": 2.59159814453125,
"step": 278000
},
{
"epoch": 1.7609418667880674,
"grad_norm": 1.6941301822662354,
"learning_rate": 5.619869992191914e-05,
"loss": 2.597547119140625,
"step": 278500
},
{
"epoch": 1.7641033423119237,
"grad_norm": 1.664022445678711,
"learning_rate": 5.611934945311658e-05,
"loss": 2.59936962890625,
"step": 279000
},
{
"epoch": 1.7672648178357804,
"grad_norm": 1.7130122184753418,
"learning_rate": 5.6039998984314e-05,
"loss": 2.59432373046875,
"step": 279500
},
{
"epoch": 1.7704262933596369,
"grad_norm": 1.6532748937606812,
"learning_rate": 5.5960648515511427e-05,
"loss": 2.594723388671875,
"step": 280000
},
{
"epoch": 1.7735877688834933,
"grad_norm": 1.6735934019088745,
"learning_rate": 5.5881298046708864e-05,
"loss": 2.594259033203125,
"step": 280500
},
{
"epoch": 1.7767492444073498,
"grad_norm": 1.60926353931427,
"learning_rate": 5.580194757790629e-05,
"loss": 2.593872314453125,
"step": 281000
},
{
"epoch": 1.7799107199312063,
"grad_norm": 1.7459192276000977,
"learning_rate": 5.5722597109103725e-05,
"loss": 2.5936220703125,
"step": 281500
},
{
"epoch": 1.7830721954550628,
"grad_norm": 1.6407498121261597,
"learning_rate": 5.564324664030115e-05,
"loss": 2.594974609375,
"step": 282000
},
{
"epoch": 1.7862336709789193,
"grad_norm": 1.6939260959625244,
"learning_rate": 5.5563896171498585e-05,
"loss": 2.58904931640625,
"step": 282500
},
{
"epoch": 1.7893951465027758,
"grad_norm": 1.7632529735565186,
"learning_rate": 5.548454570269601e-05,
"loss": 2.589850830078125,
"step": 283000
},
{
"epoch": 1.7925566220266322,
"grad_norm": 1.6395823955535889,
"learning_rate": 5.5405195233893446e-05,
"loss": 2.59438232421875,
"step": 283500
},
{
"epoch": 1.7957180975504887,
"grad_norm": 1.6666817665100098,
"learning_rate": 5.532584476509087e-05,
"loss": 2.59078076171875,
"step": 284000
},
{
"epoch": 1.7988795730743452,
"grad_norm": 1.690981149673462,
"learning_rate": 5.524649429628831e-05,
"loss": 2.58544091796875,
"step": 284500
},
{
"epoch": 1.802041048598202,
"grad_norm": 1.661834478378296,
"learning_rate": 5.516714382748574e-05,
"loss": 2.59174560546875,
"step": 285000
},
{
"epoch": 1.8052025241220582,
"grad_norm": 1.6878554821014404,
"learning_rate": 5.508779335868317e-05,
"loss": 2.58922021484375,
"step": 285500
},
{
"epoch": 1.8083639996459149,
"grad_norm": 1.6942951679229736,
"learning_rate": 5.50084428898806e-05,
"loss": 2.58524169921875,
"step": 286000
},
{
"epoch": 1.8115254751697711,
"grad_norm": 1.7384624481201172,
"learning_rate": 5.492909242107802e-05,
"loss": 2.5947763671875,
"step": 286500
},
{
"epoch": 1.8146869506936278,
"grad_norm": 1.6371098756790161,
"learning_rate": 5.484974195227546e-05,
"loss": 2.593004638671875,
"step": 287000
},
{
"epoch": 1.817848426217484,
"grad_norm": 1.6830705404281616,
"learning_rate": 5.477039148347288e-05,
"loss": 2.58744970703125,
"step": 287500
},
{
"epoch": 1.8210099017413408,
"grad_norm": 1.6495065689086914,
"learning_rate": 5.469104101467032e-05,
"loss": 2.586923828125,
"step": 288000
},
{
"epoch": 1.824171377265197,
"grad_norm": 1.6736867427825928,
"learning_rate": 5.4611690545867744e-05,
"loss": 2.58320947265625,
"step": 288500
},
{
"epoch": 1.8273328527890538,
"grad_norm": 1.7100701332092285,
"learning_rate": 5.453234007706518e-05,
"loss": 2.58325146484375,
"step": 289000
},
{
"epoch": 1.83049432831291,
"grad_norm": 1.7714853286743164,
"learning_rate": 5.4452989608262605e-05,
"loss": 2.584956787109375,
"step": 289500
},
{
"epoch": 1.8336558038367667,
"grad_norm": 1.6545295715332031,
"learning_rate": 5.437363913946004e-05,
"loss": 2.58028466796875,
"step": 290000
},
{
"epoch": 1.8368172793606232,
"grad_norm": 1.6884629726409912,
"learning_rate": 5.4294288670657465e-05,
"loss": 2.586012451171875,
"step": 290500
},
{
"epoch": 1.8399787548844797,
"grad_norm": 1.7388588190078735,
"learning_rate": 5.42149382018549e-05,
"loss": 2.5828623046875,
"step": 291000
},
{
"epoch": 1.8431402304083362,
"grad_norm": 1.7412656545639038,
"learning_rate": 5.4135587733052326e-05,
"loss": 2.580474609375,
"step": 291500
},
{
"epoch": 1.8463017059321927,
"grad_norm": 1.7147107124328613,
"learning_rate": 5.4056237264249763e-05,
"loss": 2.58169384765625,
"step": 292000
},
{
"epoch": 1.8494631814560492,
"grad_norm": 1.8051819801330566,
"learning_rate": 5.397688679544719e-05,
"loss": 2.58816796875,
"step": 292500
},
{
"epoch": 1.8526246569799056,
"grad_norm": 1.6592358350753784,
"learning_rate": 5.389753632664462e-05,
"loss": 2.579513427734375,
"step": 293000
},
{
"epoch": 1.8557861325037621,
"grad_norm": 1.7216485738754272,
"learning_rate": 5.3818185857842055e-05,
"loss": 2.580039794921875,
"step": 293500
},
{
"epoch": 1.8589476080276186,
"grad_norm": 1.7862902879714966,
"learning_rate": 5.373883538903948e-05,
"loss": 2.581174072265625,
"step": 294000
},
{
"epoch": 1.8621090835514753,
"grad_norm": 1.6968635320663452,
"learning_rate": 5.3659484920236916e-05,
"loss": 2.57760791015625,
"step": 294500
},
{
"epoch": 1.8652705590753316,
"grad_norm": 1.6866495609283447,
"learning_rate": 5.358013445143434e-05,
"loss": 2.584985107421875,
"step": 295000
},
{
"epoch": 1.8684320345991883,
"grad_norm": 1.7021806240081787,
"learning_rate": 5.3500783982631776e-05,
"loss": 2.57470361328125,
"step": 295500
},
{
"epoch": 1.8715935101230445,
"grad_norm": 1.6788595914840698,
"learning_rate": 5.34214335138292e-05,
"loss": 2.580240966796875,
"step": 296000
},
{
"epoch": 1.8747549856469012,
"grad_norm": 1.654101848602295,
"learning_rate": 5.334208304502664e-05,
"loss": 2.580859619140625,
"step": 296500
},
{
"epoch": 1.8779164611707575,
"grad_norm": 1.7617765665054321,
"learning_rate": 5.326273257622406e-05,
"loss": 2.578592041015625,
"step": 297000
},
{
"epoch": 1.8810779366946142,
"grad_norm": 1.6879280805587769,
"learning_rate": 5.31833821074215e-05,
"loss": 2.581620849609375,
"step": 297500
},
{
"epoch": 1.8842394122184705,
"grad_norm": 1.7235583066940308,
"learning_rate": 5.310403163861892e-05,
"loss": 2.576145263671875,
"step": 298000
},
{
"epoch": 1.8874008877423272,
"grad_norm": 1.7226355075836182,
"learning_rate": 5.302468116981636e-05,
"loss": 2.574718994140625,
"step": 298500
},
{
"epoch": 1.8905623632661834,
"grad_norm": 1.6874010562896729,
"learning_rate": 5.294533070101378e-05,
"loss": 2.574873046875,
"step": 299000
},
{
"epoch": 1.8937238387900401,
"grad_norm": 1.6694049835205078,
"learning_rate": 5.2865980232211206e-05,
"loss": 2.575375732421875,
"step": 299500
},
{
"epoch": 1.8968853143138966,
"grad_norm": 1.7782129049301147,
"learning_rate": 5.2786629763408643e-05,
"loss": 2.57757275390625,
"step": 300000
},
{
"epoch": 1.900046789837753,
"grad_norm": 1.6458760499954224,
"learning_rate": 5.2707279294606074e-05,
"loss": 2.574453857421875,
"step": 300500
},
{
"epoch": 1.9032082653616096,
"grad_norm": 1.7574435472488403,
"learning_rate": 5.2627928825803504e-05,
"loss": 2.57797314453125,
"step": 301000
},
{
"epoch": 1.906369740885466,
"grad_norm": 1.801379680633545,
"learning_rate": 5.2548578357000935e-05,
"loss": 2.57554833984375,
"step": 301500
},
{
"epoch": 1.9095312164093226,
"grad_norm": 1.8678525686264038,
"learning_rate": 5.246922788819837e-05,
"loss": 2.576422607421875,
"step": 302000
},
{
"epoch": 1.912692691933179,
"grad_norm": 1.6539779901504517,
"learning_rate": 5.2389877419395796e-05,
"loss": 2.571752685546875,
"step": 302500
},
{
"epoch": 1.9158541674570355,
"grad_norm": 1.7388038635253906,
"learning_rate": 5.231052695059323e-05,
"loss": 2.574066650390625,
"step": 303000
},
{
"epoch": 1.919015642980892,
"grad_norm": 1.7183165550231934,
"learning_rate": 5.2231176481790656e-05,
"loss": 2.575809326171875,
"step": 303500
},
{
"epoch": 1.9221771185047485,
"grad_norm": 1.6397099494934082,
"learning_rate": 5.2151826012988094e-05,
"loss": 2.567980712890625,
"step": 304000
},
{
"epoch": 1.925338594028605,
"grad_norm": 1.712165117263794,
"learning_rate": 5.207247554418552e-05,
"loss": 2.571453125,
"step": 304500
},
{
"epoch": 1.9285000695524617,
"grad_norm": 1.6544610261917114,
"learning_rate": 5.199312507538294e-05,
"loss": 2.575832275390625,
"step": 305000
},
{
"epoch": 1.931661545076318,
"grad_norm": 1.713889241218567,
"learning_rate": 5.191377460658038e-05,
"loss": 2.56847705078125,
"step": 305500
},
{
"epoch": 1.9348230206001746,
"grad_norm": 1.6941771507263184,
"learning_rate": 5.18344241377778e-05,
"loss": 2.57155224609375,
"step": 306000
},
{
"epoch": 1.937984496124031,
"grad_norm": 1.6438218355178833,
"learning_rate": 5.175507366897524e-05,
"loss": 2.56791015625,
"step": 306500
},
{
"epoch": 1.9411459716478876,
"grad_norm": 1.6573896408081055,
"learning_rate": 5.167572320017266e-05,
"loss": 2.572308837890625,
"step": 307000
},
{
"epoch": 1.9443074471717439,
"grad_norm": 1.6780575513839722,
"learning_rate": 5.15963727313701e-05,
"loss": 2.568292724609375,
"step": 307500
},
{
"epoch": 1.9474689226956006,
"grad_norm": 1.7285305261611938,
"learning_rate": 5.151702226256753e-05,
"loss": 2.56894970703125,
"step": 308000
},
{
"epoch": 1.9506303982194568,
"grad_norm": 1.6299684047698975,
"learning_rate": 5.143767179376496e-05,
"loss": 2.564464111328125,
"step": 308500
},
{
"epoch": 1.9537918737433135,
"grad_norm": 1.677272915840149,
"learning_rate": 5.135832132496239e-05,
"loss": 2.569608154296875,
"step": 309000
},
{
"epoch": 1.95695334926717,
"grad_norm": 1.7460423707962036,
"learning_rate": 5.127897085615982e-05,
"loss": 2.567556396484375,
"step": 309500
},
{
"epoch": 1.9601148247910265,
"grad_norm": 1.708332896232605,
"learning_rate": 5.119962038735725e-05,
"loss": 2.566654541015625,
"step": 310000
},
{
"epoch": 1.963276300314883,
"grad_norm": 1.8337763547897339,
"learning_rate": 5.112026991855469e-05,
"loss": 2.56631201171875,
"step": 310500
},
{
"epoch": 1.9664377758387395,
"grad_norm": 1.6483570337295532,
"learning_rate": 5.104091944975211e-05,
"loss": 2.568677978515625,
"step": 311000
},
{
"epoch": 1.969599251362596,
"grad_norm": 1.655570149421692,
"learning_rate": 5.0961568980949536e-05,
"loss": 2.566974609375,
"step": 311500
},
{
"epoch": 1.9727607268864524,
"grad_norm": 1.685663104057312,
"learning_rate": 5.0882218512146974e-05,
"loss": 2.5699580078125,
"step": 312000
},
{
"epoch": 1.975922202410309,
"grad_norm": 1.7235316038131714,
"learning_rate": 5.08028680433444e-05,
"loss": 2.561189208984375,
"step": 312500
},
{
"epoch": 1.9790836779341654,
"grad_norm": 1.6979799270629883,
"learning_rate": 5.0723517574541834e-05,
"loss": 2.568962890625,
"step": 313000
},
{
"epoch": 1.9822451534580219,
"grad_norm": 1.6781136989593506,
"learning_rate": 5.064416710573926e-05,
"loss": 2.559887939453125,
"step": 313500
},
{
"epoch": 1.9854066289818784,
"grad_norm": 1.7153162956237793,
"learning_rate": 5.0564816636936695e-05,
"loss": 2.5594306640625,
"step": 314000
},
{
"epoch": 1.988568104505735,
"grad_norm": 1.7000346183776855,
"learning_rate": 5.048546616813412e-05,
"loss": 2.564354248046875,
"step": 314500
},
{
"epoch": 1.9917295800295913,
"grad_norm": 1.694844126701355,
"learning_rate": 5.0406115699331556e-05,
"loss": 2.5614306640625,
"step": 315000
},
{
"epoch": 1.994891055553448,
"grad_norm": 1.641352891921997,
"learning_rate": 5.032676523052898e-05,
"loss": 2.56616259765625,
"step": 315500
},
{
"epoch": 1.9980525310773043,
"grad_norm": 1.74703049659729,
"learning_rate": 5.024741476172642e-05,
"loss": 2.562424072265625,
"step": 316000
},
{
"epoch": 2.001214006601161,
"grad_norm": 1.645979642868042,
"learning_rate": 5.016806429292385e-05,
"loss": 2.553907470703125,
"step": 316500
},
{
"epoch": 2.0043754821250173,
"grad_norm": 1.6632643938064575,
"learning_rate": 5.008871382412128e-05,
"loss": 2.561162109375,
"step": 317000
},
{
"epoch": 2.007536957648874,
"grad_norm": 1.6237820386886597,
"learning_rate": 5.000936335531871e-05,
"loss": 2.555327392578125,
"step": 317500
},
{
"epoch": 2.0106984331727302,
"grad_norm": 1.7072020769119263,
"learning_rate": 4.993001288651614e-05,
"loss": 2.555271240234375,
"step": 318000
},
{
"epoch": 2.013859908696587,
"grad_norm": 1.6668188571929932,
"learning_rate": 4.985066241771357e-05,
"loss": 2.55575634765625,
"step": 318500
},
{
"epoch": 2.017021384220443,
"grad_norm": 1.7508553266525269,
"learning_rate": 4.9771311948911e-05,
"loss": 2.55926953125,
"step": 319000
},
{
"epoch": 2.0201828597443,
"grad_norm": 1.7401902675628662,
"learning_rate": 4.969196148010843e-05,
"loss": 2.5536025390625,
"step": 319500
},
{
"epoch": 2.023344335268156,
"grad_norm": 1.682224988937378,
"learning_rate": 4.961261101130586e-05,
"loss": 2.55698779296875,
"step": 320000
},
{
"epoch": 2.026505810792013,
"grad_norm": 1.7012430429458618,
"learning_rate": 4.9533260542503284e-05,
"loss": 2.55514892578125,
"step": 320500
},
{
"epoch": 2.029667286315869,
"grad_norm": 1.7654048204421997,
"learning_rate": 4.9453910073700714e-05,
"loss": 2.5575859375,
"step": 321000
},
{
"epoch": 2.032828761839726,
"grad_norm": 1.7257287502288818,
"learning_rate": 4.9374559604898145e-05,
"loss": 2.551832275390625,
"step": 321500
},
{
"epoch": 2.0359902373635825,
"grad_norm": 1.7174879312515259,
"learning_rate": 4.9295209136095575e-05,
"loss": 2.5605341796875,
"step": 322000
},
{
"epoch": 2.039151712887439,
"grad_norm": 1.7239501476287842,
"learning_rate": 4.9215858667293006e-05,
