Safetensors
English
bert
sportsbert-small / trainer_state.json
davidmezzetti's picture
Add model
d3b8afe
Raw
History Blame Contribute Delete
105 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 15.0,
"eval_steps": 500,
"global_step": 280365,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.026750842651543524,
"grad_norm": 0.5485039353370667,
"learning_rate": 0.0001996,
"loss": 7.60828076171875,
"step": 500
},
{
"epoch": 0.05350168530308705,
"grad_norm": 1.2309740781784058,
"learning_rate": 0.0003996,
"loss": 6.32032373046875,
"step": 1000
},
{
"epoch": 0.08025252795463057,
"grad_norm": 1.0301886796951294,
"learning_rate": 0.0005996,
"loss": 6.10592138671875,
"step": 1500
},
{
"epoch": 0.1070033706061741,
"grad_norm": 1.3331760168075562,
"learning_rate": 0.0007996,
"loss": 6.0112451171875,
"step": 2000
},
{
"epoch": 0.1337542132577176,
"grad_norm": 0.9183202385902405,
"learning_rate": 0.0009996,
"loss": 5.9263505859375,
"step": 2500
},
{
"epoch": 0.16050505590926115,
"grad_norm": 1.097238540649414,
"learning_rate": 0.0009982041638925377,
"loss": 5.589072265625,
"step": 3000
},
{
"epoch": 0.18725589856080466,
"grad_norm": 1.095755696296692,
"learning_rate": 0.0009964047289151207,
"loss": 4.58474560546875,
"step": 3500
},
{
"epoch": 0.2140067412123482,
"grad_norm": 1.1769925355911255,
"learning_rate": 0.0009946052939377036,
"loss": 3.539248291015625,
"step": 4000
},
{
"epoch": 0.2407575838638917,
"grad_norm": 1.0121588706970215,
"learning_rate": 0.0009928058589602866,
"loss": 3.05837939453125,
"step": 4500
},
{
"epoch": 0.2675084265154352,
"grad_norm": 1.0649367570877075,
"learning_rate": 0.0009910064239828694,
"loss": 2.84219775390625,
"step": 5000
},
{
"epoch": 0.29425926916697875,
"grad_norm": 1.0731533765792847,
"learning_rate": 0.0009892069890054523,
"loss": 2.70942041015625,
"step": 5500
},
{
"epoch": 0.3210101118185223,
"grad_norm": 0.9980011582374573,
"learning_rate": 0.000987407554028035,
"loss": 2.608668212890625,
"step": 6000
},
{
"epoch": 0.34776095447006583,
"grad_norm": 0.957374632358551,
"learning_rate": 0.0009856081190506181,
"loss": 2.534346923828125,
"step": 6500
},
{
"epoch": 0.3745117971216093,
"grad_norm": 0.9649298191070557,
"learning_rate": 0.000983808684073201,
"loss": 2.48132666015625,
"step": 7000
},
{
"epoch": 0.40126263977315285,
"grad_norm": 0.9456714987754822,
"learning_rate": 0.000982009249095784,
"loss": 2.434343505859375,
"step": 7500
},
{
"epoch": 0.4280134824246964,
"grad_norm": 0.9873167872428894,
"learning_rate": 0.0009802098141183668,
"loss": 2.3921748046875,
"step": 8000
},
{
"epoch": 0.45476432507623993,
"grad_norm": 0.9024070501327515,
"learning_rate": 0.0009784103791409498,
"loss": 2.356698486328125,
"step": 8500
},
{
"epoch": 0.4815151677277834,
"grad_norm": 0.9391950368881226,
"learning_rate": 0.0009766109441635327,
"loss": 2.317545166015625,
"step": 9000
},
{
"epoch": 0.508266010379327,
"grad_norm": 0.846376359462738,
"learning_rate": 0.0009748115091861156,
"loss": 2.295435546875,
"step": 9500
},
{
"epoch": 0.5350168530308704,
"grad_norm": 0.8609625101089478,
"learning_rate": 0.0009730120742086985,
"loss": 2.272924072265625,
"step": 10000
},
{
"epoch": 0.561767695682414,
"grad_norm": 0.8818221092224121,
"learning_rate": 0.0009712126392312815,
"loss": 2.254301025390625,
"step": 10500
},
{
"epoch": 0.5885185383339575,
"grad_norm": 0.8799985647201538,
"learning_rate": 0.0009694132042538643,
"loss": 2.234849609375,
"step": 11000
},
{
"epoch": 0.615269380985501,
"grad_norm": 0.8812091946601868,
"learning_rate": 0.0009676137692764472,
"loss": 2.210563720703125,
"step": 11500
},
{
"epoch": 0.6420202236370446,
"grad_norm": 0.8770820498466492,
"learning_rate": 0.0009658143342990301,
"loss": 2.197633056640625,
"step": 12000
},
{
"epoch": 0.6687710662885881,
"grad_norm": 0.8292478322982788,
"learning_rate": 0.000964014899321613,
"loss": 2.180857666015625,
"step": 12500
},
{
"epoch": 0.6955219089401317,
"grad_norm": 0.8617872595787048,
"learning_rate": 0.0009622154643441959,
"loss": 2.1712119140625,
"step": 13000
},
{
"epoch": 0.7222727515916751,
"grad_norm": 0.8561051487922668,
"learning_rate": 0.0009604160293667788,
"loss": 2.147574462890625,
"step": 13500
},
{
"epoch": 0.7490235942432186,
"grad_norm": 0.860202968120575,
"learning_rate": 0.0009586165943893618,
"loss": 2.13956689453125,
"step": 14000
},
{
"epoch": 0.7757744368947622,
"grad_norm": 0.8536654114723206,
"learning_rate": 0.0009568171594119447,
"loss": 2.12505810546875,
"step": 14500
},
{
"epoch": 0.8025252795463057,
"grad_norm": 0.8831987977027893,
"learning_rate": 0.0009550177244345276,
"loss": 2.119985107421875,
"step": 15000
},
{
"epoch": 0.8292761221978492,
"grad_norm": 0.8200844526290894,
"learning_rate": 0.0009532182894571105,
"loss": 2.101698486328125,
"step": 15500
},
{
"epoch": 0.8560269648493928,
"grad_norm": 0.8576837778091431,
"learning_rate": 0.0009514188544796934,
"loss": 2.09520947265625,
"step": 16000
},
{
"epoch": 0.8827778075009363,
"grad_norm": 0.9187654256820679,
"learning_rate": 0.0009496194195022763,
"loss": 2.0851923828125,
"step": 16500
},
{
"epoch": 0.9095286501524799,
"grad_norm": 0.8142674565315247,
"learning_rate": 0.0009478199845248592,
"loss": 2.082846923828125,
"step": 17000
},
{
"epoch": 0.9362794928040233,
"grad_norm": 0.8781545758247375,
"learning_rate": 0.0009460205495474421,
"loss": 2.063587890625,
"step": 17500
},
{
"epoch": 0.9630303354555668,
"grad_norm": 0.9054897427558899,
"learning_rate": 0.000944221114570025,
"loss": 2.049908447265625,
"step": 18000
},
{
"epoch": 0.9897811781071104,
"grad_norm": 0.8344267010688782,
"learning_rate": 0.0009424216795926079,
"loss": 2.0448255615234374,
"step": 18500
},
{
"epoch": 1.016532020758654,
"grad_norm": 0.8114071488380432,
"learning_rate": 0.0009406222446151908,
"loss": 2.0421280517578126,
"step": 19000
},
{
"epoch": 1.0432828634101974,
"grad_norm": 0.8525159358978271,
"learning_rate": 0.0009388228096377738,
"loss": 2.0354322509765623,
"step": 19500
},
{
"epoch": 1.0700337060617409,
"grad_norm": 0.8453242182731628,
"learning_rate": 0.0009370233746603567,
"loss": 2.022202880859375,
"step": 20000
},
{
"epoch": 1.0967845487132846,
"grad_norm": 0.8304527997970581,
"learning_rate": 0.0009352239396829396,
"loss": 2.0183052978515623,
"step": 20500
},
{
"epoch": 1.123535391364828,
"grad_norm": 0.8553816676139832,
"learning_rate": 0.0009334245047055226,
"loss": 2.007739013671875,
"step": 21000
},
{
"epoch": 1.1502862340163715,
"grad_norm": 0.7906451225280762,
"learning_rate": 0.0009316250697281054,
"loss": 2.0093092041015623,
"step": 21500
},
{
"epoch": 1.177037076667915,
"grad_norm": 0.8397617936134338,
"learning_rate": 0.0009298256347506882,
"loss": 1.99401416015625,
"step": 22000
},
{
"epoch": 1.2037879193194585,
"grad_norm": 0.8434298038482666,
"learning_rate": 0.0009280261997732711,
"loss": 1.9919859619140625,
"step": 22500
},
{
"epoch": 1.230538761971002,
"grad_norm": 0.7914469242095947,
"learning_rate": 0.0009262267647958541,
"loss": 1.9827418212890624,
"step": 23000
},
{
"epoch": 1.2572896046225457,
"grad_norm": 0.7564200758934021,
"learning_rate": 0.000924427329818437,
"loss": 1.9770035400390624,
"step": 23500
},
{
"epoch": 1.2840404472740892,
"grad_norm": 0.7818735241889954,
"learning_rate": 0.0009226278948410199,
"loss": 1.976789306640625,
"step": 24000
},
{
"epoch": 1.3107912899256327,
"grad_norm": 0.8098034262657166,
"learning_rate": 0.0009208284598636029,
"loss": 1.96844873046875,
"step": 24500
},
{
"epoch": 1.3375421325771761,
"grad_norm": 0.7436015009880066,
"learning_rate": 0.0009190290248861858,
"loss": 1.9586038818359375,
"step": 25000
},
{
"epoch": 1.3642929752287198,
"grad_norm": 0.8264908194541931,
"learning_rate": 0.0009172295899087687,
"loss": 1.958629150390625,
"step": 25500
},
{
"epoch": 1.3910438178802633,
"grad_norm": 0.7621483206748962,
"learning_rate": 0.0009154301549313517,
"loss": 1.9556282958984375,
"step": 26000
},
{
"epoch": 1.4177946605318068,
"grad_norm": 0.7337507009506226,
"learning_rate": 0.0009136307199539344,
"loss": 1.951950927734375,
"step": 26500
},
{
"epoch": 1.4445455031833503,
"grad_norm": 0.7802164554595947,
"learning_rate": 0.0009118312849765173,
"loss": 1.9429036865234375,
"step": 27000
},
{
"epoch": 1.4712963458348938,
"grad_norm": 0.8410242795944214,
"learning_rate": 0.0009100318499991002,
"loss": 1.9374320068359374,
"step": 27500
},
{
"epoch": 1.4980471884864373,
"grad_norm": 0.8862931728363037,
"learning_rate": 0.0009082324150216832,
"loss": 1.933148681640625,
"step": 28000
},
{
"epoch": 1.5247980311379807,
"grad_norm": 0.8088354468345642,
"learning_rate": 0.0009064329800442661,
"loss": 1.935772216796875,
"step": 28500
},
{
"epoch": 1.5515488737895242,
"grad_norm": 0.854947030544281,
"learning_rate": 0.000904633545066849,
"loss": 1.920808837890625,
"step": 29000
},
{
"epoch": 1.578299716441068,
"grad_norm": 0.8204390406608582,
"learning_rate": 0.000902834110089432,
"loss": 1.9213299560546875,
"step": 29500
},
{
"epoch": 1.6050505590926114,
"grad_norm": 0.7730057239532471,
"learning_rate": 0.0009010346751120149,
"loss": 1.9188778076171875,
"step": 30000
},
{
"epoch": 1.631801401744155,
"grad_norm": 0.7853319644927979,
"learning_rate": 0.0008992352401345978,
"loss": 1.9113128662109375,
"step": 30500
},
{
"epoch": 1.6585522443956986,
"grad_norm": 0.7655641436576843,
"learning_rate": 0.0008974358051571808,
"loss": 1.901947998046875,
"step": 31000
},
{
"epoch": 1.685303087047242,
"grad_norm": 0.8110830187797546,
"learning_rate": 0.0008956363701797636,
"loss": 1.899342529296875,
"step": 31500
},
{
"epoch": 1.7120539296987856,
"grad_norm": 0.8118914365768433,
"learning_rate": 0.0008938369352023464,
"loss": 1.903105712890625,
"step": 32000
},
{
"epoch": 1.738804772350329,
"grad_norm": 0.7817990183830261,
"learning_rate": 0.0008920375002249293,
"loss": 1.89413232421875,
"step": 32500
},
{
"epoch": 1.7655556150018725,
"grad_norm": 0.7612417936325073,
"learning_rate": 0.0008902380652475123,
"loss": 1.8900601806640625,
"step": 33000
},
{
"epoch": 1.792306457653416,
"grad_norm": 0.7485489249229431,
"learning_rate": 0.0008884386302700952,
"loss": 1.8879093017578126,
"step": 33500
},
{
"epoch": 1.8190573003049595,
"grad_norm": 0.8034965395927429,
"learning_rate": 0.0008866391952926781,
"loss": 1.890444091796875,
"step": 34000
},
{
"epoch": 1.845808142956503,
"grad_norm": 0.7788478136062622,
"learning_rate": 0.0008848397603152611,
"loss": 1.8812376708984375,
"step": 34500
},
{
"epoch": 1.8725589856080467,
"grad_norm": 0.7219562530517578,
"learning_rate": 0.000883040325337844,
"loss": 1.8801744384765624,
"step": 35000
},
{
"epoch": 1.8993098282595902,
"grad_norm": 0.7952810525894165,
"learning_rate": 0.0008812408903604269,
"loss": 1.879587890625,
"step": 35500
},
{
"epoch": 1.9260606709111339,
"grad_norm": 0.7137271761894226,
"learning_rate": 0.0008794414553830097,
"loss": 1.87626904296875,
"step": 36000
},
{
"epoch": 1.9528115135626773,
"grad_norm": 0.758965015411377,
"learning_rate": 0.0008776420204055927,
"loss": 1.867668701171875,
"step": 36500
},
{
"epoch": 1.9795623562142208,
"grad_norm": 0.8050095438957214,
"learning_rate": 0.0008758425854281755,
"loss": 1.87485888671875,
"step": 37000
},
{
"epoch": 2.0063131988657643,
"grad_norm": 0.7865944504737854,
"learning_rate": 0.0008740431504507584,
"loss": 1.863081787109375,
"step": 37500
},
{
"epoch": 2.033064041517308,
"grad_norm": 0.7477171421051025,
"learning_rate": 0.0008722437154733414,
"loss": 1.8576552734375,
"step": 38000
},
{
"epoch": 2.0598148841688513,
"grad_norm": 0.7551788687705994,
"learning_rate": 0.0008704442804959243,
"loss": 1.859656494140625,
"step": 38500
},
{
"epoch": 2.0865657268203948,
"grad_norm": 0.7140872478485107,
"learning_rate": 0.0008686448455185072,
"loss": 1.8525223388671874,
"step": 39000
},
{
"epoch": 2.1133165694719382,
"grad_norm": 0.7734881043434143,
"learning_rate": 0.0008668454105410901,
"loss": 1.8510196533203125,
"step": 39500
},
{
"epoch": 2.1400674121234817,
"grad_norm": 0.8050045967102051,
"learning_rate": 0.0008650459755636731,
"loss": 1.846701171875,
"step": 40000
},
{
"epoch": 2.166818254775025,
"grad_norm": 0.7739958763122559,
"learning_rate": 0.000863246540586256,
"loss": 1.851236083984375,
"step": 40500
},
{
"epoch": 2.193569097426569,
"grad_norm": 0.8139091730117798,
"learning_rate": 0.0008614471056088388,
"loss": 1.844859130859375,
"step": 41000
},
{
"epoch": 2.2203199400781126,
"grad_norm": 0.7834722399711609,
"learning_rate": 0.0008596476706314218,
"loss": 1.8390858154296874,
"step": 41500
},
{
"epoch": 2.247070782729656,
"grad_norm": 0.7798795700073242,
"learning_rate": 0.0008578482356540046,
"loss": 1.8333323974609375,
"step": 42000
},
{
"epoch": 2.2738216253811996,
"grad_norm": 0.7604835033416748,
"learning_rate": 0.0008560488006765875,
"loss": 1.8386741943359375,
"step": 42500
},
{
"epoch": 2.300572468032743,
"grad_norm": 0.7249457836151123,
"learning_rate": 0.0008542493656991704,
"loss": 1.8327294921875,
"step": 43000
},
{
"epoch": 2.3273233106842866,
"grad_norm": 0.7734972834587097,
"learning_rate": 0.0008524499307217534,
"loss": 1.82905810546875,
"step": 43500
},
{
"epoch": 2.35407415333583,
"grad_norm": 0.7545009255409241,
"learning_rate": 0.0008506504957443363,
"loss": 1.82435205078125,
"step": 44000
},
{
"epoch": 2.3808249959873735,
"grad_norm": 0.7153519988059998,
"learning_rate": 0.0008488510607669192,
"loss": 1.8159603271484375,
"step": 44500
},
{
"epoch": 2.407575838638917,
"grad_norm": 0.7727837562561035,
"learning_rate": 0.0008470516257895022,
"loss": 1.8193516845703126,
"step": 45000
},
{
"epoch": 2.4343266812904605,
"grad_norm": 0.7596606612205505,
"learning_rate": 0.000845252190812085,
"loss": 1.8226116943359374,
"step": 45500
},
{
"epoch": 2.461077523942004,
"grad_norm": 0.7583140730857849,
"learning_rate": 0.0008434527558346679,
"loss": 1.8143336181640626,
"step": 46000
},
{
"epoch": 2.487828366593548,