"loss": 2.55483642578125,
"step": 322500
},
{
"epoch": 2.0423131884112955,
"grad_norm": 1.711594581604004,
"learning_rate": 4.9136508198490436e-05,
"loss": 2.560322998046875,
"step": 323000
},
{
"epoch": 2.0454746639351518,
"grad_norm": 1.6915425062179565,
"learning_rate": 4.9057157729687867e-05,
"loss": 2.553174072265625,
"step": 323500
},
{
"epoch": 2.0486361394590085,
"grad_norm": 1.7242716550827026,
"learning_rate": 4.89778072608853e-05,
"loss": 2.550463623046875,
"step": 324000
},
{
"epoch": 2.0517976149828647,
"grad_norm": 1.7377326488494873,
"learning_rate": 4.889845679208273e-05,
"loss": 2.555609619140625,
"step": 324500
},
{
"epoch": 2.0549590905067214,
"grad_norm": 1.7084745168685913,
"learning_rate": 4.8819106323280165e-05,
"loss": 2.555294921875,
"step": 325000
},
{
"epoch": 2.0581205660305777,
"grad_norm": 1.6371990442276,
"learning_rate": 4.8739755854477595e-05,
"loss": 2.5522607421875,
"step": 325500
},
{
"epoch": 2.0612820415544344,
"grad_norm": 1.7225629091262817,
"learning_rate": 4.8660405385675025e-05,
"loss": 2.55627197265625,
"step": 326000
},
{
"epoch": 2.0644435170782907,
"grad_norm": 1.688609004020691,
"learning_rate": 4.8581054916872456e-05,
"loss": 2.55092431640625,
"step": 326500
},
{
"epoch": 2.0676049926021474,
"grad_norm": 1.700508952140808,
"learning_rate": 4.850170444806988e-05,
"loss": 2.54750830078125,
"step": 327000
},
{
"epoch": 2.0707664681260036,
"grad_norm": 1.722130537033081,
"learning_rate": 4.842235397926731e-05,
"loss": 2.54960107421875,
"step": 327500
},
{
"epoch": 2.0739279436498603,
"grad_norm": 1.714810848236084,
"learning_rate": 4.834300351046474e-05,
"loss": 2.552972412109375,
"step": 328000
},
{
"epoch": 2.0770894191737166,
"grad_norm": 1.7323644161224365,
"learning_rate": 4.826365304166217e-05,
"loss": 2.55630126953125,
"step": 328500
},
{
"epoch": 2.0802508946975733,
"grad_norm": 1.675295114517212,
"learning_rate": 4.81843025728596e-05,
"loss": 2.54599365234375,
"step": 329000
},
{
"epoch": 2.0834123702214296,
"grad_norm": 1.7124736309051514,
"learning_rate": 4.810495210405703e-05,
"loss": 2.545865234375,
"step": 329500
},
{
"epoch": 2.0865738457452863,
"grad_norm": 1.6915316581726074,
"learning_rate": 4.802560163525446e-05,
"loss": 2.550764892578125,
"step": 330000
},
{
"epoch": 2.0897353212691425,
"grad_norm": 1.7371670007705688,
"learning_rate": 4.794625116645189e-05,
"loss": 2.545845947265625,
"step": 330500
},
{
"epoch": 2.0928967967929992,
"grad_norm": 1.793949007987976,
"learning_rate": 4.786690069764932e-05,
"loss": 2.550509765625,
"step": 331000
},
{
"epoch": 2.096058272316856,
"grad_norm": 1.7230401039123535,
"learning_rate": 4.778755022884675e-05,
"loss": 2.5508291015625,
"step": 331500
},
{
"epoch": 2.099219747840712,
"grad_norm": 1.7888908386230469,
"learning_rate": 4.7708199760044184e-05,
"loss": 2.544958984375,
"step": 332000
},
{
"epoch": 2.102381223364569,
"grad_norm": 1.6981281042099,
"learning_rate": 4.7628849291241614e-05,
"loss": 2.552451904296875,
"step": 332500
},
{
"epoch": 2.105542698888425,
"grad_norm": 1.6763311624526978,
"learning_rate": 4.7549498822439045e-05,
"loss": 2.547066162109375,
"step": 333000
},
{
"epoch": 2.108704174412282,
"grad_norm": 1.7012392282485962,
"learning_rate": 4.7470148353636475e-05,
"loss": 2.551265380859375,
"step": 333500
},
{
"epoch": 2.111865649936138,
"grad_norm": 1.7901865243911743,
"learning_rate": 4.7390797884833905e-05,
"loss": 2.543422607421875,
"step": 334000
},
{
"epoch": 2.115027125459995,
"grad_norm": 1.6667346954345703,
"learning_rate": 4.7311447416031336e-05,
"loss": 2.54710302734375,
"step": 334500
},
{
"epoch": 2.118188600983851,
"grad_norm": 1.6894967555999756,
"learning_rate": 4.7232096947228766e-05,
"loss": 2.543498046875,
"step": 335000
},
{
"epoch": 2.121350076507708,
"grad_norm": 1.6717145442962646,
"learning_rate": 4.71527464784262e-05,
"loss": 2.547709716796875,
"step": 335500
},
{
"epoch": 2.124511552031564,
"grad_norm": 1.7791911363601685,
"learning_rate": 4.707339600962363e-05,
"loss": 2.544640625,
"step": 336000
},
{
"epoch": 2.1276730275554208,
"grad_norm": 1.774796724319458,
"learning_rate": 4.699404554082106e-05,
"loss": 2.547943115234375,
"step": 336500
},
{
"epoch": 2.130834503079277,
"grad_norm": 1.7801834344863892,
"learning_rate": 4.691469507201849e-05,
"loss": 2.54091650390625,
"step": 337000
},
{
"epoch": 2.1339959786031337,
"grad_norm": 1.7217036485671997,
"learning_rate": 4.683534460321592e-05,
"loss": 2.54168896484375,
"step": 337500
},
{
"epoch": 2.13715745412699,
"grad_norm": 1.8133459091186523,
"learning_rate": 4.675599413441335e-05,
"loss": 2.546461669921875,
"step": 338000
},
{
"epoch": 2.1403189296508467,
"grad_norm": 1.7895547151565552,
"learning_rate": 4.667664366561078e-05,
"loss": 2.542771240234375,
"step": 338500
},
{
"epoch": 2.143480405174703,
"grad_norm": 1.6930867433547974,
"learning_rate": 4.659729319680821e-05,
"loss": 2.53901416015625,
"step": 339000
},
{
"epoch": 2.1466418806985597,
"grad_norm": 1.6707184314727783,
"learning_rate": 4.651794272800564e-05,
"loss": 2.540096435546875,
"step": 339500
},
{
"epoch": 2.1498033562224164,
"grad_norm": 1.6871057748794556,
"learning_rate": 4.643859225920307e-05,
"loss": 2.54280810546875,
"step": 340000
},
{
"epoch": 2.1529648317462726,
"grad_norm": 1.6985068321228027,
"learning_rate": 4.63592417904005e-05,
"loss": 2.542167724609375,
"step": 340500
},
{
"epoch": 2.156126307270129,
"grad_norm": 1.8134102821350098,
"learning_rate": 4.627989132159793e-05,
"loss": 2.54412255859375,
"step": 341000
},
{
"epoch": 2.1592877827939856,
"grad_norm": 1.6955080032348633,
"learning_rate": 4.620054085279536e-05,
"loss": 2.53557421875,
"step": 341500
},
{
"epoch": 2.1624492583178423,
"grad_norm": 1.6869953870773315,
"learning_rate": 4.612119038399279e-05,
"loss": 2.54038037109375,
"step": 342000
},
{
"epoch": 2.1656107338416986,
"grad_norm": 1.7373355627059937,
"learning_rate": 4.604183991519022e-05,
"loss": 2.542293701171875,
"step": 342500
},
{
"epoch": 2.1687722093655553,
"grad_norm": 1.7545257806777954,
"learning_rate": 4.596248944638765e-05,
"loss": 2.540019287109375,
"step": 343000
},
{
"epoch": 2.1719336848894115,
"grad_norm": 1.7184209823608398,
"learning_rate": 4.5883138977585084e-05,
"loss": 2.54190087890625,
"step": 343500
},
{
"epoch": 2.1750951604132682,
"grad_norm": 1.700035572052002,
"learning_rate": 4.5803788508782514e-05,
"loss": 2.5383876953125,
"step": 344000
},
{
"epoch": 2.1782566359371245,
"grad_norm": 1.7379525899887085,
"learning_rate": 4.5724438039979944e-05,
"loss": 2.5402470703125,
"step": 344500
},
{
"epoch": 2.181418111460981,
"grad_norm": 1.7505565881729126,
"learning_rate": 4.5645087571177375e-05,
"loss": 2.53586962890625,
"step": 345000
},
{
"epoch": 2.1845795869848375,
"grad_norm": 1.7425771951675415,
"learning_rate": 4.55657371023748e-05,
"loss": 2.537313232421875,
"step": 345500
},
{
"epoch": 2.187741062508694,
"grad_norm": 1.8148465156555176,
"learning_rate": 4.548638663357223e-05,
"loss": 2.5388896484375,
"step": 346000
},
{
"epoch": 2.1909025380325504,
"grad_norm": 1.645362138748169,
"learning_rate": 4.540703616476966e-05,
"loss": 2.538688720703125,
"step": 346500
},
{
"epoch": 2.194064013556407,
"grad_norm": 1.8049070835113525,
"learning_rate": 4.532768569596709e-05,
"loss": 2.5358134765625,
"step": 347000
},
{
"epoch": 2.1972254890802634,
"grad_norm": 1.7948477268218994,
"learning_rate": 4.524833522716452e-05,
"loss": 2.53858349609375,
"step": 347500
},
{
"epoch": 2.20038696460412,
"grad_norm": 1.8996093273162842,
"learning_rate": 4.516898475836196e-05,
"loss": 2.53972802734375,
"step": 348000
},
{
"epoch": 2.2035484401279763,
"grad_norm": 1.6910823583602905,
"learning_rate": 4.508963428955939e-05,
"loss": 2.533819580078125,
"step": 348500
},
{
"epoch": 2.206709915651833,
"grad_norm": 1.757420301437378,
"learning_rate": 4.501028382075682e-05,
"loss": 2.538610107421875,
"step": 349000
},
{
"epoch": 2.2098713911756893,
"grad_norm": 1.7346962690353394,
"learning_rate": 4.493093335195425e-05,
"loss": 2.536515625,
"step": 349500
},
{
"epoch": 2.213032866699546,
"grad_norm": 1.7200614213943481,
"learning_rate": 4.485158288315168e-05,
"loss": 2.535265625,
"step": 350000
},
{
"epoch": 2.2161943422234023,
"grad_norm": 1.7531795501708984,
"learning_rate": 4.477223241434911e-05,
"loss": 2.542253173828125,
"step": 350500
},
{
"epoch": 2.219355817747259,
"grad_norm": 1.6730948686599731,
"learning_rate": 4.469288194554654e-05,
"loss": 2.53125732421875,
"step": 351000
},
{
"epoch": 2.2225172932711157,
"grad_norm": 1.8358124494552612,
"learning_rate": 4.461353147674397e-05,
"loss": 2.53560986328125,
"step": 351500
},
{
"epoch": 2.225678768794972,
"grad_norm": 1.7008156776428223,
"learning_rate": 4.4534181007941394e-05,
"loss": 2.530701171875,
"step": 352000
},
{
"epoch": 2.2288402443188287,
"grad_norm": 1.7278190851211548,
"learning_rate": 4.4454830539138824e-05,
"loss": 2.531859130859375,
"step": 352500
},
{
"epoch": 2.232001719842685,
"grad_norm": 1.7624832391738892,
"learning_rate": 4.4375480070336255e-05,
"loss": 2.53247998046875,
"step": 353000
},
{
"epoch": 2.2351631953665416,
"grad_norm": 1.6993699073791504,
"learning_rate": 4.4296129601533685e-05,
"loss": 2.536030029296875,
"step": 353500
},
{
"epoch": 2.238324670890398,
"grad_norm": 1.7039874792099,
"learning_rate": 4.4216779132731116e-05,
"loss": 2.529044189453125,
"step": 354000
},
{
"epoch": 2.2414861464142546,
"grad_norm": 1.7265779972076416,
"learning_rate": 4.4137428663928546e-05,
"loss": 2.5344541015625,
"step": 354500
},
{
"epoch": 2.244647621938111,
"grad_norm": 1.7709705829620361,
"learning_rate": 4.4058078195125976e-05,
"loss": 2.527215576171875,
"step": 355000
},
{
"epoch": 2.2478090974619676,
"grad_norm": 1.7799522876739502,
"learning_rate": 4.397872772632341e-05,
"loss": 2.530477783203125,
"step": 355500
},
{
"epoch": 2.250970572985824,
"grad_norm": 1.7330743074417114,
"learning_rate": 4.389937725752084e-05,
"loss": 2.53780126953125,
"step": 356000
},
{
"epoch": 2.2541320485096805,
"grad_norm": 1.8422235250473022,
"learning_rate": 4.3820026788718275e-05,
"loss": 2.533418701171875,
"step": 356500
},
{
"epoch": 2.257293524033537,
"grad_norm": 1.7150598764419556,
"learning_rate": 4.3740676319915705e-05,
"loss": 2.529114013671875,
"step": 357000
},
{
"epoch": 2.2604549995573935,
"grad_norm": 1.7712043523788452,
"learning_rate": 4.3661325851113135e-05,
"loss": 2.5250224609375,
"step": 357500
},
{
"epoch": 2.2636164750812497,
"grad_norm": 1.744796633720398,
"learning_rate": 4.358197538231056e-05,
"loss": 2.52585888671875,
"step": 358000
},
{
"epoch": 2.2667779506051065,
"grad_norm": 1.7289003133773804,
"learning_rate": 4.350262491350799e-05,
"loss": 2.52629833984375,
"step": 358500
},
{
"epoch": 2.269939426128963,
"grad_norm": 1.6766456365585327,
"learning_rate": 4.342327444470542e-05,
"loss": 2.52869384765625,
"step": 359000
},
{
"epoch": 2.2731009016528194,
"grad_norm": 1.8103229999542236,
"learning_rate": 4.334392397590285e-05,
"loss": 2.526737060546875,
"step": 359500
},
{
"epoch": 2.2762623771766757,
"grad_norm": 1.7055565118789673,
"learning_rate": 4.326457350710028e-05,
"loss": 2.528530029296875,
"step": 360000
},
{
"epoch": 2.2794238527005324,
"grad_norm": 1.7404392957687378,
"learning_rate": 4.318522303829771e-05,
"loss": 2.531879638671875,
"step": 360500
},
{
"epoch": 2.282585328224389,
"grad_norm": 1.6999282836914062,
"learning_rate": 4.310587256949514e-05,
"loss": 2.525703369140625,
"step": 361000
},
{
"epoch": 2.2857468037482453,
"grad_norm": 1.716452956199646,
"learning_rate": 4.302652210069257e-05,
"loss": 2.53142041015625,
"step": 361500
},
{
"epoch": 2.288908279272102,
"grad_norm": 1.6946918964385986,
"learning_rate": 4.294717163189e-05,
"loss": 2.527634521484375,
"step": 362000
},
{
"epoch": 2.2920697547959583,
"grad_norm": 1.7284834384918213,
"learning_rate": 4.286782116308743e-05,
"loss": 2.531074462890625,
"step": 362500
},
{
"epoch": 2.295231230319815,
"grad_norm": 1.7445048093795776,
"learning_rate": 4.278847069428486e-05,
"loss": 2.528794189453125,
"step": 363000
},
{
"epoch": 2.2983927058436713,
"grad_norm": 1.7417007684707642,
"learning_rate": 4.2709120225482294e-05,
"loss": 2.525464111328125,
"step": 363500
},
{
"epoch": 2.301554181367528,
"grad_norm": 1.7060586214065552,
"learning_rate": 4.2629769756679724e-05,
"loss": 2.52593212890625,
"step": 364000
},
{
"epoch": 2.3047156568913842,
"grad_norm": 1.768139123916626,
"learning_rate": 4.2550419287877155e-05,
"loss": 2.5231650390625,
"step": 364500
},
{
"epoch": 2.307877132415241,
"grad_norm": 1.6929272413253784,
"learning_rate": 4.2471068819074585e-05,
"loss": 2.52467138671875,
"step": 365000
},
{
"epoch": 2.311038607939097,
"grad_norm": 1.6404354572296143,
"learning_rate": 4.2391718350272015e-05,
"loss": 2.52601220703125,
"step": 365500
},
{
"epoch": 2.314200083462954,
"grad_norm": 1.7828208208084106,
"learning_rate": 4.2312367881469446e-05,
"loss": 2.522073486328125,
"step": 366000
},
{
"epoch": 2.31736155898681,
"grad_norm": 1.7895129919052124,
"learning_rate": 4.2233017412666876e-05,
"loss": 2.52412353515625,
"step": 366500
},
{
"epoch": 2.320523034510667,
"grad_norm": 1.6818422079086304,
"learning_rate": 4.215366694386431e-05,