"grad_norm": 0.7500866055488586,
"learning_rate": 0.0008416533208572509,
"loss": 1.8145609130859375,
"step": 46500
},
{
"epoch": 2.5145792092450914,
"grad_norm": 0.7694220542907715,
"learning_rate": 0.0008398538858798338,
"loss": 1.8095145263671875,
"step": 47000
},
{
"epoch": 2.541330051896635,
"grad_norm": 0.7954797744750977,
"learning_rate": 0.0008380544509024166,
"loss": 1.810724365234375,
"step": 47500
},
{
"epoch": 2.5680808945481783,
"grad_norm": 0.7579993009567261,
"learning_rate": 0.0008362550159249995,
"loss": 1.8112454833984375,
"step": 48000
},
{
"epoch": 2.594831737199722,
"grad_norm": 0.8034877181053162,
"learning_rate": 0.0008344555809475825,
"loss": 1.8126644287109375,
"step": 48500
},
{
"epoch": 2.6215825798512653,
"grad_norm": 0.7455270290374756,
"learning_rate": 0.0008326561459701654,
"loss": 1.8030782470703124,
"step": 49000
},
{
"epoch": 2.648333422502809,
"grad_norm": 0.7592356204986572,
"learning_rate": 0.0008308567109927483,
"loss": 1.7986854248046875,
"step": 49500
},
{
"epoch": 2.6750842651543523,
"grad_norm": 0.7763079404830933,
"learning_rate": 0.0008290572760153312,
"loss": 1.79920166015625,
"step": 50000
},
{
"epoch": 2.7018351078058958,
"grad_norm": 0.7671984434127808,
"learning_rate": 0.0008272578410379141,
"loss": 1.7976671142578124,
"step": 50500
},
{
"epoch": 2.7285859504574397,
"grad_norm": 0.7768674492835999,
"learning_rate": 0.000825458406060497,
"loss": 1.797240234375,
"step": 51000
},
{
"epoch": 2.7553367931089827,
"grad_norm": 0.7949376702308655,
"learning_rate": 0.00082365897108308,
"loss": 1.7980609130859375,
"step": 51500
},
{
"epoch": 2.7820876357605266,
"grad_norm": 0.745426595211029,
"learning_rate": 0.0008218595361056629,
"loss": 1.792177490234375,
"step": 52000
},
{
"epoch": 2.80883847841207,
"grad_norm": 0.7549856305122375,
"learning_rate": 0.0008200601011282457,
"loss": 1.793860107421875,
"step": 52500
},
{
"epoch": 2.8355893210636136,
"grad_norm": 0.7651327252388,
"learning_rate": 0.0008182606661508286,
"loss": 1.790525146484375,
"step": 53000
},
{
"epoch": 2.862340163715157,
"grad_norm": 0.7845579981803894,
"learning_rate": 0.0008164612311734116,
"loss": 1.7910443115234376,
"step": 53500
},
{
"epoch": 2.8890910063667006,
"grad_norm": 0.7424607872962952,
"learning_rate": 0.0008146617961959945,
"loss": 1.784693603515625,
"step": 54000
},
{
"epoch": 2.915841849018244,
"grad_norm": 0.7229671478271484,
"learning_rate": 0.0008128623612185774,
"loss": 1.7821341552734375,
"step": 54500
},
{
"epoch": 2.9425926916697875,
"grad_norm": 0.7893908619880676,
"learning_rate": 0.0008110629262411602,
"loss": 1.7869644775390625,
"step": 55000
},
{
"epoch": 2.969343534321331,
"grad_norm": 0.7750260829925537,
"learning_rate": 0.0008092634912637432,
"loss": 1.7797003173828125,
"step": 55500
},
{
"epoch": 2.9960943769728745,
"grad_norm": 0.7690876126289368,
"learning_rate": 0.0008074640562863261,
"loss": 1.7731578369140626,
"step": 56000
},
{
"epoch": 3.022845219624418,
"grad_norm": 0.7383103966712952,
"learning_rate": 0.000805664621308909,
"loss": 1.7707724609375,
"step": 56500
},
{
"epoch": 3.0495960622759615,
"grad_norm": 0.8134902715682983,
"learning_rate": 0.000803865186331492,
"loss": 1.7714237060546876,
"step": 57000
},
{
"epoch": 3.0763469049275054,
"grad_norm": 0.7576144933700562,
"learning_rate": 0.0008020657513540749,
"loss": 1.7696915283203125,
"step": 57500
},
{
"epoch": 3.103097747579049,
"grad_norm": 0.7585620284080505,
"learning_rate": 0.0008002663163766577,
"loss": 1.765597900390625,
"step": 58000
},
{
"epoch": 3.1298485902305924,
"grad_norm": 0.7483615875244141,
"learning_rate": 0.0007984668813992407,
"loss": 1.7630478515625,
"step": 58500
},
{
"epoch": 3.156599432882136,
"grad_norm": 0.8620273470878601,
"learning_rate": 0.0007966674464218236,
"loss": 1.7678167724609375,
"step": 59000
},
{
"epoch": 3.1833502755336793,
"grad_norm": 0.7237071394920349,
"learning_rate": 0.0007948680114444064,
"loss": 1.76197021484375,
"step": 59500
},
{
"epoch": 3.210101118185223,
"grad_norm": 0.7121312618255615,
"learning_rate": 0.0007930685764669893,
"loss": 1.763649169921875,
"step": 60000
},
{
"epoch": 3.2368519608367663,
"grad_norm": 0.7549014687538147,
"learning_rate": 0.0007912691414895723,
"loss": 1.7582076416015624,
"step": 60500
},
{
"epoch": 3.26360280348831,
"grad_norm": 0.7666810154914856,
"learning_rate": 0.0007894697065121552,
"loss": 1.7574853515625,
"step": 61000
},
{
"epoch": 3.2903536461398533,
"grad_norm": 0.7484877705574036,
"learning_rate": 0.0007876702715347381,
"loss": 1.75898828125,
"step": 61500
},
{
"epoch": 3.3171044887913967,
"grad_norm": 0.736074686050415,
"learning_rate": 0.0007858708365573211,
"loss": 1.754572021484375,
"step": 62000
},
{
"epoch": 3.3438553314429402,
"grad_norm": 0.7367812395095825,
"learning_rate": 0.000784071401579904,
"loss": 1.75172998046875,
"step": 62500
},
{
"epoch": 3.370606174094484,
"grad_norm": 0.7703254818916321,
"learning_rate": 0.0007822719666024868,
"loss": 1.7571669921875,
"step": 63000
},
{
"epoch": 3.3973570167460276,
"grad_norm": 0.7424782514572144,
"learning_rate": 0.0007804725316250697,
"loss": 1.7445389404296876,
"step": 63500
},
{
"epoch": 3.424107859397571,
"grad_norm": 0.7220967411994934,
"learning_rate": 0.0007786730966476526,
"loss": 1.7419271240234375,
"step": 64000
},
{
"epoch": 3.4508587020491146,
"grad_norm": 0.7204769253730774,
"learning_rate": 0.0007768736616702355,
"loss": 1.7445679931640625,
"step": 64500
},
{
"epoch": 3.477609544700658,
"grad_norm": 0.7046904563903809,
"learning_rate": 0.0007750742266928184,
"loss": 1.7451087646484376,
"step": 65000
},
{
"epoch": 3.5043603873522016,
"grad_norm": 0.7349897027015686,
"learning_rate": 0.0007732747917154014,
"loss": 1.745401123046875,
"step": 65500
},
{
"epoch": 3.531111230003745,
"grad_norm": 0.775776207447052,
"learning_rate": 0.0007714753567379843,
"loss": 1.750186279296875,
"step": 66000
},
{
"epoch": 3.5578620726552885,
"grad_norm": 0.7389688491821289,
"learning_rate": 0.0007696759217605672,
"loss": 1.743634033203125,
"step": 66500
},
{
"epoch": 3.584612915306832,
"grad_norm": 0.7060774564743042,
"learning_rate": 0.0007678764867831502,
"loss": 1.74115087890625,
"step": 67000
},
{
"epoch": 3.611363757958376,
"grad_norm": 0.7634780406951904,
"learning_rate": 0.0007660770518057331,
"loss": 1.7426322021484375,
"step": 67500
},
{
"epoch": 3.638114600609919,
"grad_norm": 0.8267955780029297,
"learning_rate": 0.0007642776168283159,
"loss": 1.7281142578125,
"step": 68000
},
{
"epoch": 3.664865443261463,
"grad_norm": 0.783362865447998,
"learning_rate": 0.0007624781818508988,
"loss": 1.7331693115234375,
"step": 68500
},
{
"epoch": 3.6916162859130064,
"grad_norm": 0.7716281414031982,
"learning_rate": 0.0007606787468734817,
"loss": 1.7309781494140626,
"step": 69000
},
{
"epoch": 3.71836712856455,
"grad_norm": 0.7650080919265747,
"learning_rate": 0.0007588793118960646,
"loss": 1.731655029296875,
"step": 69500
},
{
"epoch": 3.7451179712160934,
"grad_norm": 0.7248140573501587,
"learning_rate": 0.0007570798769186475,
"loss": 1.7326429443359375,
"step": 70000
},
{
"epoch": 3.771868813867637,
"grad_norm": 0.7195979952812195,
"learning_rate": 0.0007552804419412305,
"loss": 1.72764990234375,
"step": 70500
},
{
"epoch": 3.7986196565191803,
"grad_norm": 0.7372161746025085,
"learning_rate": 0.0007534810069638134,
"loss": 1.7262598876953126,
"step": 71000
},
{
"epoch": 3.825370499170724,
"grad_norm": 0.7442724108695984,
"learning_rate": 0.0007516815719863963,
"loss": 1.7276502685546875,
"step": 71500
},
{
"epoch": 3.8521213418222673,
"grad_norm": 0.7324541211128235,
"learning_rate": 0.0007498821370089792,
"loss": 1.7292491455078125,
"step": 72000
},
{
"epoch": 3.8788721844738108,
"grad_norm": 0.7706058621406555,
"learning_rate": 0.0007480827020315622,
"loss": 1.7258216552734376,
"step": 72500
},
{
"epoch": 3.9056230271253547,
"grad_norm": 0.7710069417953491,
"learning_rate": 0.0007462832670541451,
"loss": 1.71714892578125,
"step": 73000
},
{
"epoch": 3.9323738697768977,
"grad_norm": 0.7392738461494446,
"learning_rate": 0.0007444838320767278,
"loss": 1.721431640625,
"step": 73500
},
{
"epoch": 3.9591247124284417,
"grad_norm": 0.7590123414993286,
"learning_rate": 0.0007426843970993108,
"loss": 1.7215382080078125,
"step": 74000
},
{
"epoch": 3.985875555079985,
"grad_norm": 0.7454897165298462,
"learning_rate": 0.0007408849621218937,
"loss": 1.71291064453125,
"step": 74500
},
{
"epoch": 4.012626397731529,
"grad_norm": 0.7414718866348267,
"learning_rate": 0.0007390855271444766,
"loss": 1.7112542724609374,
"step": 75000
},
{
"epoch": 4.039377240383072,
"grad_norm": 0.8065264821052551,
"learning_rate": 0.0007372860921670595,
"loss": 1.7093919677734375,
"step": 75500
},
{
"epoch": 4.066128083034616,
"grad_norm": 0.7673684358596802,
"learning_rate": 0.0007354866571896425,
"loss": 1.709315673828125,
"step": 76000
},
{
"epoch": 4.0928789256861595,
"grad_norm": 0.7181776762008667,
"learning_rate": 0.0007336872222122254,
"loss": 1.707806396484375,
"step": 76500
},
{
"epoch": 4.119629768337703,
"grad_norm": 0.746005654335022,
"learning_rate": 0.0007318877872348083,
"loss": 1.7106917724609374,
"step": 77000
},
{
"epoch": 4.1463806109892465,
"grad_norm": 0.7312119603157043,
"learning_rate": 0.0007300883522573913,
"loss": 1.703173095703125,
"step": 77500
},
{
"epoch": 4.1731314536407895,
"grad_norm": 0.7580727338790894,
"learning_rate": 0.0007282889172799742,
"loss": 1.7133687744140624,
"step": 78000
},
{
"epoch": 4.1998822962923335,
"grad_norm": 0.7454116344451904,
"learning_rate": 0.0007264894823025569,
"loss": 1.703045654296875,
"step": 78500
},
{
"epoch": 4.2266331389438765,
"grad_norm": 0.7500846982002258,
"learning_rate": 0.0007246900473251398,
"loss": 1.707533447265625,
"step": 79000
},
{
"epoch": 4.25338398159542,
"grad_norm": 0.7599788904190063,
"learning_rate": 0.0007228906123477228,
"loss": 1.7102509765625,
"step": 79500
},
{
"epoch": 4.2801348242469635,
"grad_norm": 0.7639349102973938,
"learning_rate": 0.0007210911773703057,
"loss": 1.7015328369140625,
"step": 80000
},
{
"epoch": 4.306885666898507,
"grad_norm": 0.7287936806678772,
"learning_rate": 0.0007192917423928886,
"loss": 1.6905933837890625,
"step": 80500
},
{
"epoch": 4.33363650955005,
"grad_norm": 0.7246515154838562,
"learning_rate": 0.0007174923074154716,
"loss": 1.690707275390625,
"step": 81000
},
{
"epoch": 4.360387352201594,
"grad_norm": 0.7810876965522766,
"learning_rate": 0.0007156928724380545,
"loss": 1.70224951171875,
"step": 81500
},
{
"epoch": 4.387138194853138,
"grad_norm": 0.7806092500686646,
"learning_rate": 0.0007138934374606374,
"loss": 1.692428466796875,
"step": 82000
},
{
"epoch": 4.413889037504681,
"grad_norm": 0.7730904221534729,
"learning_rate": 0.0007120940024832204,
"loss": 1.6919620361328125,
"step": 82500
},
{
"epoch": 4.440639880156225,
"grad_norm": 0.795973002910614,
"learning_rate": 0.0007102945675058032,
"loss": 1.6974315185546875,
"step": 83000
},
{
"epoch": 4.467390722807768,
"grad_norm": 0.7632553577423096,
"learning_rate": 0.0007084951325283861,
"loss": 1.6860362548828125,
"step": 83500
},
{
"epoch": 4.494141565459312,
"grad_norm": 0.7489613890647888,
"learning_rate": 0.0007066956975509689,
"loss": 1.69228857421875,
"step": 84000
},
{
"epoch": 4.520892408110855,
"grad_norm": 0.7227621674537659,
"learning_rate": 0.0007048962625735519,
"loss": 1.6939288330078126,
"step": 84500
},
{
"epoch": 4.547643250762399,
"grad_norm": 0.7552118301391602,
"learning_rate": 0.0007030968275961348,
"loss": 1.691815673828125,
"step": 85000
},
{
"epoch": 4.574394093413942,
"grad_norm": 0.7512255907058716,
"learning_rate": 0.0007012973926187177,
"loss": 1.687352783203125,
"step": 85500
},
{
"epoch": 4.601144936065486,
"grad_norm": 0.8033685088157654,
"learning_rate": 0.0006994979576413007,
"loss": 1.6840885009765625,
"step": 86000
},
{
"epoch": 4.627895778717029,
"grad_norm": 0.7521221041679382,
"learning_rate": 0.0006976985226638836,
"loss": 1.682308837890625,
"step": 86500
},
{
"epoch": 4.654646621368573,
"grad_norm": 0.6734914779663086,
"learning_rate": 0.0006958990876864665,
"loss": 1.678454833984375,
"step": 87000
},
{
"epoch": 4.681397464020117,
"grad_norm": 0.8489283919334412,
"learning_rate": 0.0006940996527090493,
"loss": 1.6863057861328126,
"step": 87500
},
{
"epoch": 4.70814830667166,
"grad_norm": 0.7051050662994385,
"learning_rate": 0.0006923002177316323,
"loss": 1.6883282470703125,
"step": 88000
},
{
"epoch": 4.734899149323204,
"grad_norm": 0.7193682193756104,
"learning_rate": 0.0006905007827542152,
"loss": 1.682403076171875,
"step": 88500
},
{
"epoch": 4.761649991974747,
"grad_norm": 0.863974392414093,
"learning_rate": 0.000688701347776798,
"loss": 1.684949462890625,
"step": 89000
},
{
"epoch": 4.788400834626291,
"grad_norm": 0.7394677400588989,
"learning_rate": 0.000686901912799381,
"loss": 1.683350341796875,
"step": 89500
},
{
"epoch": 4.815151677277834,
"grad_norm": 0.725368082523346,
"learning_rate": 0.0006851024778219639,
"loss": 1.6859183349609375,
"step": 90000
},
{
"epoch": 4.841902519929378,
"grad_norm": 0.7815067172050476,
"learning_rate": 0.0006833030428445468,
"loss": 1.6746912841796875,
"step": 90500
},
{
"epoch": 4.868653362580921,
"grad_norm": 0.8002517223358154,
"learning_rate": 0.0006815036078671298,
"loss": 1.6739063720703125,
"step": 91000
},
{
"epoch": 4.895404205232465,
"grad_norm": 0.7966354489326477,
"learning_rate": 0.0006797041728897127,
"loss": 1.68167041015625,
"step": 91500
},
{
"epoch": 4.922155047884008,
"grad_norm": 0.7149737477302551,
"learning_rate": 0.0006779047379122956,
"loss": 1.6688140869140624,
"step": 92000
},
{
"epoch": 4.948905890535552,
"grad_norm": 0.7511692643165588,
"learning_rate": 0.0006761053029348784,
"loss": 1.6625345458984375,
"step": 92500
},
{
"epoch": 4.975656733187096,
"grad_norm": 0.8088135123252869,
"learning_rate": 0.0006743058679574614,
"loss": 1.67000341796875,
"step": 93000
},
{