"loss": 2.525711181640625,
"step": 367000
},
{
"epoch": 2.323684510034523,
"grad_norm": 1.7184971570968628,
"learning_rate": 4.207431647506174e-05,
"loss": 2.5273662109375,
"step": 367500
},
{
"epoch": 2.32684598555838,
"grad_norm": 1.7267197370529175,
"learning_rate": 4.199496600625917e-05,
"loss": 2.522635498046875,
"step": 368000
},
{
"epoch": 2.330007461082236,
"grad_norm": 1.7006652355194092,
"learning_rate": 4.19156155374566e-05,
"loss": 2.52476513671875,
"step": 368500
},
{
"epoch": 2.333168936606093,
"grad_norm": 1.7603662014007568,
"learning_rate": 4.183626506865403e-05,
"loss": 2.5251904296875,
"step": 369000
},
{
"epoch": 2.336330412129949,
"grad_norm": 1.8043859004974365,
"learning_rate": 4.175691459985146e-05,
"loss": 2.5203720703125,
"step": 369500
},
{
"epoch": 2.339491887653806,
"grad_norm": 1.739791750907898,
"learning_rate": 4.167756413104889e-05,
"loss": 2.52461767578125,
"step": 370000
},
{
"epoch": 2.3426533631776625,
"grad_norm": 1.7414002418518066,
"learning_rate": 4.159821366224631e-05,
"loss": 2.527456298828125,
"step": 370500
},
{
"epoch": 2.3458148387015187,
"grad_norm": 1.6841999292373657,
"learning_rate": 4.151886319344374e-05,
"loss": 2.52438671875,
"step": 371000
},
{
"epoch": 2.3489763142253755,
"grad_norm": 1.7478948831558228,
"learning_rate": 4.143951272464118e-05,
"loss": 2.5241103515625,
"step": 371500
},
{
"epoch": 2.3521377897492317,
"grad_norm": 1.7525583505630493,
"learning_rate": 4.136016225583861e-05,
"loss": 2.5187880859375,
"step": 372000
},
{
"epoch": 2.3552992652730884,
"grad_norm": 1.7401121854782104,
"learning_rate": 4.128081178703604e-05,
"loss": 2.51843017578125,
"step": 372500
},
{
"epoch": 2.3584607407969447,
"grad_norm": 1.6875884532928467,
"learning_rate": 4.120146131823347e-05,
"loss": 2.518933837890625,
"step": 373000
},
{
"epoch": 2.3616222163208014,
"grad_norm": 1.7762725353240967,
"learning_rate": 4.11221108494309e-05,
"loss": 2.523000244140625,
"step": 373500
},
{
"epoch": 2.3647836918446576,
"grad_norm": 1.6478955745697021,
"learning_rate": 4.104276038062833e-05,
"loss": 2.52421044921875,
"step": 374000
},
{
"epoch": 2.3679451673685143,
"grad_norm": 1.7773512601852417,
"learning_rate": 4.096340991182576e-05,
"loss": 2.517908203125,
"step": 374500
},
{
"epoch": 2.3711066428923706,
"grad_norm": 1.7414214611053467,
"learning_rate": 4.0884059443023193e-05,
"loss": 2.52295166015625,
"step": 375000
},
{
"epoch": 2.3742681184162273,
"grad_norm": 1.7533800601959229,
"learning_rate": 4.0804708974220624e-05,
"loss": 2.5224453125,
"step": 375500
},
{
"epoch": 2.3774295939400836,
"grad_norm": 1.7633564472198486,
"learning_rate": 4.0725358505418054e-05,
"loss": 2.51393896484375,
"step": 376000
},
{
"epoch": 2.3805910694639403,
"grad_norm": 1.7669663429260254,
"learning_rate": 4.0646008036615485e-05,
"loss": 2.519831787109375,
"step": 376500
},
{
"epoch": 2.3837525449877965,
"grad_norm": 1.82060968875885,
"learning_rate": 4.056665756781291e-05,
"loss": 2.520356689453125,
"step": 377000
},
{
"epoch": 2.3869140205116532,
"grad_norm": 1.7347828149795532,
"learning_rate": 4.048730709901034e-05,
"loss": 2.5166962890625,
"step": 377500
},
{
"epoch": 2.3900754960355095,
"grad_norm": 1.7392734289169312,
"learning_rate": 4.040795663020777e-05,
"loss": 2.5205,
"step": 378000
},
{
"epoch": 2.393236971559366,
"grad_norm": 1.6537641286849976,
"learning_rate": 4.03286061614052e-05,
"loss": 2.518643310546875,
"step": 378500
},
{
"epoch": 2.3963984470832225,
"grad_norm": 1.739017128944397,
"learning_rate": 4.024925569260263e-05,
"loss": 2.51444775390625,
"step": 379000
},
{
"epoch": 2.399559922607079,
"grad_norm": 1.6847864389419556,
"learning_rate": 4.016990522380007e-05,
"loss": 2.515763916015625,
"step": 379500
},
{
"epoch": 2.402721398130936,
"grad_norm": 1.735995888710022,
"learning_rate": 4.00905547549975e-05,
"loss": 2.51568115234375,
"step": 380000
},
{
"epoch": 2.405882873654792,
"grad_norm": 1.7095977067947388,
"learning_rate": 4.001120428619493e-05,
"loss": 2.5167392578125,
"step": 380500
},
{
"epoch": 2.409044349178649,
"grad_norm": 1.7414978742599487,
"learning_rate": 3.993185381739236e-05,
"loss": 2.50940869140625,
"step": 381000
},
{
"epoch": 2.412205824702505,
"grad_norm": 1.779842734336853,
"learning_rate": 3.985250334858979e-05,
"loss": 2.5131240234375,
"step": 381500
},
{
"epoch": 2.415367300226362,
"grad_norm": 1.7040811777114868,
"learning_rate": 3.977315287978722e-05,
"loss": 2.51479248046875,
"step": 382000
},
{
"epoch": 2.418528775750218,
"grad_norm": 1.8120027780532837,
"learning_rate": 3.969380241098465e-05,
"loss": 2.510103515625,
"step": 382500
},
{
"epoch": 2.421690251274075,
"grad_norm": 1.7283341884613037,
"learning_rate": 3.9614451942182073e-05,
"loss": 2.510373046875,
"step": 383000
},
{
"epoch": 2.424851726797931,
"grad_norm": 1.6391911506652832,
"learning_rate": 3.9535101473379504e-05,
"loss": 2.513380615234375,
"step": 383500
},
{
"epoch": 2.4280132023217877,
"grad_norm": 1.7967621088027954,
"learning_rate": 3.9455751004576934e-05,
"loss": 2.51839453125,
"step": 384000
},
{
"epoch": 2.431174677845644,
"grad_norm": 1.7125979661941528,
"learning_rate": 3.9376400535774365e-05,
"loss": 2.516982177734375,
"step": 384500
},
{
"epoch": 2.4343361533695007,
"grad_norm": 1.7376694679260254,
"learning_rate": 3.9297050066971795e-05,
"loss": 2.511137939453125,
"step": 385000
},
{
"epoch": 2.437497628893357,
"grad_norm": 1.7806309461593628,
"learning_rate": 3.9217699598169226e-05,
"loss": 2.51102783203125,
"step": 385500
},
{
"epoch": 2.4406591044172137,
"grad_norm": 1.7650015354156494,
"learning_rate": 3.9138349129366656e-05,
"loss": 2.5102392578125,
"step": 386000
},
{
"epoch": 2.44382057994107,
"grad_norm": 1.779952883720398,
"learning_rate": 3.9058998660564086e-05,
"loss": 2.5139140625,
"step": 386500
},
{
"epoch": 2.4469820554649266,
"grad_norm": 1.758278250694275,
"learning_rate": 3.897964819176152e-05,
"loss": 2.51097900390625,
"step": 387000
},
{
"epoch": 2.450143530988783,
"grad_norm": 1.7586311101913452,
"learning_rate": 3.890029772295895e-05,
"loss": 2.511375732421875,
"step": 387500
},
{
"epoch": 2.4533050065126396,
"grad_norm": 1.8121308088302612,
"learning_rate": 3.8820947254156384e-05,
"loss": 2.51412060546875,
"step": 388000
},
{
"epoch": 2.456466482036496,
"grad_norm": 1.8297014236450195,
"learning_rate": 3.8741596785353815e-05,
"loss": 2.511367431640625,
"step": 388500
},
{
"epoch": 2.4596279575603526,
"grad_norm": 1.721234679222107,
"learning_rate": 3.8662246316551245e-05,
"loss": 2.51208837890625,
"step": 389000
},
{
"epoch": 2.4627894330842093,
"grad_norm": 1.7084195613861084,
"learning_rate": 3.858289584774867e-05,
"loss": 2.5123017578125,
"step": 389500
},
{
"epoch": 2.4659509086080655,
"grad_norm": 1.7288706302642822,
"learning_rate": 3.85035453789461e-05,
"loss": 2.506217529296875,
"step": 390000
},
{
"epoch": 2.4691123841319222,
"grad_norm": 1.8153363466262817,
"learning_rate": 3.842419491014353e-05,
"loss": 2.509179443359375,
"step": 390500
},
{
"epoch": 2.4722738596557785,
"grad_norm": 1.804592251777649,
"learning_rate": 3.834484444134096e-05,
"loss": 2.5018798828125,
"step": 391000
},
{
"epoch": 2.475435335179635,
"grad_norm": 1.7740651369094849,
"learning_rate": 3.826549397253839e-05,
"loss": 2.503543212890625,
"step": 391500
},
{
"epoch": 2.4785968107034915,
"grad_norm": 1.738162636756897,
"learning_rate": 3.818614350373582e-05,
"loss": 2.509318115234375,
"step": 392000
},
{
"epoch": 2.481758286227348,
"grad_norm": 1.8145672082901,
"learning_rate": 3.810679303493325e-05,
"loss": 2.510130615234375,
"step": 392500
},
{
"epoch": 2.4849197617512044,
"grad_norm": 1.7516227960586548,
"learning_rate": 3.802744256613068e-05,
"loss": 2.509041748046875,
"step": 393000
},
{
"epoch": 2.488081237275061,
"grad_norm": 1.7258633375167847,
"learning_rate": 3.794809209732811e-05,
"loss": 2.509795166015625,
"step": 393500
},
{
"epoch": 2.4912427127989174,
"grad_norm": 1.8278143405914307,
"learning_rate": 3.786874162852554e-05,
"loss": 2.51022412109375,
"step": 394000
},
{
"epoch": 2.494404188322774,
"grad_norm": 1.7449432611465454,
"learning_rate": 3.778939115972297e-05,
"loss": 2.50709521484375,
"step": 394500
},
{
"epoch": 2.4975656638466304,
"grad_norm": 1.7232692241668701,
"learning_rate": 3.7710040690920404e-05,
"loss": 2.50889990234375,
"step": 395000
},
{
"epoch": 2.500727139370487,
"grad_norm": 1.8552279472351074,
"learning_rate": 3.7630690222117834e-05,
"loss": 2.507832763671875,
"step": 395500
},
{
"epoch": 2.5038886148943433,
"grad_norm": 1.7707123756408691,
"learning_rate": 3.7551339753315264e-05,
"loss": 2.50444140625,
"step": 396000
},
{
"epoch": 2.5070500904182,
"grad_norm": 1.7885921001434326,
"learning_rate": 3.7471989284512695e-05,
"loss": 2.506232666015625,
"step": 396500
},
{
"epoch": 2.5102115659420567,
"grad_norm": 1.825189232826233,
"learning_rate": 3.7392638815710125e-05,
"loss": 2.50288623046875,
"step": 397000
},
{
"epoch": 2.513373041465913,
"grad_norm": 1.7198543548583984,
"learning_rate": 3.7313288346907556e-05,
"loss": 2.5014560546875,
"step": 397500
},
{
"epoch": 2.5165345169897693,
"grad_norm": 1.7541420459747314,
"learning_rate": 3.7233937878104986e-05,
"loss": 2.50422509765625,
"step": 398000
},
{
"epoch": 2.519695992513626,
"grad_norm": 1.7437357902526855,
"learning_rate": 3.7154587409302417e-05,
"loss": 2.50784521484375,
"step": 398500
},
{
"epoch": 2.5228574680374827,
"grad_norm": 1.8088346719741821,
"learning_rate": 3.707523694049985e-05,
"loss": 2.50709716796875,
"step": 399000
},
{
"epoch": 2.526018943561339,
"grad_norm": 1.7170387506484985,
"learning_rate": 3.699588647169728e-05,
"loss": 2.5020654296875,
"step": 399500
},
{
"epoch": 2.529180419085195,
"grad_norm": 1.7514601945877075,
"learning_rate": 3.691653600289471e-05,
"loss": 2.50452490234375,
"step": 400000
},
{
"epoch": 2.532341894609052,
"grad_norm": 1.8122878074645996,
"learning_rate": 3.683718553409214e-05,
"loss": 2.501074951171875,
"step": 400500
},
{
"epoch": 2.5355033701329086,
"grad_norm": 1.714368224143982,
"learning_rate": 3.675783506528957e-05,
"loss": 2.50412353515625,
"step": 401000
},
{
"epoch": 2.538664845656765,
"grad_norm": 1.804879903793335,
"learning_rate": 3.6678484596487e-05,
"loss": 2.503176025390625,
"step": 401500
},
{
"epoch": 2.5418263211806216,
"grad_norm": 1.7031537294387817,
"learning_rate": 3.659913412768442e-05,
"loss": 2.50793310546875,
"step": 402000
},
{
"epoch": 2.544987796704478,
"grad_norm": 1.716870903968811,
"learning_rate": 3.651978365888185e-05,
"loss": 2.504349365234375,
"step": 402500
},
{
"epoch": 2.5481492722283345,
"grad_norm": 1.8532987833023071,
"learning_rate": 3.644043319007929e-05,
"loss": 2.50275390625,
"step": 403000
},
{
"epoch": 2.551310747752191,
"grad_norm": 1.7822057008743286,
"learning_rate": 3.636108272127672e-05,
"loss": 2.504682861328125,
"step": 403500
},
{
"epoch": 2.5544722232760475,
"grad_norm": 1.7340317964553833,
"learning_rate": 3.628173225247415e-05,
"loss": 2.492722412109375,
"step": 404000
},
{
"epoch": 2.5576336987999038,
"grad_norm": 1.8002653121948242,
"learning_rate": 3.620238178367158e-05,
"loss": 2.50096435546875,
"step": 404500
},
{
"epoch": 2.5607951743237605,
"grad_norm": 1.726341724395752,
"learning_rate": 3.612303131486901e-05,
"loss": 2.50090625,
"step": 405000
},
{
"epoch": 2.5639566498476167,
"grad_norm": 1.8281654119491577,
"learning_rate": 3.604368084606644e-05,
"loss": 2.507187255859375,
"step": 405500
},
{
"epoch": 2.5671181253714734,
"grad_norm": 1.7688469886779785,
"learning_rate": 3.596433037726387e-05,
"loss": 2.499770751953125,
"step": 406000
},
{
"epoch": 2.57027960089533,
"grad_norm": 1.762968897819519,
"learning_rate": 3.58849799084613e-05,
"loss": 2.502153076171875,
"step": 406500
},
{
"epoch": 2.5734410764191864,
"grad_norm": 1.7306631803512573,
"learning_rate": 3.5805629439658734e-05,
"loss": 2.501606689453125,
"step": 407000
},
{
"epoch": 2.5766025519430427,
"grad_norm": 1.7622700929641724,
"learning_rate": 3.5726278970856164e-05,
"loss": 2.499221923828125,
"step": 407500
},
{
"epoch": 2.5797640274668994,
"grad_norm": 1.6837431192398071,
"learning_rate": 3.564692850205359e-05,
"loss": 2.500308349609375,
"step": 408000
},
{
"epoch": 2.582925502990756,
"grad_norm": 1.802474021911621,
"learning_rate": 3.556757803325102e-05,
"loss": 2.50134814453125,
"step": 408500
},
{
"epoch": 2.5860869785146123,
"grad_norm": 1.7845810651779175,
"learning_rate": 3.548822756444845e-05,
"loss": 2.49598828125,
"step": 409000
},
{
"epoch": 2.5892484540384686,
"grad_norm": 1.710442066192627,
"learning_rate": 3.540887709564588e-05,
"loss": 2.498289306640625,
"step": 409500
},
{
"epoch": 2.5924099295623253,
"grad_norm": 1.7537379264831543,
"learning_rate": 3.532952662684331e-05,
"loss": 2.502031982421875,
"step": 410000
},
{
"epoch": 2.595571405086182,
"grad_norm": 1.864090919494629,
"learning_rate": 3.525017615804074e-05,
"loss": 2.493902099609375,
"step": 410500
},
{
"epoch": 2.5987328806100383,
"grad_norm": 1.7609754800796509,
"learning_rate": 3.517082568923818e-05,
"loss": 2.49700537109375,
"step": 411000
},
{
"epoch": 2.601894356133895,
"grad_norm": 1.9026283025741577,
"learning_rate": 3.509147522043561e-05,
"loss": 2.4989794921875,
"step": 411500
},
{