"epoch": 5.002407575838639,
"grad_norm": 0.6949880719184875,
"learning_rate": 0.0006725064329800443,
"loss": 1.6705894775390624,
"step": 93500
},
{
"epoch": 5.029158418490183,
"grad_norm": 0.7841982245445251,
"learning_rate": 0.0006707069980026272,
"loss": 1.6609549560546875,
"step": 94000
},
{
"epoch": 5.055909261141726,
"grad_norm": 0.6889007687568665,
"learning_rate": 0.0006689075630252101,
"loss": 1.6680516357421875,
"step": 94500
},
{
"epoch": 5.08266010379327,
"grad_norm": 0.7443187832832336,
"learning_rate": 0.000667108128047793,
"loss": 1.6604012451171875,
"step": 95000
},
{
"epoch": 5.109410946444813,
"grad_norm": 0.7050021886825562,
"learning_rate": 0.0006653086930703759,
"loss": 1.6658201904296874,
"step": 95500
},
{
"epoch": 5.136161789096357,
"grad_norm": 0.7626831531524658,
"learning_rate": 0.0006635092580929588,
"loss": 1.661433349609375,
"step": 96000
},
{
"epoch": 5.1629126317479,
"grad_norm": 0.7269205451011658,
"learning_rate": 0.0006617098231155418,
"loss": 1.651887451171875,
"step": 96500
},
{
"epoch": 5.189663474399444,
"grad_norm": 0.7043991088867188,
"learning_rate": 0.0006599103881381246,
"loss": 1.6586444091796875,
"step": 97000
},
{
"epoch": 5.216414317050987,
"grad_norm": 0.7657971978187561,
"learning_rate": 0.0006581109531607075,
"loss": 1.6585179443359375,
"step": 97500
},
{
"epoch": 5.243165159702531,
"grad_norm": 0.732473611831665,
"learning_rate": 0.0006563115181832905,
"loss": 1.6574713134765624,
"step": 98000
},
{
"epoch": 5.2699160023540745,
"grad_norm": 0.8044590353965759,
"learning_rate": 0.0006545120832058734,
"loss": 1.6489119873046876,
"step": 98500
},
{
"epoch": 5.296666845005618,
"grad_norm": 0.7754623293876648,
"learning_rate": 0.0006527126482284563,
"loss": 1.655093505859375,
"step": 99000
},
{
"epoch": 5.3234176876571615,
"grad_norm": 0.7008805274963379,
"learning_rate": 0.0006509132132510391,
"loss": 1.6585677490234374,
"step": 99500
},
{
"epoch": 5.3501685303087045,
"grad_norm": 0.7523577213287354,
"learning_rate": 0.0006491137782736221,
"loss": 1.650102783203125,
"step": 100000
},
{
"epoch": 5.3769193729602485,
"grad_norm": 0.7980531454086304,
"learning_rate": 0.000647314343296205,
"loss": 1.6546085205078125,
"step": 100500
},
{
"epoch": 5.4036702156117915,
"grad_norm": 0.7081476449966431,
"learning_rate": 0.0006455149083187879,
"loss": 1.6536826171875,
"step": 101000
},
{
"epoch": 5.430421058263335,
"grad_norm": 0.7847213745117188,
"learning_rate": 0.0006437154733413708,
"loss": 1.6518145751953126,
"step": 101500
},
{
"epoch": 5.4571719009148785,
"grad_norm": 0.7400954365730286,
"learning_rate": 0.0006419160383639537,
"loss": 1.6445126953125,
"step": 102000
},
{
"epoch": 5.483922743566422,
"grad_norm": 0.7608935236930847,
"learning_rate": 0.0006401166033865366,
"loss": 1.65615234375,
"step": 102500
},
{
"epoch": 5.510673586217965,
"grad_norm": 0.7587541341781616,
"learning_rate": 0.0006383171684091196,
"loss": 1.6529322509765625,
"step": 103000
},
{
"epoch": 5.537424428869509,
"grad_norm": 0.7919000387191772,
"learning_rate": 0.0006365177334317025,
"loss": 1.641713623046875,
"step": 103500
},
{
"epoch": 5.564175271521053,
"grad_norm": 0.7423938512802124,
"learning_rate": 0.0006347182984542854,
"loss": 1.6508837890625,
"step": 104000
},
{
"epoch": 5.590926114172596,
"grad_norm": 0.7470756769180298,
"learning_rate": 0.0006329188634768682,
"loss": 1.6429993896484374,
"step": 104500
},
{
"epoch": 5.61767695682414,
"grad_norm": 0.7286596894264221,
"learning_rate": 0.0006311194284994512,
"loss": 1.644008544921875,
"step": 105000
},
{
"epoch": 5.644427799475683,
"grad_norm": 0.7663483023643494,
"learning_rate": 0.0006293199935220341,
"loss": 1.64073681640625,
"step": 105500
},
{
"epoch": 5.671178642127227,
"grad_norm": 0.710423469543457,
"learning_rate": 0.000627520558544617,
"loss": 1.6406298828125,
"step": 106000
},
{
"epoch": 5.69792948477877,
"grad_norm": 0.75254225730896,
"learning_rate": 0.0006257211235671999,
"loss": 1.639049560546875,
"step": 106500
},
{
"epoch": 5.724680327430314,
"grad_norm": 0.8236501812934875,
"learning_rate": 0.0006239216885897828,
"loss": 1.65065087890625,
"step": 107000
},
{
"epoch": 5.751431170081857,
"grad_norm": 0.7569979429244995,
"learning_rate": 0.0006221222536123657,
"loss": 1.6329951171875,
"step": 107500
},
{
"epoch": 5.778182012733401,
"grad_norm": 0.6858986020088196,
"learning_rate": 0.0006203228186349486,
"loss": 1.6375078125,
"step": 108000
},
{
"epoch": 5.804932855384944,
"grad_norm": 0.7040875554084778,
"learning_rate": 0.0006185233836575316,
"loss": 1.636608154296875,
"step": 108500
},
{
"epoch": 5.831683698036488,
"grad_norm": 0.7736778259277344,
"learning_rate": 0.0006167239486801145,
"loss": 1.63436181640625,
"step": 109000
},
{
"epoch": 5.858434540688032,
"grad_norm": 0.7413392663002014,
"learning_rate": 0.0006149245137026974,
"loss": 1.632391357421875,
"step": 109500
},
{
"epoch": 5.885185383339575,
"grad_norm": 0.73166424036026,
"learning_rate": 0.0006131250787252803,
"loss": 1.6343319091796875,
"step": 110000
},
{
"epoch": 5.911936225991119,
"grad_norm": 0.7656437158584595,
"learning_rate": 0.0006113256437478632,
"loss": 1.636669189453125,
"step": 110500
},
{
"epoch": 5.938687068642662,
"grad_norm": 0.7018763422966003,
"learning_rate": 0.000609526208770446,
"loss": 1.6316654052734374,
"step": 111000
},
{
"epoch": 5.965437911294206,
"grad_norm": 0.7571375966072083,
"learning_rate": 0.000607726773793029,
"loss": 1.6229813232421875,
"step": 111500
},
{
"epoch": 5.992188753945749,
"grad_norm": 0.741434633731842,
"learning_rate": 0.0006059273388156119,
"loss": 1.62279638671875,
"step": 112000
},
{
"epoch": 6.018939596597293,
"grad_norm": 0.6990076899528503,
"learning_rate": 0.0006041279038381948,
"loss": 1.6260076904296874,
"step": 112500
},
{
"epoch": 6.045690439248836,
"grad_norm": 0.7509827613830566,
"learning_rate": 0.0006023284688607777,
"loss": 1.6244715576171875,
"step": 113000
},
{
"epoch": 6.07244128190038,
"grad_norm": 0.7830501198768616,
"learning_rate": 0.0006005290338833607,
"loss": 1.6222431640625,
"step": 113500
},
{
"epoch": 6.099192124551923,
"grad_norm": 0.7406209707260132,
"learning_rate": 0.0005987295989059436,
"loss": 1.62652294921875,
"step": 114000
},
{
"epoch": 6.125942967203467,
"grad_norm": 0.7363091707229614,
"learning_rate": 0.0005969301639285265,
"loss": 1.6254371337890625,
"step": 114500
},
{
"epoch": 6.152693809855011,
"grad_norm": 0.7810644507408142,
"learning_rate": 0.0005951307289511094,
"loss": 1.6230723876953126,
"step": 115000
},
{
"epoch": 6.179444652506554,
"grad_norm": 0.7414225339889526,
"learning_rate": 0.0005933312939736922,
"loss": 1.6193028564453125,
"step": 115500
},
{
"epoch": 6.206195495158098,
"grad_norm": 0.7613883018493652,
"learning_rate": 0.0005915318589962751,
"loss": 1.6171964111328125,
"step": 116000
},
{
"epoch": 6.232946337809641,
"grad_norm": 0.7151041626930237,
"learning_rate": 0.000589732424018858,
"loss": 1.6166754150390625,
"step": 116500