"epoch": 2.6050558316577512,
"grad_norm": 1.7925771474838257,
"learning_rate": 3.501212475163304e-05,
"loss": 2.502332275390625,
"step": 412000
},
{
"epoch": 2.608217307181608,
"grad_norm": 1.741355061531067,
"learning_rate": 3.493277428283047e-05,
"loss": 2.49816748046875,
"step": 412500
},
{
"epoch": 2.611378782705464,
"grad_norm": 1.7714658975601196,
"learning_rate": 3.48534238140279e-05,
"loss": 2.497150634765625,
"step": 413000
},
{
"epoch": 2.614540258229321,
"grad_norm": 1.8279527425765991,
"learning_rate": 3.477407334522533e-05,
"loss": 2.5073349609375,
"step": 413500
},
{
"epoch": 2.617701733753177,
"grad_norm": 1.7267725467681885,
"learning_rate": 3.469472287642276e-05,
"loss": 2.49571728515625,
"step": 414000
},
{
"epoch": 2.620863209277034,
"grad_norm": 1.801113247871399,
"learning_rate": 3.461537240762018e-05,
"loss": 2.495616943359375,
"step": 414500
},
{
"epoch": 2.62402468480089,
"grad_norm": 1.752238154411316,
"learning_rate": 3.4536021938817614e-05,
"loss": 2.495669921875,
"step": 415000
},
{
"epoch": 2.627186160324747,
"grad_norm": 1.7667412757873535,
"learning_rate": 3.4456671470015044e-05,
"loss": 2.49341064453125,
"step": 415500
},
{
"epoch": 2.630347635848603,
"grad_norm": 1.7637271881103516,
"learning_rate": 3.4377321001212475e-05,
"loss": 2.496296142578125,
"step": 416000
},
{
"epoch": 2.63350911137246,
"grad_norm": 1.769901156425476,
"learning_rate": 3.4297970532409905e-05,
"loss": 2.498442138671875,
"step": 416500
},
{
"epoch": 2.636670586896316,
"grad_norm": 1.6878020763397217,
"learning_rate": 3.4218620063607335e-05,
"loss": 2.503362548828125,
"step": 417000
},
{
"epoch": 2.6398320624201728,
"grad_norm": 1.6940027475357056,
"learning_rate": 3.4139269594804766e-05,
"loss": 2.49743603515625,
"step": 417500
},
{
"epoch": 2.6429935379440295,
"grad_norm": 1.7306866645812988,
"learning_rate": 3.4059919126002196e-05,
"loss": 2.494462646484375,
"step": 418000
},
{
"epoch": 2.6461550134678857,
"grad_norm": 1.8407728672027588,
"learning_rate": 3.398056865719963e-05,
"loss": 2.4992060546875,
"step": 418500
},
{
"epoch": 2.649316488991742,
"grad_norm": 1.725067377090454,
"learning_rate": 3.390121818839706e-05,
"loss": 2.493571533203125,
"step": 419000
},
{
"epoch": 2.6524779645155987,
"grad_norm": 1.8143824338912964,
"learning_rate": 3.3821867719594494e-05,
"loss": 2.493434326171875,
"step": 419500
},
{
"epoch": 2.6556394400394554,
"grad_norm": 1.836155652999878,
"learning_rate": 3.3742517250791925e-05,
"loss": 2.49705419921875,
"step": 420000
},
{
"epoch": 2.6588009155633117,
"grad_norm": 1.6706143617630005,
"learning_rate": 3.366316678198935e-05,
"loss": 2.491701416015625,
"step": 420500
},
{
"epoch": 2.6619623910871684,
"grad_norm": 1.7750943899154663,
"learning_rate": 3.358381631318678e-05,
"loss": 2.496917236328125,
"step": 421000
},
{
"epoch": 2.6651238666110246,
"grad_norm": 1.7143902778625488,
"learning_rate": 3.350446584438421e-05,
"loss": 2.49254150390625,
"step": 421500
},
{
"epoch": 2.6682853421348813,
"grad_norm": 1.755142092704773,
"learning_rate": 3.342511537558164e-05,
"loss": 2.496576416015625,
"step": 422000
},
{
"epoch": 2.6714468176587376,
"grad_norm": 1.7843048572540283,
"learning_rate": 3.334576490677907e-05,
"loss": 2.492592529296875,
"step": 422500
},
{
"epoch": 2.6746082931825943,
"grad_norm": 1.7168258428573608,
"learning_rate": 3.32664144379765e-05,
"loss": 2.492325927734375,
"step": 423000
},
{
"epoch": 2.6777697687064506,
"grad_norm": 1.8280614614486694,
"learning_rate": 3.318706396917393e-05,
"loss": 2.4921962890625,
"step": 423500
},
{
"epoch": 2.6809312442303073,
"grad_norm": 1.7062269449234009,
"learning_rate": 3.310771350037136e-05,
"loss": 2.491731201171875,
"step": 424000
},
{
"epoch": 2.6840927197541635,
"grad_norm": 1.786206841468811,
"learning_rate": 3.302836303156879e-05,
"loss": 2.4906005859375,
"step": 424500
},
{
"epoch": 2.6872541952780202,
"grad_norm": 1.7371267080307007,
"learning_rate": 3.294901256276622e-05,
"loss": 2.495618896484375,
"step": 425000
},
{
"epoch": 2.6904156708018765,
"grad_norm": 1.736602783203125,
"learning_rate": 3.286966209396365e-05,
"loss": 2.487023681640625,
"step": 425500
},
{
"epoch": 2.693577146325733,
"grad_norm": 1.806997537612915,
"learning_rate": 3.279031162516108e-05,
"loss": 2.494110107421875,
"step": 426000
},
{
"epoch": 2.6967386218495895,
"grad_norm": 1.8468589782714844,
"learning_rate": 3.2710961156358513e-05,
"loss": 2.494903076171875,
"step": 426500
},
{
"epoch": 2.699900097373446,
"grad_norm": 1.822914719581604,
"learning_rate": 3.2631610687555944e-05,
"loss": 2.4949716796875,
"step": 427000
},
{
"epoch": 2.703061572897303,
"grad_norm": 1.7953490018844604,
"learning_rate": 3.2552260218753374e-05,
"loss": 2.494130615234375,
"step": 427500
},
{
"epoch": 2.706223048421159,
"grad_norm": 1.791394829750061,
"learning_rate": 3.2472909749950805e-05,
"loss": 2.49253466796875,
"step": 428000
},
{
"epoch": 2.7093845239450154,
"grad_norm": 1.7596778869628906,
"learning_rate": 3.2393559281148235e-05,
"loss": 2.482137451171875,
"step": 428500
},
{
"epoch": 2.712545999468872,
"grad_norm": 1.8860998153686523,
"learning_rate": 3.2314208812345666e-05,
"loss": 2.489312255859375,
"step": 429000
},
{
"epoch": 2.715707474992729,
"grad_norm": 1.7928045988082886,
"learning_rate": 3.2234858343543096e-05,
"loss": 2.488930419921875,
"step": 429500
},
{
"epoch": 2.718868950516585,
"grad_norm": 1.7422815561294556,
"learning_rate": 3.2155507874740526e-05,
"loss": 2.491107177734375,
"step": 430000
},
{
"epoch": 2.7220304260404413,
"grad_norm": 1.796434998512268,
"learning_rate": 3.207615740593796e-05,
"loss": 2.49074853515625,
"step": 430500
},
{
"epoch": 2.725191901564298,
"grad_norm": 1.6872786283493042,
"learning_rate": 3.199680693713539e-05,
"loss": 2.493439453125,
"step": 431000
},
{
"epoch": 2.7283533770881547,
"grad_norm": 1.803020715713501,
"learning_rate": 3.191745646833282e-05,
"loss": 2.490026123046875,
"step": 431500
},
{
"epoch": 2.731514852612011,
"grad_norm": 1.7220793962478638,
"learning_rate": 3.183810599953025e-05,
"loss": 2.485221923828125,
"step": 432000
},
{
"epoch": 2.7346763281358677,
"grad_norm": 1.8259172439575195,
"learning_rate": 3.175875553072768e-05,
"loss": 2.48187060546875,
"step": 432500
},
{
"epoch": 2.737837803659724,
"grad_norm": 1.7771124839782715,
"learning_rate": 3.16794050619251e-05,
"loss": 2.49085986328125,
"step": 433000
},
{
"epoch": 2.7409992791835807,
"grad_norm": 1.7622228860855103,
"learning_rate": 3.160005459312253e-05,
"loss": 2.486861572265625,
"step": 433500
},
{
"epoch": 2.744160754707437,
"grad_norm": 1.7843176126480103,
"learning_rate": 3.152070412431996e-05,
"loss": 2.491146240234375,
"step": 434000
},
{
"epoch": 2.7473222302312936,
"grad_norm": 1.8698512315750122,
"learning_rate": 3.14413536555174e-05,
"loss": 2.48747216796875,
"step": 434500
},
{
"epoch": 2.75048370575515,
"grad_norm": 1.7659764289855957,
"learning_rate": 3.136200318671483e-05,
"loss": 2.48459423828125,
"step": 435000
},
{
"epoch": 2.7536451812790066,
"grad_norm": 1.8431531190872192,
"learning_rate": 3.128265271791226e-05,
"loss": 2.487949462890625,
"step": 435500
},
{
"epoch": 2.756806656802863,
"grad_norm": 1.777239441871643,
"learning_rate": 3.120330224910969e-05,
"loss": 2.484468994140625,
"step": 436000
},
{
"epoch": 2.7599681323267196,
"grad_norm": 1.8007045984268188,
"learning_rate": 3.112395178030712e-05,
"loss": 2.490224365234375,
"step": 436500
},
{
"epoch": 2.7631296078505763,
"grad_norm": 1.7780418395996094,
"learning_rate": 3.104460131150455e-05,
"loss": 2.4871572265625,
"step": 437000
},
{
"epoch": 2.7662910833744325,
"grad_norm": 1.800548791885376,
"learning_rate": 3.096525084270198e-05,
"loss": 2.486268798828125,
"step": 437500
},
{
"epoch": 2.769452558898289,
"grad_norm": 1.7479411363601685,
"learning_rate": 3.088590037389941e-05,
"loss": 2.490397705078125,
"step": 438000
},
{
"epoch": 2.7726140344221455,
"grad_norm": 1.803053855895996,
"learning_rate": 3.0806549905096844e-05,
"loss": 2.49248876953125,
"step": 438500
},
{
"epoch": 2.775775509946002,
"grad_norm": 1.8069261312484741,
"learning_rate": 3.072719943629427e-05,
"loss": 2.4850234375,
"step": 439000
},
{
"epoch": 2.7789369854698585,
"grad_norm": 1.6864376068115234,
"learning_rate": 3.06478489674917e-05,
"loss": 2.4844697265625,
"step": 439500
},
{
"epoch": 2.7820984609937147,
"grad_norm": 1.741611123085022,
"learning_rate": 3.056849849868913e-05,
"loss": 2.480401611328125,
"step": 440000
},
{
"epoch": 2.7852599365175714,
"grad_norm": 1.8807125091552734,
"learning_rate": 3.0489148029886562e-05,
"loss": 2.487471923828125,
"step": 440500
},
{
"epoch": 2.788421412041428,
"grad_norm": 1.7675764560699463,
"learning_rate": 3.0409797561083992e-05,
"loss": 2.489609619140625,
"step": 441000
},
{
"epoch": 2.7915828875652844,
"grad_norm": 1.7800296545028687,
"learning_rate": 3.0330447092281423e-05,
"loss": 2.48851904296875,
"step": 441500
},
{
"epoch": 2.794744363089141,
"grad_norm": 1.7281991243362427,
"learning_rate": 3.0251096623478853e-05,
"loss": 2.478299072265625,
"step": 442000
},
{
"epoch": 2.7979058386129974,
"grad_norm": 1.7676992416381836,
"learning_rate": 3.0171746154676284e-05,
"loss": 2.486189453125,
"step": 442500
},
{
"epoch": 2.801067314136854,
"grad_norm": 1.7467396259307861,
"learning_rate": 3.0092395685873714e-05,
"loss": 2.488348876953125,
"step": 443000
},
{
"epoch": 2.8042287896607103,
"grad_norm": 1.7488715648651123,
"learning_rate": 3.0013045217071144e-05,
"loss": 2.481170166015625,
"step": 443500
},
{
"epoch": 2.807390265184567,
"grad_norm": 1.8407244682312012,
"learning_rate": 2.9933694748268575e-05,
"loss": 2.482872802734375,
"step": 444000
},
{
"epoch": 2.8105517407084233,
"grad_norm": 1.7635235786437988,
"learning_rate": 2.9854344279466005e-05,
"loss": 2.48202587890625,
"step": 444500
},
{
"epoch": 2.81371321623228,
"grad_norm": 1.732746958732605,
"learning_rate": 2.9774993810663436e-05,
"loss": 2.4831953125,
"step": 445000
},
{
"epoch": 2.8168746917561363,
"grad_norm": 1.7351652383804321,
"learning_rate": 2.9695643341860863e-05,
"loss": 2.48315771484375,
"step": 445500
},
{
"epoch": 2.820036167279993,
"grad_norm": 1.795088768005371,
"learning_rate": 2.9616292873058293e-05,
"loss": 2.475879150390625,
"step": 446000
},
{
"epoch": 2.8231976428038497,
"grad_norm": 1.842271089553833,
"learning_rate": 2.9536942404255724e-05,
"loss": 2.484837890625,
"step": 446500
},
{
"epoch": 2.826359118327706,
"grad_norm": 1.7460764646530151,
"learning_rate": 2.9457591935453154e-05,
"loss": 2.482647705078125,
"step": 447000
},
{
"epoch": 2.829520593851562,
"grad_norm": 1.8485976457595825,
"learning_rate": 2.9378241466650584e-05,
"loss": 2.486168701171875,
"step": 447500
},
{
"epoch": 2.832682069375419,
"grad_norm": 1.741240382194519,
"learning_rate": 2.9298890997848015e-05,
"loss": 2.47640673828125,
"step": 448000
},
{
"epoch": 2.8358435448992756,
"grad_norm": 1.7596064805984497,
"learning_rate": 2.921954052904545e-05,
"loss": 2.488303466796875,
"step": 448500
},
{
"epoch": 2.839005020423132,
"grad_norm": 1.7422022819519043,
"learning_rate": 2.914019006024288e-05,
"loss": 2.482251220703125,
"step": 449000
},
{
"epoch": 2.842166495946988,
"grad_norm": 1.860239028930664,
"learning_rate": 2.906083959144031e-05,
"loss": 2.47701904296875,
"step": 449500
},
{
"epoch": 2.845327971470845,
"grad_norm": 1.7305396795272827,
"learning_rate": 2.898148912263774e-05,
"loss": 2.47966357421875,
"step": 450000
},
{
"epoch": 2.8484894469947015,
"grad_norm": 1.811401128768921,
"learning_rate": 2.890213865383517e-05,
"loss": 2.47669970703125,
"step": 450500
},
{
"epoch": 2.851650922518558,
"grad_norm": 1.8047016859054565,
"learning_rate": 2.88227881850326e-05,
"loss": 2.48571875,
"step": 451000
},
{
"epoch": 2.8548123980424145,
"grad_norm": 1.8378208875656128,
"learning_rate": 2.8743437716230024e-05,
"loss": 2.4792705078125,
"step": 451500
},
{
"epoch": 2.8579738735662708,
"grad_norm": 1.8144879341125488,
"learning_rate": 2.8664087247427458e-05,
"loss": 2.479243896484375,
"step": 452000
},
{
"epoch": 2.8611353490901275,
"grad_norm": 1.7534644603729248,
"learning_rate": 2.858473677862489e-05,
"loss": 2.476955078125,
"step": 452500
},
{
"epoch": 2.8642968246139837,
"grad_norm": 1.7392510175704956,
"learning_rate": 2.850538630982232e-05,
"loss": 2.48573095703125,
"step": 453000
},
{
"epoch": 2.8674583001378404,
"grad_norm": 1.7234078645706177,
"learning_rate": 2.842603584101975e-05,
"loss": 2.480677978515625,
"step": 453500
},
{
"epoch": 2.8706197756616967,
"grad_norm": 1.7838243246078491,
"learning_rate": 2.834668537221718e-05,
"loss": 2.47841943359375,
"step": 454000
},
{
"epoch": 2.8737812511855534,
"grad_norm": 1.8608052730560303,
"learning_rate": 2.826733490341461e-05,
"loss": 2.47799267578125,
"step": 454500
},
{
"epoch": 2.8769427267094096,
"grad_norm": 1.7855165004730225,
"learning_rate": 2.818798443461204e-05,
"loss": 2.47879736328125,
"step": 455000
},
{
"epoch": 2.8801042022332664,
"grad_norm": 1.8582948446273804,
"learning_rate": 2.810863396580947e-05,
"loss": 2.480887939453125,
"step": 455500
},
{
"epoch": 2.883265677757123,
"grad_norm": 1.854804277420044,
"learning_rate": 2.80292834970069e-05,
"loss": 2.4784150390625,
"step": 456000
},
{
"epoch": 2.8864271532809793,
"grad_norm": 1.7963296175003052,
"learning_rate": 2.7949933028204332e-05,
"loss": 2.4748212890625,
"step": 456500
},
{
"epoch": 2.8895886288048356,
"grad_norm": 1.7743266820907593,
"learning_rate": 2.7870582559401766e-05,
"loss": 2.475849609375,
"step": 457000
},
{
"epoch": 2.8927501043286923,
"grad_norm": 1.680433750152588,
"learning_rate": 2.7791232090599196e-05,
"loss": 2.481010009765625,
"step": 457500
},
{
"epoch": 2.895911579852549,
"grad_norm": 1.8547523021697998,
"learning_rate": 2.771188162179662e-05,
"loss": 2.47914453125,
"step": 458000
},
{
"epoch": 2.8990730553764053,
"grad_norm": 1.7567074298858643,
"learning_rate": 2.763253115299405e-05,
"loss": 2.483482177734375,
"step": 458500
},
{
"epoch": 2.9022345309002615,
"grad_norm": 1.906812310218811,
"learning_rate": 2.755318068419148e-05,
"loss": 2.48186962890625,
"step": 459000
},
{
"epoch": 2.905396006424118,
"grad_norm": 1.8662272691726685,
"learning_rate": 2.747383021538891e-05,
"loss": 2.480791748046875,
"step": 459500
},
{
"epoch": 2.908557481947975,
"grad_norm": 1.6749777793884277,
"learning_rate": 2.7394479746586345e-05,
"loss": 2.4720859375,
"step": 460000
},
{
"epoch": 2.911718957471831,
"grad_norm": 1.7413032054901123,
"learning_rate": 2.7315129277783775e-05,
"loss": 2.473732421875,
"step": 460500
},
{
"epoch": 2.914880432995688,
"grad_norm": 1.7523657083511353,
"learning_rate": 2.7235778808981206e-05,
"loss": 2.476798583984375,
"step": 461000
},
{
"epoch": 2.918041908519544,
"grad_norm": 1.7592319250106812,
"learning_rate": 2.7156428340178636e-05,
"loss": 2.47803955078125,
"step": 461500
},
{
"epoch": 2.921203384043401,
"grad_norm": 1.8783061504364014,
"learning_rate": 2.7077077871376067e-05,
"loss": 2.47321630859375,
"step": 462000
},
{
"epoch": 2.924364859567257,
"grad_norm": 1.8487108945846558,
"learning_rate": 2.6997727402573497e-05,
"loss": 2.4805830078125,
"step": 462500
},
{
"epoch": 2.927526335091114,
"grad_norm": 1.7998102903366089,
"learning_rate": 2.6918376933770928e-05,
"loss": 2.479827392578125,
"step": 463000
},
{
"epoch": 2.93068781061497,
"grad_norm": 1.8088874816894531,
"learning_rate": 2.6839026464968358e-05,
"loss": 2.470238037109375,
"step": 463500
},
{
"epoch": 2.933849286138827,
"grad_norm": 1.7920806407928467,
"learning_rate": 2.6759675996165785e-05,
"loss": 2.477951171875,
"step": 464000
},
{
"epoch": 2.937010761662683,
"grad_norm": 1.7425010204315186,
"learning_rate": 2.6680325527363215e-05,
"loss": 2.4783095703125,
"step": 464500
},
{
"epoch": 2.9401722371865397,
"grad_norm": 1.8238309621810913,
"learning_rate": 2.6600975058560646e-05,
"loss": 2.47339697265625,
"step": 465000
},
{
"epoch": 2.9433337127103965,
"grad_norm": 1.7441800832748413,
"learning_rate": 2.6521624589758076e-05,
"loss": 2.47821630859375,
"step": 465500
},
{
"epoch": 2.9464951882342527,
"grad_norm": 1.793269395828247,
"learning_rate": 2.6442274120955507e-05,
"loss": 2.473937744140625,
"step": 466000
},
{
"epoch": 2.949656663758109,
"grad_norm": 1.7536602020263672,
"learning_rate": 2.6362923652152937e-05,
"loss": 2.47687158203125,
"step": 466500
},
{
"epoch": 2.9528181392819657,
"grad_norm": 1.826667070388794,
"learning_rate": 2.6283573183350368e-05,
"loss": 2.471275390625,
"step": 467000
},
{
"epoch": 2.9559796148058224,
"grad_norm": 1.7831002473831177,
"learning_rate": 2.6204222714547798e-05,
"loss": 2.475607421875,
"step": 467500
},
{
"epoch": 2.9591410903296786,
"grad_norm": 1.92928946018219,
"learning_rate": 2.612487224574523e-05,
"loss": 2.47490771484375,
"step": 468000
},
{
"epoch": 2.962302565853535,
"grad_norm": 1.7333531379699707,
"learning_rate": 2.6045521776942662e-05,
"loss": 2.470122314453125,
"step": 468500
},
{
"epoch": 2.9654640413773916,
"grad_norm": 1.7594388723373413,
"learning_rate": 2.5966171308140093e-05,
"loss": 2.477195068359375,
"step": 469000
},
{
"epoch": 2.9686255169012483,
"grad_norm": 1.7311286926269531,
"learning_rate": 2.5886820839337523e-05,
"loss": 2.474933349609375,
"step": 469500
},
{
"epoch": 2.9717869924251046,
"grad_norm": 1.825905442237854,
"learning_rate": 2.5807470370534954e-05,
"loss": 2.4705478515625,
"step": 470000
},
{
"epoch": 2.9749484679489613,
"grad_norm": 1.7995409965515137,
"learning_rate": 2.5728119901732377e-05,
"loss": 2.470817626953125,
"step": 470500
},
{
"epoch": 2.9781099434728175,
"grad_norm": 1.8015352487564087,
"learning_rate": 2.5648769432929808e-05,
"loss": 2.466350341796875,
"step": 471000
},
{
"epoch": 2.9812714189966742,
"grad_norm": 1.7570463418960571,
"learning_rate": 2.5569418964127238e-05,
"loss": 2.4701015625,
"step": 471500
},
{
"epoch": 2.9844328945205305,
"grad_norm": 1.7860133647918701,
"learning_rate": 2.5490068495324672e-05,
"loss": 2.474818359375,
"step": 472000
},
{
"epoch": 2.987594370044387,
"grad_norm": 1.7863579988479614,
"learning_rate": 2.5410718026522102e-05,
"loss": 2.477245849609375,
"step": 472500
},
{
"epoch": 2.9907558455682435,
"grad_norm": 1.755395531654358,
"learning_rate": 2.5331367557719533e-05,
"loss": 2.463203857421875,
"step": 473000
},
{
"epoch": 2.9939173210921,
"grad_norm": 1.7714825868606567,
"learning_rate": 2.5252017088916963e-05,
"loss": 2.474131591796875,
"step": 473500
},
{
"epoch": 2.9970787966159564,
"grad_norm": 1.7622647285461426,
"learning_rate": 2.5172666620114394e-05,
"loss": 2.4758662109375,
"step": 474000
},
{
"epoch": 3.000240272139813,
"grad_norm": 1.8382084369659424,
"learning_rate": 2.5093316151311824e-05,
"loss": 2.472795166015625,
"step": 474500
},
{
"epoch": 3.0034017476636694,
"grad_norm": 1.7811071872711182,
"learning_rate": 2.5013965682509254e-05,
"loss": 2.472397705078125,
"step": 475000
},
{
"epoch": 3.006563223187526,
"grad_norm": 1.766930103302002,
"learning_rate": 2.493461521370668e-05,
"loss": 2.4680751953125,
"step": 475500
},
{
"epoch": 3.0097246987113824,
"grad_norm": 1.7869229316711426,
"learning_rate": 2.4855264744904115e-05,
"loss": 2.467306396484375,
"step": 476000
},
{
"epoch": 3.012886174235239,
"grad_norm": 1.842387080192566,
"learning_rate": 2.4775914276101546e-05,
"loss": 2.465614501953125,
"step": 476500
},
{
"epoch": 3.016047649759096,
"grad_norm": 1.873660922050476,
"learning_rate": 2.4696563807298976e-05,
"loss": 2.46312646484375,
"step": 477000
},
{
"epoch": 3.019209125282952,
"grad_norm": 1.8362675905227661,
"learning_rate": 2.4617213338496406e-05,
"loss": 2.471597412109375,
"step": 477500
},
{
"epoch": 3.0223706008068087,
"grad_norm": 1.9381550550460815,
"learning_rate": 2.4537862869693834e-05,
"loss": 2.469968017578125,
"step": 478000
},
{
"epoch": 3.025532076330665,
"grad_norm": 1.8130383491516113,
"learning_rate": 2.4458512400891264e-05,
"loss": 2.46311328125,
"step": 478500
},
{
"epoch": 3.0286935518545217,
"grad_norm": 1.9205710887908936,
"learning_rate": 2.4379161932088694e-05,
"loss": 2.467774658203125,
"step": 479000
},
{
"epoch": 3.031855027378378,
"grad_norm": 1.795658826828003,
"learning_rate": 2.4299811463286125e-05,
"loss": 2.466140380859375,
"step": 479500
},
{
"epoch": 3.0350165029022347,
"grad_norm": 1.8448054790496826,
"learning_rate": 2.422046099448356e-05,
"loss": 2.4651552734375,
"step": 480000
},
{
"epoch": 3.038177978426091,
"grad_norm": 1.7483762502670288,
"learning_rate": 2.414111052568099e-05,
"loss": 2.4660029296875,
"step": 480500
},
{
"epoch": 3.0413394539499476,
"grad_norm": 1.8066811561584473,
"learning_rate": 2.406176005687842e-05,
"loss": 2.466659912109375,
"step": 481000
},
{
"epoch": 3.044500929473804,
"grad_norm": 1.9203633069992065,
"learning_rate": 2.3982409588075846e-05,
"loss": 2.471052734375,
"step": 481500
},
{
"epoch": 3.0476624049976606,
"grad_norm": 1.8127508163452148,
"learning_rate": 2.3903059119273277e-05,
"loss": 2.469372802734375,
"step": 482000
},
{
"epoch": 3.050823880521517,
"grad_norm": 1.8026906251907349,
"learning_rate": 2.3823708650470707e-05,
"loss": 2.469861328125,
"step": 482500
},
{
"epoch": 3.0539853560453736,
"grad_norm": 1.7605341672897339,
"learning_rate": 2.3744358181668138e-05,
"loss": 2.4647705078125,
"step": 483000
},
{
"epoch": 3.05714683156923,
"grad_norm": 1.7471855878829956,
"learning_rate": 2.3665007712865568e-05,
"loss": 2.46558447265625,
"step": 483500
},
{
"epoch": 3.0603083070930865,
"grad_norm": 1.8125495910644531,
"learning_rate": 2.3585657244063e-05,
"loss": 2.4705,
"step": 484000
},
{
"epoch": 3.063469782616943,
"grad_norm": 1.7905769348144531,
"learning_rate": 2.350630677526043e-05,
"loss": 2.463659423828125,
"step": 484500
},
{
"epoch": 3.0666312581407995,
"grad_norm": 1.7178443670272827,
"learning_rate": 2.342695630645786e-05,
"loss": 2.467238525390625,
"step": 485000
},
{
"epoch": 3.0697927336646558,
"grad_norm": 1.829702615737915,
"learning_rate": 2.334760583765529e-05,
"loss": 2.465052001953125,
"step": 485500
},
{
"epoch": 3.0729542091885125,
"grad_norm": 1.8007923364639282,
"learning_rate": 2.326825536885272e-05,
"loss": 2.4671669921875,
"step": 486000
},
{
"epoch": 3.076115684712369,
"grad_norm": 2.0038607120513916,
"learning_rate": 2.318890490005015e-05,
"loss": 2.467210693359375,
"step": 486500
},
{
"epoch": 3.0792771602362254,
"grad_norm": 1.9250867366790771,
"learning_rate": 2.310955443124758e-05,
"loss": 2.462173095703125,
"step": 487000
},
{
"epoch": 3.082438635760082,
"grad_norm": 1.776657223701477,
"learning_rate": 2.303020396244501e-05,
"loss": 2.461007080078125,
"step": 487500
},
{
"epoch": 3.0856001112839384,
"grad_norm": 1.8971377611160278,
"learning_rate": 2.2950853493642442e-05,
"loss": 2.46502392578125,
"step": 488000
},
{
"epoch": 3.088761586807795,
"grad_norm": 1.8631473779678345,
"learning_rate": 2.2871503024839872e-05,
"loss": 2.46151708984375,
"step": 488500
},
{
"epoch": 3.0919230623316514,
"grad_norm": 1.8068488836288452,
"learning_rate": 2.2792152556037303e-05,
"loss": 2.46356640625,
"step": 489000
},
{
"epoch": 3.095084537855508,
"grad_norm": 1.7869725227355957,
"learning_rate": 2.2712802087234733e-05,
"loss": 2.46430712890625,
"step": 489500
},
{
"epoch": 3.0982460133793643,
"grad_norm": 1.9093724489212036,
"learning_rate": 2.2633451618432164e-05,
"loss": 2.46559716796875,
"step": 490000
},
{
"epoch": 3.101407488903221,
"grad_norm": 1.8199621438980103,
"learning_rate": 2.255410114962959e-05,
"loss": 2.463968994140625,
"step": 490500
},
{
"epoch": 3.1045689644270773,
"grad_norm": 1.8101762533187866,
"learning_rate": 2.247475068082702e-05,
"loss": 2.46714013671875,
"step": 491000
},
{
"epoch": 3.107730439950934,
"grad_norm": 1.753417730331421,
"learning_rate": 2.2395400212024455e-05,
"loss": 2.469317626953125,
"step": 491500
},
{
"epoch": 3.1108919154747903,
"grad_norm": 1.917372465133667,
"learning_rate": 2.2316049743221885e-05,
"loss": 2.46452490234375,
"step": 492000
},
{
"epoch": 3.114053390998647,
"grad_norm": 1.8102434873580933,
"learning_rate": 2.2236699274419316e-05,
"loss": 2.461546142578125,
"step": 492500
},
{
"epoch": 3.1172148665225032,
"grad_norm": 1.9709382057189941,
"learning_rate": 2.2157348805616746e-05,
"loss": 2.46670068359375,
"step": 493000
},
{
"epoch": 3.12037634204636,
"grad_norm": 1.7828607559204102,
"learning_rate": 2.2077998336814177e-05,
"loss": 2.462189208984375,
"step": 493500
},
{
"epoch": 3.123537817570216,
"grad_norm": 1.7755306959152222,
"learning_rate": 2.1998647868011604e-05,
"loss": 2.46380029296875,
"step": 494000
},
{
"epoch": 3.126699293094073,
"grad_norm": 1.8087832927703857,
"learning_rate": 2.1919297399209034e-05,
"loss": 2.462564697265625,
"step": 494500
},
{
"epoch": 3.129860768617929,
"grad_norm": 1.831347942352295,
"learning_rate": 2.1839946930406465e-05,
"loss": 2.461643798828125,
"step": 495000
},
{
"epoch": 3.133022244141786,
"grad_norm": 1.8205162286758423,
"learning_rate": 2.1760596461603895e-05,
"loss": 2.46226171875,
"step": 495500
},
{
"epoch": 3.1361837196656426,
"grad_norm": 1.8381644487380981,
"learning_rate": 2.168124599280133e-05,
"loss": 2.46469873046875,
"step": 496000
},
{
"epoch": 3.139345195189499,
"grad_norm": 1.8561416864395142,
"learning_rate": 2.160189552399876e-05,
"loss": 2.46030322265625,
"step": 496500
},
{
"epoch": 3.1425066707133555,
"grad_norm": 1.8560101985931396,
"learning_rate": 2.1522545055196186e-05,
"loss": 2.46278662109375,
"step": 497000
},
{
"epoch": 3.145668146237212,
"grad_norm": 1.90720534324646,
"learning_rate": 2.1443194586393617e-05,
"loss": 2.464942626953125,
"step": 497500
},
{
"epoch": 3.1488296217610685,
"grad_norm": 1.744273066520691,
"learning_rate": 2.1363844117591047e-05,
"loss": 2.45955029296875,
"step": 498000
},
{
"epoch": 3.1519910972849248,
"grad_norm": 1.7981034517288208,
"learning_rate": 2.1284493648788477e-05,
"loss": 2.463569091796875,
"step": 498500
},
{
"epoch": 3.1551525728087815,
"grad_norm": 1.9599028825759888,
"learning_rate": 2.1205143179985908e-05,
"loss": 2.455825927734375,
"step": 499000
},
{
"epoch": 3.1583140483326377,
"grad_norm": 1.7581430673599243,
"learning_rate": 2.112579271118334e-05,
"loss": 2.462685546875,
"step": 499500
},
{
"epoch": 3.1614755238564944,
"grad_norm": 1.7590006589889526,
"learning_rate": 2.104644224238077e-05,
"loss": 2.4585322265625,
"step": 500000
},
{
"epoch": 3.1646369993803507,
"grad_norm": 1.7604933977127075,
"learning_rate": 2.09670917735782e-05,
"loss": 2.460159423828125,
"step": 500500