},
{
"epoch": 6.259697180461185,
"grad_norm": 0.7034673094749451,
"learning_rate": 0.000587932989041441,
"loss": 1.6149678955078124,
"step": 117000
},
{
"epoch": 6.286448023112728,
"grad_norm": 0.7332485318183899,
"learning_rate": 0.0005861335540640239,
"loss": 1.6184937744140624,
"step": 117500
},
{
"epoch": 6.313198865764272,
"grad_norm": 0.674255907535553,
"learning_rate": 0.0005843341190866068,
"loss": 1.61947412109375,
"step": 118000
},
{
"epoch": 6.339949708415815,
"grad_norm": 0.7474187612533569,
"learning_rate": 0.0005825346841091898,
"loss": 1.6153465576171875,
"step": 118500
},
{
"epoch": 6.366700551067359,
"grad_norm": 0.7223332524299622,
"learning_rate": 0.0005807352491317727,
"loss": 1.6163701171875,
"step": 119000
},
{
"epoch": 6.393451393718902,
"grad_norm": 0.7460731267929077,
"learning_rate": 0.0005789358141543556,
"loss": 1.6072003173828124,
"step": 119500
},
{
"epoch": 6.420202236370446,
"grad_norm": 0.6917471289634705,
"learning_rate": 0.0005771363791769386,
"loss": 1.611141357421875,
"step": 120000
},
{
"epoch": 6.4469530790219896,
"grad_norm": 0.791117787361145,
"learning_rate": 0.0005753369441995213,
"loss": 1.610268798828125,
"step": 120500
},
{
"epoch": 6.473703921673533,
"grad_norm": 0.7223137021064758,
"learning_rate": 0.0005735375092221042,
"loss": 1.6085001220703126,
"step": 121000
},
{
"epoch": 6.5004547643250765,
"grad_norm": 0.7701982259750366,
"learning_rate": 0.0005717380742446871,
"loss": 1.6051036376953125,
"step": 121500
},
{
"epoch": 6.52720560697662,
"grad_norm": 0.762233316898346,
"learning_rate": 0.0005699386392672701,
"loss": 1.61201123046875,
"step": 122000
},
{
"epoch": 6.5539564496281635,
"grad_norm": 0.7160345911979675,
"learning_rate": 0.000568139204289853,
"loss": 1.6138345947265624,
"step": 122500
},
{
"epoch": 6.5807072922797065,
"grad_norm": 0.735724151134491,
"learning_rate": 0.0005663397693124359,
"loss": 1.60338623046875,
"step": 123000
},
{
"epoch": 6.6074581349312504,
"grad_norm": 0.6977682113647461,
"learning_rate": 0.0005645403343350189,
"loss": 1.6083431396484376,
"step": 123500
},
{
"epoch": 6.6342089775827935,
"grad_norm": 0.7749923467636108,
"learning_rate": 0.0005627408993576018,
"loss": 1.601582275390625,
"step": 124000
},
{
"epoch": 6.660959820234337,
"grad_norm": 0.7442880868911743,
"learning_rate": 0.0005609414643801847,
"loss": 1.602933837890625,
"step": 124500
},
{
"epoch": 6.6877106628858805,
"grad_norm": 0.7766671776771545,
"learning_rate": 0.0005591420294027675,
"loss": 1.5964852294921874,
"step": 125000
},
{
"epoch": 6.714461505537424,
"grad_norm": 0.7357870936393738,
"learning_rate": 0.0005573425944253504,
"loss": 1.6035784912109374,
"step": 125500
},
{
"epoch": 6.741212348188968,
"grad_norm": 0.740944504737854,
"learning_rate": 0.0005555431594479333,
"loss": 1.60151123046875,
"step": 126000
},
{
"epoch": 6.767963190840511,
"grad_norm": 0.7281274795532227,
"learning_rate": 0.0005537437244705162,
"loss": 1.60528955078125,
"step": 126500
},
{
"epoch": 6.794714033492055,
"grad_norm": 0.7036533355712891,
"learning_rate": 0.0005519442894930992,
"loss": 1.601577392578125,
"step": 127000
},
{
"epoch": 6.821464876143598,
"grad_norm": 0.7406377792358398,
"learning_rate": 0.0005501448545156821,
"loss": 1.5987823486328125,
"step": 127500
},
{
"epoch": 6.848215718795142,
"grad_norm": 0.7304438352584839,
"learning_rate": 0.000548345419538265,
"loss": 1.5988348388671876,
"step": 128000
},
{
"epoch": 6.874966561446685,
"grad_norm": 0.7306429147720337,
"learning_rate": 0.000546545984560848,
"loss": 1.5863153076171874,
"step": 128500
},
{
"epoch": 6.901717404098229,
"grad_norm": 0.7541327476501465,
"learning_rate": 0.0005447465495834309,
"loss": 1.590626220703125,
"step": 129000
},
{
"epoch": 6.928468246749772,
"grad_norm": 0.751590371131897,
"learning_rate": 0.0005429471146060138,
"loss": 1.5914866943359376,
"step": 129500
},
{
"epoch": 6.955219089401316,
"grad_norm": 0.7596741914749146,
"learning_rate": 0.0005411476796285966,
"loss": 1.5921947021484375,
"step": 130000
},
{
"epoch": 6.981969932052859,
"grad_norm": 0.7774979472160339,
"learning_rate": 0.0005393482446511795,
"loss": 1.5906837158203124,
"step": 130500
},
{
"epoch": 7.008720774704403,
"grad_norm": 0.7522944211959839,
"learning_rate": 0.0005375488096737624,
"loss": 1.58911572265625,
"step": 131000
},
{
"epoch": 7.035471617355947,
"grad_norm": 0.710144579410553,
"learning_rate": 0.0005357493746963453,
"loss": 1.5968592529296874,
"step": 131500
},
{
"epoch": 7.06222246000749,
"grad_norm": 0.8234675526618958,
"learning_rate": 0.0005339499397189282,
"loss": 1.59000537109375,
"step": 132000
},
{
"epoch": 7.088973302659034,
"grad_norm": 0.7598490715026855,
"learning_rate": 0.0005321505047415112,
"loss": 1.5829810791015626,
"step": 132500
},
{
"epoch": 7.115724145310577,
"grad_norm": 0.7888147234916687,
"learning_rate": 0.0005303510697640941,
"loss": 1.5845816650390625,
"step": 133000
},
{
"epoch": 7.142474987962121,
"grad_norm": 0.7426342964172363,
"learning_rate": 0.000528551634786677,
"loss": 1.5815279541015625,
"step": 133500
},
{
"epoch": 7.169225830613664,
"grad_norm": 0.7744437456130981,
"learning_rate": 0.00052675219980926,
"loss": 1.581222412109375,
"step": 134000
},
{
"epoch": 7.195976673265208,
"grad_norm": 0.8049531579017639,
"learning_rate": 0.0005249527648318428,
"loss": 1.583486328125,
"step": 134500
},
{
"epoch": 7.222727515916751,
"grad_norm": 0.7207815051078796,
"learning_rate": 0.0005231533298544257,
"loss": 1.5812940673828124,
"step": 135000
},
{
"epoch": 7.249478358568295,
"grad_norm": 0.7493980526924133,
"learning_rate": 0.0005213538948770087,
"loss": 1.57850244140625,
"step": 135500
},
{
"epoch": 7.276229201219838,
"grad_norm": 0.7564734816551208,
"learning_rate": 0.0005195544598995915,
"loss": 1.5837806396484375,
"step": 136000
},
{
"epoch": 7.302980043871382,
"grad_norm": 0.7123373746871948,
"learning_rate": 0.0005177550249221744,
"loss": 1.5882777099609375,
"step": 136500
},
{
"epoch": 7.329730886522926,
"grad_norm": 0.7281443476676941,
"learning_rate": 0.0005159555899447573,
"loss": 1.5734921875,
"step": 137000
},
{
"epoch": 7.356481729174469,
"grad_norm": 0.751957356929779,
"learning_rate": 0.0005141561549673403,
"loss": 1.577795166015625,
"step": 137500
},
{
"epoch": 7.383232571826013,
"grad_norm": 0.7183213829994202,
"learning_rate": 0.0005123567199899232,
"loss": 1.5784765625,
"step": 138000
},
{
"epoch": 7.409983414477556,
"grad_norm": 0.7079334855079651,
"learning_rate": 0.0005105572850125061,
"loss": 1.5767528076171875,
"step": 138500
},
{
"epoch": 7.4367342571291,
"grad_norm": 0.759462833404541,
"learning_rate": 0.000508757850035089,
"loss": 1.576964111328125,
"step": 139000
},
{
"epoch": 7.463485099780643,
"grad_norm": 0.7075770497322083,
"learning_rate": 0.0005069584150576719,
"loss": 1.5788411865234375,
"step": 139500
},
{
"epoch": 7.490235942432187,
"grad_norm": 0.7932926416397095,
"learning_rate": 0.0005051589800802548,