},
{
"epoch": 3.1677984749042074,
"grad_norm": 1.797680139541626,
"learning_rate": 2.088774130477563e-05,
"loss": 2.45772900390625,
"step": 501000
},
{
"epoch": 3.1709599504280637,
"grad_norm": 1.818292498588562,
"learning_rate": 2.080839083597306e-05,
"loss": 2.4672841796875,
"step": 501500
},
{
"epoch": 3.1741214259519204,
"grad_norm": 1.8139883279800415,
"learning_rate": 2.072904036717049e-05,
"loss": 2.4628232421875,
"step": 502000
},
{
"epoch": 3.1772829014757766,
"grad_norm": 1.7951339483261108,
"learning_rate": 2.064968989836792e-05,
"loss": 2.4601259765625,
"step": 502500
},
{
"epoch": 3.1804443769996333,
"grad_norm": 1.737992286682129,
"learning_rate": 2.0570339429565348e-05,
"loss": 2.457323486328125,
"step": 503000
},
{
"epoch": 3.1836058525234896,
"grad_norm": 1.9223175048828125,
"learning_rate": 2.0490988960762782e-05,
"loss": 2.460111572265625,
"step": 503500
},
{
"epoch": 3.1867673280473463,
"grad_norm": 1.7282299995422363,
"learning_rate": 2.0411638491960212e-05,
"loss": 2.456501953125,
"step": 504000
},
{
"epoch": 3.1899288035712026,
"grad_norm": 1.7651430368423462,
"learning_rate": 2.0332288023157643e-05,
"loss": 2.456777587890625,
"step": 504500
},
{
"epoch": 3.1930902790950593,
"grad_norm": 1.7545665502548218,
"learning_rate": 2.0252937554355073e-05,
"loss": 2.4595517578125,
"step": 505000
},
{
"epoch": 3.196251754618916,
"grad_norm": 1.8404427766799927,
"learning_rate": 2.0173587085552503e-05,
"loss": 2.459908203125,
"step": 505500
},
{
"epoch": 3.1994132301427722,
"grad_norm": 1.9075236320495605,
"learning_rate": 2.0094236616749934e-05,
"loss": 2.46445654296875,
"step": 506000
},
{
"epoch": 3.2025747056666285,
"grad_norm": 1.8555846214294434,
"learning_rate": 2.001488614794736e-05,
"loss": 2.455956787109375,
"step": 506500
},
{
"epoch": 3.205736181190485,
"grad_norm": 1.863501787185669,
"learning_rate": 1.993553567914479e-05,
"loss": 2.46444140625,
"step": 507000
},
{
"epoch": 3.208897656714342,
"grad_norm": 1.9332226514816284,
"learning_rate": 1.9856185210342225e-05,
"loss": 2.45934619140625,
"step": 507500
},
{
"epoch": 3.212059132238198,
"grad_norm": 1.8140562772750854,
"learning_rate": 1.9776834741539656e-05,
"loss": 2.4611875,
"step": 508000
},
{
"epoch": 3.215220607762055,
"grad_norm": 1.8269511461257935,
"learning_rate": 1.9697484272737086e-05,
"loss": 2.460216796875,
"step": 508500
},
{
"epoch": 3.218382083285911,
"grad_norm": 1.7976815700531006,
"learning_rate": 1.9618133803934516e-05,
"loss": 2.45228466796875,
"step": 509000
},
{
"epoch": 3.221543558809768,
"grad_norm": 1.9040664434432983,
"learning_rate": 1.9538783335131943e-05,
"loss": 2.459642822265625,
"step": 509500
},
{
"epoch": 3.224705034333624,
"grad_norm": 1.8873611688613892,
"learning_rate": 1.9459432866329374e-05,
"loss": 2.467588134765625,
"step": 510000
},
{
"epoch": 3.227866509857481,
"grad_norm": 1.8360077142715454,
"learning_rate": 1.9380082397526804e-05,
"loss": 2.454847412109375,
"step": 510500
},
{
"epoch": 3.231027985381337,
"grad_norm": 1.8908779621124268,
"learning_rate": 1.9300731928724235e-05,
"loss": 2.46505126953125,
"step": 511000
},
{
"epoch": 3.2341894609051938,
"grad_norm": 1.8892560005187988,
"learning_rate": 1.922138145992167e-05,
"loss": 2.45915576171875,
"step": 511500
},
{
"epoch": 3.23735093642905,
"grad_norm": 1.8367403745651245,
"learning_rate": 1.91420309911191e-05,
"loss": 2.46047216796875,
"step": 512000
},
{
"epoch": 3.2405124119529067,
"grad_norm": 1.856690764427185,
"learning_rate": 1.9062680522316526e-05,
"loss": 2.4579599609375,
"step": 512500
},
{
"epoch": 3.243673887476763,
"grad_norm": 1.7953070402145386,
"learning_rate": 1.8983330053513956e-05,
"loss": 2.46224560546875,
"step": 513000
},
{
"epoch": 3.2468353630006197,
"grad_norm": 1.8923710584640503,
"learning_rate": 1.8903979584711387e-05,
"loss": 2.452962890625,
"step": 513500
},
{
"epoch": 3.249996838524476,
"grad_norm": 1.8990341424942017,
"learning_rate": 1.8824629115908817e-05,
"loss": 2.46110400390625,
"step": 514000
},
{
"epoch": 3.2531583140483327,
"grad_norm": 1.8265435695648193,
"learning_rate": 1.8745278647106248e-05,
"loss": 2.45503515625,
"step": 514500
},
{
"epoch": 3.2563197895721894,
"grad_norm": 1.8529694080352783,
"learning_rate": 1.8665928178303678e-05,
"loss": 2.451604248046875,
"step": 515000
},
{
"epoch": 3.2594812650960456,
"grad_norm": 1.832459568977356,
"learning_rate": 1.858657770950111e-05,
"loss": 2.452759033203125,
"step": 515500
},
{
"epoch": 3.262642740619902,
"grad_norm": 1.7773107290267944,
"learning_rate": 1.850722724069854e-05,
"loss": 2.451248046875,
"step": 516000
},
{
"epoch": 3.2658042161437586,
"grad_norm": 1.7747048139572144,
"learning_rate": 1.842787677189597e-05,
"loss": 2.45845458984375,
"step": 516500
},
{
"epoch": 3.2689656916676153,
"grad_norm": 1.8650330305099487,
"learning_rate": 1.83485263030934e-05,
"loss": 2.455768310546875,
"step": 517000
},
{
"epoch": 3.2721271671914716,
"grad_norm": 1.9840600490570068,
"learning_rate": 1.826917583429083e-05,
"loss": 2.45290283203125,
"step": 517500
},
{
"epoch": 3.2752886427153283,
"grad_norm": 1.8055514097213745,
"learning_rate": 1.818982536548826e-05,
"loss": 2.459531005859375,
"step": 518000
},
{
"epoch": 3.2784501182391845,
"grad_norm": 1.92037034034729,
"learning_rate": 1.811047489668569e-05,
"loss": 2.455253662109375,
"step": 518500
},
{
"epoch": 3.2816115937630412,
"grad_norm": 1.8511393070220947,
"learning_rate": 1.803112442788312e-05,
"loss": 2.454196533203125,
"step": 519000
},
{
"epoch": 3.2847730692868975,
"grad_norm": 1.873582124710083,
"learning_rate": 1.7951773959080552e-05,
"loss": 2.453672119140625,
"step": 519500
},
{
"epoch": 3.287934544810754,
"grad_norm": 1.8843079805374146,
"learning_rate": 1.7872423490277982e-05,
"loss": 2.456384765625,
"step": 520000
},
{
"epoch": 3.2910960203346105,
"grad_norm": 1.9618260860443115,
"learning_rate": 1.7793073021475413e-05,
"loss": 2.456720458984375,
"step": 520500
},
{
"epoch": 3.294257495858467,
"grad_norm": 1.9127881526947021,
"learning_rate": 1.7713722552672843e-05,
"loss": 2.45427880859375,
"step": 521000
},
{
"epoch": 3.2974189713823234,
"grad_norm": 1.7935330867767334,
"learning_rate": 1.7634372083870274e-05,
"loss": 2.451572998046875,
"step": 521500
},
{
"epoch": 3.30058044690618,
"grad_norm": 1.8051173686981201,
"learning_rate": 1.75550216150677e-05,
"loss": 2.457842041015625,
"step": 522000
},
{
"epoch": 3.3037419224300364,
"grad_norm": 1.899842619895935,
"learning_rate": 1.747567114626513e-05,
"loss": 2.454799072265625,
"step": 522500
},
{
"epoch": 3.306903397953893,
"grad_norm": 1.769952416419983,
"learning_rate": 1.739632067746256e-05,
"loss": 2.454308349609375,
"step": 523000
},
{
"epoch": 3.3100648734777494,
"grad_norm": 1.8426693677902222,
"learning_rate": 1.7316970208659995e-05,
"loss": 2.45426904296875,
"step": 523500
},
{
"epoch": 3.313226349001606,
"grad_norm": 1.8101383447647095,
"learning_rate": 1.7237619739857426e-05,
"loss": 2.453549072265625,
"step": 524000
},
{
"epoch": 3.3163878245254628,
"grad_norm": 1.8529447317123413,
"learning_rate": 1.7158269271054856e-05,
"loss": 2.453557861328125,
"step": 524500
},
{
"epoch": 3.319549300049319,
"grad_norm": 1.8531192541122437,
"learning_rate": 1.7078918802252283e-05,
"loss": 2.4491357421875,
"step": 525000
},
{
"epoch": 3.3227107755731753,
"grad_norm": 1.842976450920105,
"learning_rate": 1.6999568333449714e-05,
"loss": 2.4548642578125,
"step": 525500
},
{
"epoch": 3.325872251097032,
"grad_norm": 1.9068312644958496,
"learning_rate": 1.6920217864647144e-05,
"loss": 2.451682373046875,
"step": 526000
},
{
"epoch": 3.3290337266208887,
"grad_norm": 1.8012440204620361,
"learning_rate": 1.6840867395844574e-05,
"loss": 2.454109130859375,
"step": 526500
},
{
"epoch": 3.332195202144745,
"grad_norm": 1.8149930238723755,
"learning_rate": 1.6761516927042005e-05,
"loss": 2.452175048828125,
"step": 527000
},
{
"epoch": 3.3353566776686017,
"grad_norm": 1.8541282415390015,
"learning_rate": 1.668216645823944e-05,
"loss": 2.451037109375,
"step": 527500
},
{
"epoch": 3.338518153192458,
"grad_norm": 1.8474128246307373,
"learning_rate": 1.6602815989436866e-05,
"loss": 2.450594482421875,
"step": 528000
},
{
"epoch": 3.3416796287163146,
"grad_norm": 2.0225117206573486,
"learning_rate": 1.6523465520634296e-05,
"loss": 2.452268310546875,
"step": 528500
},
{
"epoch": 3.344841104240171,
"grad_norm": 1.8677397966384888,
"learning_rate": 1.6444115051831727e-05,
"loss": 2.4503798828125,
"step": 529000
},
{
"epoch": 3.3480025797640276,
"grad_norm": 1.8365589380264282,
"learning_rate": 1.6364764583029157e-05,
"loss": 2.454198486328125,
"step": 529500
},
{
"epoch": 3.351164055287884,
"grad_norm": 1.8791484832763672,
"learning_rate": 1.6285414114226587e-05,
"loss": 2.4436337890625,
"step": 530000
},
{
"epoch": 3.3543255308117406,
"grad_norm": 1.8730679750442505,
"learning_rate": 1.6206063645424018e-05,
"loss": 2.45110107421875,
"step": 530500
},
{
"epoch": 3.357487006335597,
"grad_norm": 1.8307521343231201,
"learning_rate": 1.6126713176621448e-05,
"loss": 2.450311279296875,
"step": 531000
},
{
"epoch": 3.3606484818594535,
"grad_norm": 1.8459255695343018,
"learning_rate": 1.604736270781888e-05,
"loss": 2.446857666015625,
"step": 531500
},
{
"epoch": 3.36380995738331,
"grad_norm": 1.8690210580825806,
"learning_rate": 1.596801223901631e-05,
"loss": 2.447150634765625,
"step": 532000
},
{
"epoch": 3.3669714329071665,
"grad_norm": 1.919494867324829,
"learning_rate": 1.588866177021374e-05,
"loss": 2.445688720703125,
"step": 532500
},
{
"epoch": 3.3701329084310228,
"grad_norm": 1.866140365600586,
"learning_rate": 1.580931130141117e-05,
"loss": 2.452443603515625,
"step": 533000
},
{
"epoch": 3.3732943839548795,
"grad_norm": 1.8772399425506592,
"learning_rate": 1.57299608326086e-05,
"loss": 2.4491044921875,
"step": 533500
},
{
"epoch": 3.3764558594787357,
"grad_norm": 1.8784788846969604,
"learning_rate": 1.565061036380603e-05,
"loss": 2.44586328125,
"step": 534000
},
{
"epoch": 3.3796173350025924,
"grad_norm": 1.8612768650054932,
"learning_rate": 1.5571259895003458e-05,
"loss": 2.450836181640625,
"step": 534500
},
{
"epoch": 3.3827788105264487,
"grad_norm": 1.7726970911026,
"learning_rate": 1.549190942620089e-05,
"loss": 2.45125537109375,
"step": 535000
},
{
"epoch": 3.3859402860503054,
"grad_norm": 1.8198189735412598,
"learning_rate": 1.5412558957398322e-05,
"loss": 2.452689697265625,
"step": 535500
},
{
"epoch": 3.389101761574162,
"grad_norm": 1.8851004838943481,
"learning_rate": 1.5333208488595752e-05,
"loss": 2.453731689453125,
"step": 536000
},
{
"epoch": 3.3922632370980184,
"grad_norm": 1.7908233404159546,
"learning_rate": 1.5253858019793183e-05,
"loss": 2.4510439453125,
"step": 536500
},
{
"epoch": 3.395424712621875,
"grad_norm": 1.8189315795898438,
"learning_rate": 1.5174507550990613e-05,
"loss": 2.446853271484375,
"step": 537000
},
{
"epoch": 3.3985861881457313,
"grad_norm": 1.9149062633514404,
"learning_rate": 1.509515708218804e-05,
"loss": 2.44935107421875,
"step": 537500
},
{
"epoch": 3.401747663669588,
"grad_norm": 1.8869186639785767,
"learning_rate": 1.5015806613385472e-05,
"loss": 2.450443603515625,
"step": 538000
},
{
"epoch": 3.4049091391934443,
"grad_norm": 1.9008139371871948,
"learning_rate": 1.4936456144582903e-05,
"loss": 2.449310546875,
"step": 538500
},
{
"epoch": 3.408070614717301,
"grad_norm": 1.89591646194458,
"learning_rate": 1.4857105675780333e-05,
"loss": 2.452410400390625,
"step": 539000
},
{
"epoch": 3.4112320902411573,
"grad_norm": 1.8382610082626343,
"learning_rate": 1.4777755206977764e-05,
"loss": 2.44575146484375,
"step": 539500
},
{
"epoch": 3.414393565765014,
"grad_norm": 1.9256197214126587,
"learning_rate": 1.4698404738175194e-05,
"loss": 2.4523486328125,
"step": 540000
},
{
"epoch": 3.41755504128887,
"grad_norm": 1.8464560508728027,
"learning_rate": 1.4619054269372623e-05,
"loss": 2.445962646484375,
"step": 540500
},
{
"epoch": 3.420716516812727,
"grad_norm": 1.8324875831604004,
"learning_rate": 1.4539703800570053e-05,
"loss": 2.44436669921875,
"step": 541000
},
{
"epoch": 3.423877992336583,
"grad_norm": 1.814501404762268,
"learning_rate": 1.4460353331767484e-05,
"loss": 2.44619580078125,
"step": 541500
},
{
"epoch": 3.42703946786044,
"grad_norm": 1.8813005685806274,
"learning_rate": 1.4381002862964916e-05,
"loss": 2.448772705078125,
"step": 542000
},
{
"epoch": 3.430200943384296,
"grad_norm": 1.8425790071487427,
"learning_rate": 1.4301652394162346e-05,
"loss": 2.445496337890625,
"step": 542500
},
{
"epoch": 3.433362418908153,
"grad_norm": 1.8446074724197388,
"learning_rate": 1.4222301925359777e-05,
"loss": 2.446618408203125,
"step": 543000
},
{
"epoch": 3.436523894432009,
"grad_norm": 1.82999587059021,
"learning_rate": 1.4142951456557205e-05,
"loss": 2.448857421875,
"step": 543500
},
{
"epoch": 3.439685369955866,
"grad_norm": 1.877970576286316,
"learning_rate": 1.4063600987754636e-05,
"loss": 2.44383154296875,
"step": 544000
},
{
"epoch": 3.442846845479722,
"grad_norm": 1.8475658893585205,
"learning_rate": 1.3984250518952066e-05,
"loss": 2.4407734375,
"step": 544500
},
{
"epoch": 3.446008321003579,
"grad_norm": 1.773900032043457,
"learning_rate": 1.3904900050149497e-05,
"loss": 2.4462802734375,
"step": 545000
},
{
"epoch": 3.4491697965274355,
"grad_norm": 1.8530097007751465,
"learning_rate": 1.3825549581346927e-05,
"loss": 2.45056982421875,
"step": 545500
},
{
"epoch": 3.4523312720512918,
"grad_norm": 1.8936686515808105,
"learning_rate": 1.374619911254436e-05,
"loss": 2.448995849609375,
"step": 546000
},
{
"epoch": 3.4554927475751485,
"grad_norm": 1.94265878200531,
"learning_rate": 1.366684864374179e-05,
"loss": 2.447746826171875,
"step": 546500
},
{
"epoch": 3.4586542230990047,
"grad_norm": 1.8119094371795654,
"learning_rate": 1.3587498174939217e-05,
"loss": 2.445254150390625,
"step": 547000
},
{
"epoch": 3.4618156986228614,
"grad_norm": 1.8417136669158936,
"learning_rate": 1.3508147706136647e-05,
"loss": 2.447227783203125,
"step": 547500
},
{
"epoch": 3.4649771741467177,
"grad_norm": 1.8808671236038208,
"learning_rate": 1.342879723733408e-05,
"loss": 2.447950927734375,
"step": 548000
},
{
"epoch": 3.4681386496705744,
"grad_norm": 1.8847047090530396,
"learning_rate": 1.334944676853151e-05,
"loss": 2.445017822265625,
"step": 548500
},
{
"epoch": 3.4713001251944307,
"grad_norm": 1.9782164096832275,
"learning_rate": 1.327009629972894e-05,
"loss": 2.448794921875,
"step": 549000
},
{
"epoch": 3.4744616007182874,
"grad_norm": 1.8095091581344604,
"learning_rate": 1.319074583092637e-05,
"loss": 2.445968017578125,
"step": 549500
},
{
"epoch": 3.4776230762421436,
"grad_norm": 1.8193165063858032,
"learning_rate": 1.31113953621238e-05,
"loss": 2.4478671875,
"step": 550000
},
{
"epoch": 3.4807845517660003,
"grad_norm": 1.8594002723693848,
"learning_rate": 1.303204489332123e-05,
"loss": 2.446626708984375,
"step": 550500
},
{
"epoch": 3.4839460272898566,
"grad_norm": 1.9951618909835815,
"learning_rate": 1.295269442451866e-05,
"loss": 2.444819580078125,
"step": 551000
},
{
"epoch": 3.4871075028137133,
"grad_norm": 1.8649407625198364,
"learning_rate": 1.287334395571609e-05,
"loss": 2.444177978515625,
"step": 551500
},
{
"epoch": 3.4902689783375695,
"grad_norm": 1.8845462799072266,
"learning_rate": 1.2793993486913523e-05,
"loss": 2.448691650390625,
"step": 552000
},
{
"epoch": 3.4934304538614263,
"grad_norm": 1.9054014682769775,
"learning_rate": 1.2714643018110953e-05,
"loss": 2.4426123046875,
"step": 552500
},
{
"epoch": 3.4965919293852825,
"grad_norm": 1.885680913925171,
"learning_rate": 1.263529254930838e-05,
"loss": 2.443014892578125,
"step": 553000
},
{
"epoch": 3.499753404909139,
"grad_norm": 1.8507832288742065,
"learning_rate": 1.2555942080505812e-05,
"loss": 2.445064697265625,
"step": 553500
},
{
"epoch": 3.5029148804329955,
"grad_norm": 1.8325011730194092,
"learning_rate": 1.2476591611703243e-05,
"loss": 2.444703857421875,
"step": 554000
},
{
"epoch": 3.506076355956852,
"grad_norm": 1.9163662195205688,
"learning_rate": 1.2397241142900673e-05,
"loss": 2.44217919921875,
"step": 554500
},
{
"epoch": 3.509237831480709,
"grad_norm": 1.7925523519515991,
"learning_rate": 1.2317890674098103e-05,
"loss": 2.4436337890625,
"step": 555000
},
{
"epoch": 3.512399307004565,
"grad_norm": 1.9178029298782349,
"learning_rate": 1.2238540205295534e-05,
"loss": 2.444232421875,
"step": 555500
},
{
"epoch": 3.5155607825284214,
"grad_norm": 1.8192750215530396,
"learning_rate": 1.2159189736492964e-05,
"loss": 2.449568603515625,
"step": 556000
},
{
"epoch": 3.518722258052278,
"grad_norm": 1.890152931213379,
"learning_rate": 1.2079839267690395e-05,
"loss": 2.438968994140625,
"step": 556500
},
{
"epoch": 3.521883733576135,
"grad_norm": 1.8981515169143677,
"learning_rate": 1.2000488798887823e-05,
"loss": 2.44128857421875,
"step": 557000
},
{
"epoch": 3.525045209099991,
"grad_norm": 1.9135870933532715,
"learning_rate": 1.1921138330085254e-05,
"loss": 2.44830810546875,
"step": 557500
},
{
"epoch": 3.528206684623848,
"grad_norm": 1.8042473793029785,
"learning_rate": 1.1841787861282686e-05,
"loss": 2.443125732421875,
"step": 558000
},
{
"epoch": 3.531368160147704,
"grad_norm": 1.9115724563598633,
"learning_rate": 1.1762437392480115e-05,
"loss": 2.438267333984375,
"step": 558500
},
{
"epoch": 3.5345296356715608,
"grad_norm": 1.8535972833633423,
"learning_rate": 1.1683086923677545e-05,
"loss": 2.435645263671875,
"step": 559000
},
{
"epoch": 3.537691111195417,
"grad_norm": 1.828763723373413,
"learning_rate": 1.1603736454874976e-05,
"loss": 2.44226953125,
"step": 559500
},
{
"epoch": 3.5408525867192737,
"grad_norm": 1.8682254552841187,
"learning_rate": 1.1524385986072406e-05,
"loss": 2.44458984375,
"step": 560000
},
{
"epoch": 3.54401406224313,
"grad_norm": 1.8215090036392212,
"learning_rate": 1.1445035517269836e-05,
"loss": 2.4424501953125,
"step": 560500
},
{
"epoch": 3.5471755377669867,
"grad_norm": 1.8779679536819458,
"learning_rate": 1.1365685048467267e-05,
"loss": 2.440673095703125,
"step": 561000
},
{
"epoch": 3.550337013290843,
"grad_norm": 1.8382534980773926,
"learning_rate": 1.1286334579664697e-05,
"loss": 2.43988916015625,
"step": 561500
},
{
"epoch": 3.5534984888146997,
"grad_norm": 1.8693525791168213,
"learning_rate": 1.1206984110862128e-05,
"loss": 2.440709716796875,
"step": 562000
},
{
"epoch": 3.5566599643385564,
"grad_norm": 1.8565537929534912,
"learning_rate": 1.1127633642059558e-05,
"loss": 2.44348388671875,
"step": 562500
},
{
"epoch": 3.5598214398624126,
"grad_norm": 1.8466014862060547,
"learning_rate": 1.1048283173256987e-05,
"loss": 2.44055078125,
"step": 563000
},
{
"epoch": 3.562982915386269,
"grad_norm": 1.8586146831512451,
"learning_rate": 1.0968932704454419e-05,
"loss": 2.444595458984375,
"step": 563500
},
{
"epoch": 3.5661443909101256,
"grad_norm": 1.8469187021255493,
"learning_rate": 1.088958223565185e-05,
"loss": 2.43764404296875,
"step": 564000
},
{
"epoch": 3.5693058664339823,
"grad_norm": 1.908569097518921,
"learning_rate": 1.081023176684928e-05,
"loss": 2.440992919921875,
"step": 564500
},
{
"epoch": 3.5724673419578385,
"grad_norm": 1.8871933221817017,
"learning_rate": 1.0730881298046709e-05,
"loss": 2.444171142578125,
"step": 565000
},
{
"epoch": 3.575628817481695,
"grad_norm": 1.884829044342041,
"learning_rate": 1.065153082924414e-05,
"loss": 2.44093310546875,
"step": 565500
},
{
"epoch": 3.5787902930055515,
"grad_norm": 1.8414987325668335,
"learning_rate": 1.0572180360441571e-05,
"loss": 2.447325927734375,
"step": 566000
},
{
"epoch": 3.581951768529408,
"grad_norm": 1.8464195728302002,
"learning_rate": 1.0492829891639e-05,
"loss": 2.441845458984375,
"step": 566500
},
{
"epoch": 3.5851132440532645,
"grad_norm": 1.8108004331588745,
"learning_rate": 1.041347942283643e-05,
"loss": 2.440102294921875,
"step": 567000
},
{
"epoch": 3.588274719577121,
"grad_norm": 1.7962620258331299,
"learning_rate": 1.0334128954033862e-05,
"loss": 2.443111328125,
"step": 567500
},
{
"epoch": 3.5914361951009774,
"grad_norm": 1.8455406427383423,
"learning_rate": 1.0254778485231291e-05,
"loss": 2.442737060546875,
"step": 568000
},
{
"epoch": 3.594597670624834,
"grad_norm": 1.828534722328186,
"learning_rate": 1.0175428016428721e-05,
"loss": 2.435753662109375,
"step": 568500
},
{
"epoch": 3.5977591461486904,
"grad_norm": 1.810401201248169,
"learning_rate": 1.0096077547626152e-05,
"loss": 2.44002880859375,
"step": 569000
},
{
"epoch": 3.600920621672547,
"grad_norm": 1.9466774463653564,
"learning_rate": 1.0016727078823582e-05,
"loss": 2.441895263671875,
"step": 569500
},
{
"epoch": 3.6040820971964034,
"grad_norm": 1.9174081087112427,
"learning_rate": 9.937376610021013e-06,
"loss": 2.44030810546875,
"step": 570000
},
{
"epoch": 3.60724357272026,
"grad_norm": 1.8186631202697754,
"learning_rate": 9.858026141218443e-06,
"loss": 2.441958251953125,
"step": 570500
},
{
"epoch": 3.6104050482441163,
"grad_norm": 1.7663170099258423,
"learning_rate": 9.778675672415872e-06,
"loss": 2.44289599609375,
"step": 571000
},
{
"epoch": 3.613566523767973,
"grad_norm": 1.883945345878601,
"learning_rate": 9.699325203613304e-06,
"loss": 2.44241796875,
"step": 571500
},
{
"epoch": 3.6167279992918298,
"grad_norm": 1.9665865898132324,
"learning_rate": 9.619974734810734e-06,
"loss": 2.4351689453125,
"step": 572000
},
{
"epoch": 3.619889474815686,
"grad_norm": 1.8201476335525513,
"learning_rate": 9.540624266008163e-06,
"loss": 2.43949072265625,
"step": 572500
},
{
"epoch": 3.6230509503395423,
"grad_norm": 1.8966400623321533,
"learning_rate": 9.461273797205594e-06,
"loss": 2.438548828125,
"step": 573000
},
{
"epoch": 3.626212425863399,
"grad_norm": 1.8687849044799805,
"learning_rate": 9.381923328403026e-06,
"loss": 2.441631591796875,
"step": 573500
},
{
"epoch": 3.6293739013872557,
"grad_norm": 1.8166216611862183,
"learning_rate": 9.302572859600454e-06,
"loss": 2.43788916015625,
"step": 574000
},
{
"epoch": 3.632535376911112,
"grad_norm": 1.811508297920227,
"learning_rate": 9.223222390797885e-06,
"loss": 2.437996826171875,
"step": 574500
},
{
"epoch": 3.635696852434968,
"grad_norm": 1.9841291904449463,
"learning_rate": 9.143871921995315e-06,
"loss": 2.4378740234375,
"step": 575000
},
{
"epoch": 3.638858327958825,
"grad_norm": 1.8423569202423096,
"learning_rate": 9.064521453192746e-06,
"loss": 2.44152001953125,
"step": 575500
},
{
"epoch": 3.6420198034826816,
"grad_norm": 1.9433730840682983,
"learning_rate": 8.985170984390176e-06,
"loss": 2.442005615234375,
"step": 576000
},
{
"epoch": 3.645181279006538,
"grad_norm": 1.7779185771942139,
"learning_rate": 8.905820515587607e-06,
"loss": 2.44034765625,
"step": 576500
},
{
"epoch": 3.6483427545303946,
"grad_norm": 1.8657695055007935,
"learning_rate": 8.826470046785035e-06,
"loss": 2.438537841796875,
"step": 577000
},
{
"epoch": 3.651504230054251,
"grad_norm": 1.7974389791488647,
"learning_rate": 8.747119577982467e-06,
"loss": 2.43519384765625,
"step": 577500
},
{
"epoch": 3.6546657055781075,
"grad_norm": 1.9149723052978516,
"learning_rate": 8.667769109179898e-06,
"loss": 2.438570556640625,
"step": 578000
},
{
"epoch": 3.657827181101964,
"grad_norm": 1.9174214601516724,
"learning_rate": 8.588418640377328e-06,
"loss": 2.4412353515625,
"step": 578500
},
{
"epoch": 3.6609886566258205,
"grad_norm": 1.8361644744873047,
"learning_rate": 8.509068171574757e-06,
"loss": 2.440362548828125,
"step": 579000
},
{
"epoch": 3.6641501321496768,
"grad_norm": 1.855931282043457,
"learning_rate": 8.429717702772189e-06,
"loss": 2.43415625,
"step": 579500
},
{
"epoch": 3.6673116076735335,
"grad_norm": 1.8884010314941406,
"learning_rate": 8.35036723396962e-06,
"loss": 2.435652587890625,
"step": 580000
},
{
"epoch": 3.6704730831973897,
"grad_norm": 1.8420894145965576,
"learning_rate": 8.271016765167048e-06,
"loss": 2.435760009765625,
"step": 580500
},
{
"epoch": 3.6736345587212464,
"grad_norm": 1.9250836372375488,
"learning_rate": 8.191666296364479e-06,
"loss": 2.436638671875,
"step": 581000
},
{
"epoch": 3.676796034245103,
"grad_norm": 1.8133349418640137,
"learning_rate": 8.11231582756191e-06,
"loss": 2.442661376953125,
"step": 581500
},
{
"epoch": 3.6799575097689594,
"grad_norm": 1.8360625505447388,
"learning_rate": 8.03296535875934e-06,
"loss": 2.43381982421875,
"step": 582000
},
{
"epoch": 3.6831189852928157,
"grad_norm": 1.8330854177474976,
"learning_rate": 7.95361488995677e-06,
"loss": 2.439815185546875,
"step": 582500
},
{
"epoch": 3.6862804608166724,
"grad_norm": 1.851251244544983,
"learning_rate": 7.8742644211542e-06,
"loss": 2.44084228515625,
"step": 583000
},
{
"epoch": 3.689441936340529,
"grad_norm": 1.8458049297332764,
"learning_rate": 7.79491395235163e-06,
"loss": 2.432794677734375,
"step": 583500
},
{
"epoch": 3.6926034118643853,
"grad_norm": 1.8132705688476562,
"learning_rate": 7.715563483549061e-06,
"loss": 2.4350302734375,
"step": 584000
},
{
"epoch": 3.6957648873882416,
"grad_norm": 1.8961392641067505,
"learning_rate": 7.636213014746492e-06,
"loss": 2.437624755859375,
"step": 584500
},
{
"epoch": 3.6989263629120983,
"grad_norm": 1.8140231370925903,
"learning_rate": 7.556862545943921e-06,
"loss": 2.43572509765625,
"step": 585000
},
{
"epoch": 3.702087838435955,
"grad_norm": 1.9114110469818115,
"learning_rate": 7.477512077141352e-06,
"loss": 2.43691748046875,
"step": 585500
},
{
"epoch": 3.7052493139598113,
"grad_norm": 1.8391425609588623,
"learning_rate": 7.398161608338783e-06,
"loss": 2.434448974609375,
"step": 586000
},
{
"epoch": 3.708410789483668,
"grad_norm": 1.8610597848892212,
"learning_rate": 7.3188111395362125e-06,
"loss": 2.438320556640625,
"step": 586500
},
{
"epoch": 3.7115722650075242,
"grad_norm": 1.8171026706695557,
"learning_rate": 7.239460670733643e-06,
"loss": 2.43872998046875,
"step": 587000
},
{
"epoch": 3.714733740531381,
"grad_norm": 1.9607990980148315,
"learning_rate": 7.160110201931073e-06,
"loss": 2.441789306640625,
"step": 587500
},
{
"epoch": 3.717895216055237,
"grad_norm": 1.9308041334152222,
"learning_rate": 7.080759733128503e-06,
"loss": 2.43645703125,
"step": 588000
},
{
"epoch": 3.721056691579094,
"grad_norm": 1.915954351425171,
"learning_rate": 7.001409264325934e-06,
"loss": 2.435857666015625,
"step": 588500
},
{
"epoch": 3.72421816710295,
"grad_norm": 1.846395492553711,
"learning_rate": 6.922058795523365e-06,
"loss": 2.438666015625,
"step": 589000
},
{
"epoch": 3.727379642626807,
"grad_norm": 1.9419134855270386,
"learning_rate": 6.842708326720794e-06,
"loss": 2.433604248046875,