"loss": 1.5738316650390625,
"step": 140000
},
{
"epoch": 7.51698678508373,
"grad_norm": 0.7543522715568542,
"learning_rate": 0.0005033595451028377,
"loss": 1.5748197021484376,
"step": 140500
},
{
"epoch": 7.543737627735274,
"grad_norm": 0.7194728255271912,
"learning_rate": 0.0005015601101254206,
"loss": 1.5730728759765624,
"step": 141000
},
{
"epoch": 7.570488470386817,
"grad_norm": 0.8007258772850037,
"learning_rate": 0.0004997606751480035,
"loss": 1.5670592041015625,
"step": 141500
},
{
"epoch": 7.597239313038361,
"grad_norm": 0.7342287302017212,
"learning_rate": 0.0004979612401705864,
"loss": 1.5717366943359374,
"step": 142000
},
{
"epoch": 7.623990155689905,
"grad_norm": 0.7782418131828308,
"learning_rate": 0.0004961618051931694,
"loss": 1.572611328125,
"step": 142500
},
{
"epoch": 7.650740998341448,
"grad_norm": 0.8047422170639038,
"learning_rate": 0.0004943623702157523,
"loss": 1.571184326171875,
"step": 143000
},
{
"epoch": 7.6774918409929915,
"grad_norm": 0.753219723701477,
"learning_rate": 0.0004925629352383352,
"loss": 1.56966943359375,
"step": 143500
},
{
"epoch": 7.704242683644535,
"grad_norm": 0.7463161945343018,
"learning_rate": 0.000490763500260918,
"loss": 1.566420654296875,
"step": 144000
},
{
"epoch": 7.7309935262960785,
"grad_norm": 0.7027589082717896,
"learning_rate": 0.000488964065283501,
"loss": 1.5628912353515625,
"step": 144500
},
{
"epoch": 7.7577443689476215,
"grad_norm": 0.7905352711677551,
"learning_rate": 0.00048716463030608386,
"loss": 1.563440673828125,
"step": 145000
},
{
"epoch": 7.7844952115991655,
"grad_norm": 0.7792290449142456,
"learning_rate": 0.0004853651953286668,
"loss": 1.5622423095703124,
"step": 145500
},
{
"epoch": 7.811246054250709,
"grad_norm": 0.7368639707565308,
"learning_rate": 0.0004835657603512497,
"loss": 1.565147216796875,
"step": 146000
},
{
"epoch": 7.837996896902252,
"grad_norm": 0.7256583571434021,
"learning_rate": 0.00048176632537383265,
"loss": 1.5632242431640626,
"step": 146500
},
{
"epoch": 7.8647477395537955,
"grad_norm": 0.7611755132675171,
"learning_rate": 0.0004799668903964156,
"loss": 1.55699609375,
"step": 147000
},
{
"epoch": 7.891498582205339,
"grad_norm": 0.7408255934715271,
"learning_rate": 0.0004781674554189984,
"loss": 1.5623577880859374,
"step": 147500
},
{
"epoch": 7.918249424856883,
"grad_norm": 0.7804815173149109,
"learning_rate": 0.00047636802044158133,
"loss": 1.5642452392578126,
"step": 148000
},
{
"epoch": 7.945000267508426,
"grad_norm": 0.7388637065887451,
"learning_rate": 0.00047456858546416426,
"loss": 1.561642333984375,
"step": 148500
},
{
"epoch": 7.97175111015997,
"grad_norm": 0.7259577512741089,
"learning_rate": 0.0004727691504867472,
"loss": 1.5545775146484375,
"step": 149000
},
{
"epoch": 7.998501952811513,
"grad_norm": 0.7551413774490356,
"learning_rate": 0.0004709697155093301,
"loss": 1.5569154052734375,
"step": 149500
},
{
"epoch": 8.025252795463057,
"grad_norm": 0.8080924153327942,
"learning_rate": 0.00046917028053191295,
"loss": 1.557394775390625,
"step": 150000
},
{
"epoch": 8.0520036381146,
"grad_norm": 0.7601260542869568,
"learning_rate": 0.0004673708455544959,
"loss": 1.5542308349609375,
"step": 150500
},
{
"epoch": 8.078754480766143,
"grad_norm": 0.7100428938865662,
"learning_rate": 0.0004655714105770788,
"loss": 1.5542353515625,
"step": 151000
},
{
"epoch": 8.105505323417688,
"grad_norm": 0.7985648512840271,
"learning_rate": 0.00046377197559966174,
"loss": 1.5555567626953124,
"step": 151500
},
{
"epoch": 8.132256166069231,
"grad_norm": 0.7503368854522705,
"learning_rate": 0.0004619725406222446,
"loss": 1.5509652099609375,
"step": 152000
},
{
"epoch": 8.159007008720774,
"grad_norm": 0.789909839630127,
"learning_rate": 0.00046017310564482755,
"loss": 1.55783251953125,
"step": 152500
},
{
"epoch": 8.185757851372319,
"grad_norm": 0.7529017925262451,
"learning_rate": 0.0004583736706674104,
"loss": 1.5534405517578125,
"step": 153000
},
{
"epoch": 8.212508694023862,
"grad_norm": 0.7600626349449158,
"learning_rate": 0.00045657423568999336,
"loss": 1.54395947265625,
"step": 153500
},
{
"epoch": 8.239259536675405,
"grad_norm": 0.7473117709159851,
"learning_rate": 0.0004547748007125763,
"loss": 1.5501925048828125,
"step": 154000
},
{
"epoch": 8.266010379326948,
"grad_norm": 0.724808931350708,
"learning_rate": 0.00045297536573515916,
"loss": 1.547697265625,
"step": 154500
},
{
"epoch": 8.292761221978493,
"grad_norm": 0.7340389490127563,
"learning_rate": 0.0004511759307577421,
"loss": 1.5496689453125,
"step": 155000
},
{
"epoch": 8.319512064630036,
"grad_norm": 0.7394249439239502,
"learning_rate": 0.00044937649578032497,
"loss": 1.539451416015625,
"step": 155500
},
{
"epoch": 8.346262907281579,
"grad_norm": 0.761161744594574,
"learning_rate": 0.0004475770608029079,
"loss": 1.548199951171875,
"step": 156000
},
{
"epoch": 8.373013749933122,
"grad_norm": 0.7441539764404297,
"learning_rate": 0.00044577762582549083,
"loss": 1.547853271484375,
"step": 156500
},
{
"epoch": 8.399764592584667,
"grad_norm": 0.7313327193260193,
"learning_rate": 0.0004439781908480737,
"loss": 1.5455986328125,
"step": 157000
},
{
"epoch": 8.42651543523621,
"grad_norm": 0.7411492466926575,
"learning_rate": 0.00044217875587065664,
"loss": 1.536748779296875,
"step": 157500
},
{
"epoch": 8.453266277887753,
"grad_norm": 0.7260663509368896,
"learning_rate": 0.0004403793208932395,
"loss": 1.5397850341796875,
"step": 158000
},
{
"epoch": 8.480017120539298,
"grad_norm": 0.7800888419151306,
"learning_rate": 0.00043857988591582245,
"loss": 1.5440208740234376,
"step": 158500
},
{
"epoch": 8.50676796319084,
"grad_norm": 0.7656701803207397,
"learning_rate": 0.0004367804509384053,
"loss": 1.546975341796875,
"step": 159000
},
{
"epoch": 8.533518805842384,
"grad_norm": 0.7696455121040344,
"learning_rate": 0.00043498101596098826,
"loss": 1.537057861328125,
"step": 159500
},
{
"epoch": 8.560269648493927,
"grad_norm": 0.7813064455986023,
"learning_rate": 0.0004331815809835712,
"loss": 1.54501953125,
"step": 160000
},
{
"epoch": 8.587020491145472,
"grad_norm": 0.7358993291854858,
"learning_rate": 0.00043138214600615406,
"loss": 1.5387371826171874,
"step": 160500
},
{
"epoch": 8.613771333797015,
"grad_norm": 0.767827033996582,
"learning_rate": 0.000429582711028737,
"loss": 1.5395753173828124,
"step": 161000
},
{
"epoch": 8.640522176448558,
"grad_norm": 0.7464857697486877,
"learning_rate": 0.00042778327605131987,
"loss": 1.5391500244140626,
"step": 161500
},
{
"epoch": 8.6672730191001,
"grad_norm": 0.762269914150238,
"learning_rate": 0.0004259838410739028,
"loss": 1.53144677734375,
"step": 162000
},
{
"epoch": 8.694023861751646,
"grad_norm": 0.7448802590370178,
"learning_rate": 0.00042418440609648573,
"loss": 1.53069091796875,
"step": 162500
},
{
"epoch": 8.720774704403189,
"grad_norm": 0.7250141501426697,
"learning_rate": 0.0004223849711190686,
"loss": 1.5396180419921874,
"step": 163000
},
{
"epoch": 8.747525547054732,
"grad_norm": 0.7308769226074219,