"step": 589500
},
{
"epoch": 3.730541118150663,
"grad_norm": 1.9746551513671875,
"learning_rate": 6.763357857918225e-06,
"loss": 2.4411376953125,
"step": 590000
},
{
"epoch": 3.73370259367452,
"grad_norm": 1.8442648649215698,
"learning_rate": 6.684007389115655e-06,
"loss": 2.43439501953125,
"step": 590500
},
{
"epoch": 3.736864069198376,
"grad_norm": 1.8980733156204224,
"learning_rate": 6.604656920313086e-06,
"loss": 2.43585400390625,
"step": 591000
},
{
"epoch": 3.740025544722233,
"grad_norm": 1.895359754562378,
"learning_rate": 6.525306451510516e-06,
"loss": 2.437965087890625,
"step": 591500
},
{
"epoch": 3.743187020246089,
"grad_norm": 1.8792415857315063,
"learning_rate": 6.445955982707946e-06,
"loss": 2.437791015625,
"step": 592000
},
{
"epoch": 3.7463484957699458,
"grad_norm": 1.8697702884674072,
"learning_rate": 6.366605513905377e-06,
"loss": 2.436945068359375,
"step": 592500
},
{
"epoch": 3.7495099712938025,
"grad_norm": 1.910974383354187,
"learning_rate": 6.287255045102806e-06,
"loss": 2.43442578125,
"step": 593000
},
{
"epoch": 3.7526714468176587,
"grad_norm": 1.8729419708251953,
"learning_rate": 6.2079045763002376e-06,
"loss": 2.43822216796875,
"step": 593500
},
{
"epoch": 3.755832922341515,
"grad_norm": 1.827894926071167,
"learning_rate": 6.128554107497667e-06,
"loss": 2.434888671875,
"step": 594000
},
{
"epoch": 3.7589943978653717,
"grad_norm": 1.8874504566192627,
"learning_rate": 6.049203638695098e-06,
"loss": 2.4370263671875,
"step": 594500
},
{
"epoch": 3.7621558733892284,
"grad_norm": 1.8262746334075928,
"learning_rate": 5.969853169892528e-06,
"loss": 2.445139404296875,
"step": 595000
},
{
"epoch": 3.7653173489130847,
"grad_norm": 1.8492931127548218,
"learning_rate": 5.890502701089958e-06,
"loss": 2.4354609375,
"step": 595500
},
{
"epoch": 3.768478824436941,
"grad_norm": 1.8327908515930176,
"learning_rate": 5.811152232287389e-06,
"loss": 2.436000732421875,
"step": 596000
},
{
"epoch": 3.7716402999607976,
"grad_norm": 2.0110836029052734,
"learning_rate": 5.731801763484819e-06,
"loss": 2.433419677734375,
"step": 596500
},
{
"epoch": 3.7748017754846543,
"grad_norm": 1.8390324115753174,
"learning_rate": 5.652451294682249e-06,
"loss": 2.434158203125,
"step": 597000
},
{
"epoch": 3.7779632510085106,
"grad_norm": 1.8950042724609375,
"learning_rate": 5.57310082587968e-06,
"loss": 2.43294140625,
"step": 597500
},
{
"epoch": 3.7811247265323673,
"grad_norm": 1.874310851097107,
"learning_rate": 5.49375035707711e-06,
"loss": 2.436384765625,
"step": 598000
},
{
"epoch": 3.7842862020562236,
"grad_norm": 1.846126675605774,
"learning_rate": 5.41439988827454e-06,
"loss": 2.431458251953125,
"step": 598500
},
{
"epoch": 3.7874476775800803,
"grad_norm": 1.8770949840545654,
"learning_rate": 5.3350494194719705e-06,
"loss": 2.437216064453125,
"step": 599000
},
{
"epoch": 3.7906091531039365,
"grad_norm": 1.9436699151992798,
"learning_rate": 5.255698950669401e-06,
"loss": 2.436260986328125,
"step": 599500
},
{
"epoch": 3.7937706286277932,
"grad_norm": 1.9122369289398193,
"learning_rate": 5.1763484818668305e-06,
"loss": 2.43215478515625,
"step": 600000
},
{
"epoch": 3.7969321041516495,
"grad_norm": 1.8902041912078857,
"learning_rate": 5.096998013064262e-06,
"loss": 2.4327666015625,
"step": 600500
},
{
"epoch": 3.800093579675506,
"grad_norm": 1.8429012298583984,
"learning_rate": 5.017647544261691e-06,
"loss": 2.43523828125,
"step": 601000
},
{
"epoch": 3.8032550551993625,
"grad_norm": 1.7744569778442383,
"learning_rate": 4.938297075459123e-06,
"loss": 2.43950927734375,
"step": 601500
},
{
"epoch": 3.806416530723219,
"grad_norm": 1.9168301820755005,
"learning_rate": 4.858946606656552e-06,
"loss": 2.434794921875,
"step": 602000
},
{
"epoch": 3.809578006247076,
"grad_norm": 1.8670744895935059,
"learning_rate": 4.779596137853983e-06,
"loss": 2.431747314453125,
"step": 602500
},
{
"epoch": 3.812739481770932,
"grad_norm": 1.8653948307037354,
"learning_rate": 4.700245669051413e-06,
"loss": 2.432641845703125,
"step": 603000
},
{
"epoch": 3.8159009572947884,
"grad_norm": 1.9850834608078003,
"learning_rate": 4.6208952002488435e-06,
"loss": 2.435759033203125,
"step": 603500
},
{
"epoch": 3.819062432818645,
"grad_norm": 1.8792845010757446,
"learning_rate": 4.541544731446273e-06,
"loss": 2.428632080078125,
"step": 604000
},
{
"epoch": 3.822223908342502,
"grad_norm": 1.856383204460144,
"learning_rate": 4.462194262643704e-06,
"loss": 2.43512841796875,
"step": 604500
},
{
"epoch": 3.825385383866358,
"grad_norm": 1.9730868339538574,
"learning_rate": 4.382843793841134e-06,
"loss": 2.43049755859375,
"step": 605000
},
{
"epoch": 3.8285468593902143,
"grad_norm": 1.8719202280044556,
"learning_rate": 4.303493325038564e-06,
"loss": 2.4310849609375,
"step": 605500
},
{
"epoch": 3.831708334914071,
"grad_norm": 1.9522733688354492,
"learning_rate": 4.224142856235995e-06,
"loss": 2.432956298828125,
"step": 606000
},
{
"epoch": 3.8348698104379277,
"grad_norm": 1.8189289569854736,
"learning_rate": 4.144792387433425e-06,
"loss": 2.426925537109375,
"step": 606500
},
{
"epoch": 3.838031285961784,
"grad_norm": 1.863081455230713,
"learning_rate": 4.065441918630856e-06,
"loss": 2.423583251953125,
"step": 607000
},
{
"epoch": 3.8411927614856407,
"grad_norm": 1.9260568618774414,
"learning_rate": 3.986091449828286e-06,
"loss": 2.43239111328125,
"step": 607500
},
{
"epoch": 3.844354237009497,
"grad_norm": 1.9263060092926025,
"learning_rate": 3.906740981025716e-06,
"loss": 2.42907373046875,
"step": 608000
},
{
"epoch": 3.8475157125333537,
"grad_norm": 1.8558999300003052,
"learning_rate": 3.827390512223147e-06,
"loss": 2.4392568359375,
"step": 608500
},
{
"epoch": 3.85067718805721,
"grad_norm": 1.9042233228683472,
"learning_rate": 3.748040043420577e-06,
"loss": 2.430118896484375,
"step": 609000
},
{
"epoch": 3.8538386635810666,
"grad_norm": 1.8519482612609863,
"learning_rate": 3.668689574618007e-06,
"loss": 2.434379638671875,
"step": 609500
},
{
"epoch": 3.857000139104923,
"grad_norm": 1.8047749996185303,
"learning_rate": 3.5893391058154377e-06,
"loss": 2.430823974609375,
"step": 610000
},
{
"epoch": 3.8601616146287796,
"grad_norm": 1.924051284790039,
"learning_rate": 3.5099886370128677e-06,
"loss": 2.43474755859375,
"step": 610500
},
{
"epoch": 3.863323090152636,
"grad_norm": 1.822155237197876,
"learning_rate": 3.4306381682102977e-06,
"loss": 2.43241357421875,
"step": 611000
},
{
"epoch": 3.8664845656764926,
"grad_norm": 1.9684028625488281,
"learning_rate": 3.3512876994077286e-06,
"loss": 2.42247705078125,
"step": 611500
},
{
"epoch": 3.8696460412003493,
"grad_norm": 1.9292751550674438,
"learning_rate": 3.2719372306051586e-06,
"loss": 2.431715087890625,
"step": 612000
},
{
"epoch": 3.8728075167242055,
"grad_norm": 1.9320871829986572,
"learning_rate": 3.1925867618025886e-06,
"loss": 2.428010986328125,
"step": 612500
},
{
"epoch": 3.875968992248062,
"grad_norm": 1.9704115390777588,
"learning_rate": 3.1132362930000194e-06,
"loss": 2.43111767578125,
"step": 613000
},
{
"epoch": 3.8791304677719185,
"grad_norm": 1.9325289726257324,
"learning_rate": 3.0338858241974494e-06,
"loss": 2.434203125,
"step": 613500
},
{
"epoch": 3.882291943295775,
"grad_norm": 1.8643882274627686,
"learning_rate": 2.95453535539488e-06,
"loss": 2.4321611328125,
"step": 614000
},
{
"epoch": 3.8854534188196315,
"grad_norm": 1.786916971206665,
"learning_rate": 2.8751848865923103e-06,
"loss": 2.434728515625,
"step": 614500
},
{
"epoch": 3.8886148943434877,
"grad_norm": 1.937774658203125,
"learning_rate": 2.7958344177897407e-06,
"loss": 2.433371337890625,
"step": 615000
},
{
"epoch": 3.8917763698673444,
"grad_norm": 1.8756780624389648,
"learning_rate": 2.7164839489871707e-06,
"loss": 2.4320380859375,
"step": 615500
},
{
"epoch": 3.894937845391201,
"grad_norm": 1.8986276388168335,
"learning_rate": 2.637133480184601e-06,
"loss": 2.429661376953125,
"step": 616000
},
{
"epoch": 3.8980993209150574,
"grad_norm": 1.8440965414047241,
"learning_rate": 2.5577830113820315e-06,
"loss": 2.432765625,
"step": 616500
},
{
"epoch": 3.901260796438914,
"grad_norm": 1.9017361402511597,
"learning_rate": 2.4784325425794615e-06,
"loss": 2.429876708984375,
"step": 617000
},
{
"epoch": 3.9044222719627704,
"grad_norm": 1.912746787071228,
"learning_rate": 2.399082073776892e-06,
"loss": 2.432318603515625,
"step": 617500
},
{
"epoch": 3.907583747486627,
"grad_norm": 1.8669275045394897,
"learning_rate": 2.3197316049743224e-06,
"loss": 2.431711669921875,
"step": 618000
},
{
"epoch": 3.9107452230104833,
"grad_norm": 1.882030963897705,
"learning_rate": 2.240381136171753e-06,
"loss": 2.433046142578125,
"step": 618500
},
{
"epoch": 3.91390669853434,
"grad_norm": 1.92763352394104,
"learning_rate": 2.161030667369183e-06,
"loss": 2.433512939453125,
"step": 619000
},
{
"epoch": 3.9170681740581963,
"grad_norm": 1.832667350769043,
"learning_rate": 2.0816801985666132e-06,
"loss": 2.431409423828125,
"step": 619500
},
{
"epoch": 3.920229649582053,
"grad_norm": 1.8729039430618286,
"learning_rate": 2.0023297297640437e-06,
"loss": 2.4308095703125,
"step": 620000
},
{
"epoch": 3.9233911251059093,
"grad_norm": 1.9077296257019043,
"learning_rate": 1.9229792609614737e-06,
"loss": 2.429420654296875,
"step": 620500
},
{
"epoch": 3.926552600629766,
"grad_norm": 1.9839016199111938,
"learning_rate": 1.843628792158904e-06,
"loss": 2.425022705078125,
"step": 621000
},
{
"epoch": 3.9297140761536227,
"grad_norm": 1.852299690246582,
"learning_rate": 1.7642783233563345e-06,
"loss": 2.427850341796875,
"step": 621500
},
{
"epoch": 3.932875551677479,
"grad_norm": 1.9137951135635376,
"learning_rate": 1.684927854553765e-06,
"loss": 2.429792236328125,
"step": 622000
},
{
"epoch": 3.936037027201335,
"grad_norm": 1.906032919883728,
"learning_rate": 1.605577385751195e-06,
"loss": 2.424141845703125,
"step": 622500
},
{
"epoch": 3.939198502725192,
"grad_norm": 1.9414933919906616,
"learning_rate": 1.5262269169486253e-06,
"loss": 2.43278125,
"step": 623000
},
{
"epoch": 3.9423599782490486,
"grad_norm": 1.8752532005310059,
"learning_rate": 1.4468764481460556e-06,
"loss": 2.4314853515625,
"step": 623500
},
{
"epoch": 3.945521453772905,
"grad_norm": 1.867332935333252,
"learning_rate": 1.367525979343486e-06,
"loss": 2.4332744140625,
"step": 624000
},
{
"epoch": 3.948682929296761,
"grad_norm": 1.8243566751480103,
"learning_rate": 1.2881755105409162e-06,
"loss": 2.43213720703125,
"step": 624500
},
{
"epoch": 3.951844404820618,
"grad_norm": 1.9058722257614136,
"learning_rate": 1.2088250417383466e-06,
"loss": 2.431479248046875,
"step": 625000
},
{
"epoch": 3.9550058803444745,
"grad_norm": 1.8012287616729736,
"learning_rate": 1.1294745729357768e-06,
"loss": 2.431847900390625,
"step": 625500
},
{
"epoch": 3.958167355868331,
"grad_norm": 1.9177396297454834,
"learning_rate": 1.0501241041332073e-06,
"loss": 2.431138916015625,
"step": 626000
},
{
"epoch": 3.9613288313921875,
"grad_norm": 1.9069550037384033,
"learning_rate": 9.707736353306377e-07,
"loss": 2.4272275390625,
"step": 626500
},
{
"epoch": 3.9644903069160438,
"grad_norm": 1.8532055616378784,
"learning_rate": 8.914231665280678e-07,
"loss": 2.42448779296875,
"step": 627000
},
{
"epoch": 3.9676517824399005,
"grad_norm": 1.8894760608673096,
"learning_rate": 8.120726977254982e-07,
"loss": 2.433758056640625,
"step": 627500
},
{
"epoch": 3.9708132579637567,
"grad_norm": 1.841554045677185,
"learning_rate": 7.327222289229285e-07,
"loss": 2.42707958984375,
"step": 628000
},
{
"epoch": 3.9739747334876134,
"grad_norm": 1.8873754739761353,
"learning_rate": 6.533717601203588e-07,
"loss": 2.434695068359375,
"step": 628500
},
{
"epoch": 3.9771362090114697,
"grad_norm": 1.979251742362976,
"learning_rate": 5.74021291317789e-07,
"loss": 2.426421630859375,
"step": 629000
},
{
"epoch": 3.9802976845353264,
"grad_norm": 1.832371473312378,
"learning_rate": 4.946708225152194e-07,
"loss": 2.431343505859375,
"step": 629500
},
{
"epoch": 3.9834591600591827,
"grad_norm": 1.8508740663528442,
"learning_rate": 4.1532035371264974e-07,
"loss": 2.429571044921875,
"step": 630000
},
{
"epoch": 3.9866206355830394,
"grad_norm": 1.9497724771499634,
"learning_rate": 3.3596988491008006e-07,
"loss": 2.430441162109375,
"step": 630500
},
{
"epoch": 3.989782111106896,
"grad_norm": 1.8936879634857178,
"learning_rate": 2.566194161075104e-07,
"loss": 2.430122314453125,
"step": 631000
},
{
"epoch": 3.9929435866307523,
"grad_norm": 1.899812936782837,
"learning_rate": 1.772689473049407e-07,
"loss": 2.424232666015625,
"step": 631500
},
{
"epoch": 3.9961050621546086,
"grad_norm": 1.9367576837539673,
"learning_rate": 9.791847850237101e-08,
"loss": 2.429995361328125,
"step": 632000
},
{
"epoch": 3.9992665376784653,
"grad_norm": 1.9522849321365356,
"learning_rate": 1.8568009699801307e-08,
"loss": 2.424099609375,
"step": 632500
},
{
"epoch": 4.0,
"step": 632616,
"total_flos": 1.346593530204758e+18,
"train_loss": 3.0332739559738817,
"train_runtime": 103501.5888,
"train_samples_per_second": 391.177,
"train_steps_per_second": 6.112
}
],
"logging_steps": 500,
"max_steps": 632616,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 5000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.346593530204758e+18,
"train_batch_size": 64,
"trial_name": null,
"trial_params": null
}