"learning_rate": 0.00042058553614165154,
"loss": 1.5377947998046875,
"step": 163500
},
{
"epoch": 8.774276389706277,
"grad_norm": 0.6930657625198364,
"learning_rate": 0.0004187861011642344,
"loss": 1.5352694091796875,
"step": 164000
},
{
"epoch": 8.80102723235782,
"grad_norm": 0.7253273725509644,
"learning_rate": 0.00041698666618681735,
"loss": 1.533189208984375,
"step": 164500
},
{
"epoch": 8.827778075009363,
"grad_norm": 0.7447793483734131,
"learning_rate": 0.0004151872312094003,
"loss": 1.528855224609375,
"step": 165000
},
{
"epoch": 8.854528917660906,
"grad_norm": 0.6933104991912842,
"learning_rate": 0.0004133877962319832,
"loss": 1.5295169677734375,
"step": 165500
},
{
"epoch": 8.88127976031245,
"grad_norm": 0.7189187407493591,
"learning_rate": 0.00041158836125456603,
"loss": 1.532017333984375,
"step": 166000
},
{
"epoch": 8.908030602963994,
"grad_norm": 0.7505865693092346,
"learning_rate": 0.00040978892627714896,
"loss": 1.5301812744140626,
"step": 166500
},
{
"epoch": 8.934781445615537,
"grad_norm": 0.7124701142311096,
"learning_rate": 0.0004079894912997319,
"loss": 1.5288656005859376,
"step": 167000
},
{
"epoch": 8.96153228826708,
"grad_norm": 0.7343282699584961,
"learning_rate": 0.0004061900563223148,
"loss": 1.5274080810546875,
"step": 167500
},
{
"epoch": 8.988283130918624,
"grad_norm": 0.7689876556396484,
"learning_rate": 0.00040439062134489776,
"loss": 1.523005126953125,
"step": 168000
},
{
"epoch": 9.015033973570167,
"grad_norm": 0.7539162039756775,
"learning_rate": 0.0004025911863674806,
"loss": 1.522071044921875,
"step": 168500
},
{
"epoch": 9.04178481622171,
"grad_norm": 0.7118518352508545,
"learning_rate": 0.0004007917513900635,
"loss": 1.5295810546875,
"step": 169000
},
{
"epoch": 9.068535658873255,
"grad_norm": 0.7213994264602661,
"learning_rate": 0.00039899231641264644,
"loss": 1.51806982421875,
"step": 169500
},
{
"epoch": 9.095286501524798,
"grad_norm": 0.6974011659622192,
"learning_rate": 0.00039719288143522937,
"loss": 1.51844580078125,
"step": 170000
},
{
"epoch": 9.122037344176341,
"grad_norm": 0.716418445110321,
"learning_rate": 0.0003953934464578123,
"loss": 1.519795654296875,
"step": 170500
},
{
"epoch": 9.148788186827884,
"grad_norm": 0.7481328248977661,
"learning_rate": 0.0003935940114803951,
"loss": 1.52125390625,
"step": 171000
},
{
"epoch": 9.17553902947943,
"grad_norm": 0.7122513651847839,
"learning_rate": 0.00039179457650297806,
"loss": 1.5144580078125,
"step": 171500
},
{
"epoch": 9.202289872130972,
"grad_norm": 0.7269628047943115,
"learning_rate": 0.000389995141525561,
"loss": 1.515474365234375,
"step": 172000
},
{
"epoch": 9.229040714782515,
"grad_norm": 0.7624303698539734,
"learning_rate": 0.0003881957065481439,
"loss": 1.5189844970703126,
"step": 172500
},
{
"epoch": 9.255791557434058,
"grad_norm": 0.7570505142211914,
"learning_rate": 0.0003863962715707268,
"loss": 1.5225419921875,
"step": 173000
},
{
"epoch": 9.282542400085603,
"grad_norm": 0.693884015083313,
"learning_rate": 0.00038459683659330967,
"loss": 1.5088955078125,
"step": 173500
},
{
"epoch": 9.309293242737146,
"grad_norm": 0.7123017311096191,
"learning_rate": 0.0003827974016158926,
"loss": 1.5151614990234374,
"step": 174000
},
{
"epoch": 9.33604408538869,
"grad_norm": 0.7079525589942932,
"learning_rate": 0.00038099796663847553,
"loss": 1.51560400390625,
"step": 174500
},
{
"epoch": 9.362794928040234,
"grad_norm": 0.6983911395072937,
"learning_rate": 0.00037919853166105846,
"loss": 1.5168739013671875,
"step": 175000
},
{
"epoch": 9.389545770691777,
"grad_norm": 0.7500879764556885,
"learning_rate": 0.00037739909668364134,
"loss": 1.5098924560546876,
"step": 175500
},
{
"epoch": 9.41629661334332,
"grad_norm": 0.7747844457626343,
"learning_rate": 0.0003755996617062242,
"loss": 1.5116580810546876,
"step": 176000
},
{
"epoch": 9.443047455994863,
"grad_norm": 0.7102504372596741,
"learning_rate": 0.00037380022672880715,
"loss": 1.513258544921875,
"step": 176500
},
{
"epoch": 9.469798298646408,
"grad_norm": 0.7059140801429749,
"learning_rate": 0.0003720007917513901,
"loss": 1.5125245361328126,
"step": 177000
},
{
"epoch": 9.496549141297951,
"grad_norm": 0.749907910823822,
"learning_rate": 0.000370201356773973,
"loss": 1.515197509765625,
"step": 177500
},
{
"epoch": 9.523299983949494,
"grad_norm": 0.8103958964347839,
"learning_rate": 0.0003684019217965559,
"loss": 1.5046624755859375,
"step": 178000
},
{
"epoch": 9.550050826601037,
"grad_norm": 0.7142824530601501,
"learning_rate": 0.0003666024868191388,
"loss": 1.5047467041015625,
"step": 178500
},
{
"epoch": 9.576801669252582,
"grad_norm": 0.7496501803398132,
"learning_rate": 0.0003648030518417217,
"loss": 1.5018792724609376,
"step": 179000
},
{
"epoch": 9.603552511904125,
"grad_norm": 0.7446944713592529,
"learning_rate": 0.0003630036168643046,
"loss": 1.5039600830078126,
"step": 179500
},
{
"epoch": 9.630303354555668,
"grad_norm": 0.7408098578453064,
"learning_rate": 0.0003612041818868875,
"loss": 1.51191552734375,
"step": 180000
},
{
"epoch": 9.657054197207213,
"grad_norm": 0.7211296558380127,
"learning_rate": 0.00035940474690947043,
"loss": 1.5069207763671875,
"step": 180500
},
{
"epoch": 9.683805039858756,
"grad_norm": 0.7187151312828064,
"learning_rate": 0.00035760531193205336,
"loss": 1.5100963134765626,
"step": 181000
},
{
"epoch": 9.710555882510299,
"grad_norm": 0.7317778468132019,
"learning_rate": 0.00035580587695463624,
"loss": 1.5019161376953125,
"step": 181500
},
{
"epoch": 9.737306725161842,
"grad_norm": 0.7727632522583008,
"learning_rate": 0.00035400644197721917,
"loss": 1.4982386474609375,
"step": 182000
},
{
"epoch": 9.764057567813387,
"grad_norm": 0.7385058403015137,
"learning_rate": 0.00035220700699980205,
"loss": 1.503185302734375,
"step": 182500
},
{
"epoch": 9.79080841046493,
"grad_norm": 0.7632474899291992,
"learning_rate": 0.000350407572022385,
"loss": 1.5056253662109376,
"step": 183000
},
{
"epoch": 9.817559253116473,
"grad_norm": 0.7118545174598694,
"learning_rate": 0.0003486081370449679,
"loss": 1.5031114501953124,
"step": 183500
},
{
"epoch": 9.844310095768016,
"grad_norm": 0.7350465059280396,
"learning_rate": 0.0003468087020675508,
"loss": 1.5033663330078124,
"step": 184000
},
{
"epoch": 9.87106093841956,
"grad_norm": 0.7367674112319946,
"learning_rate": 0.0003450092670901337,
"loss": 1.49630322265625,
"step": 184500
},
{
"epoch": 9.897811781071104,
"grad_norm": 0.7782302498817444,
"learning_rate": 0.0003432098321127166,
"loss": 1.4930860595703126,
"step": 185000
},
{
"epoch": 9.924562623722647,
"grad_norm": 0.7644042372703552,
"learning_rate": 0.0003414103971352995,
"loss": 1.4961455078125,
"step": 185500
},
{
"epoch": 9.951313466374192,
"grad_norm": 0.7162057161331177,
"learning_rate": 0.00033961096215788246,
"loss": 1.495660888671875,
"step": 186000
},
{
"epoch": 9.978064309025735,
"grad_norm": 0.7434723377227783,
"learning_rate": 0.00033781152718046533,
"loss": 1.502974853515625,
"step": 186500
},
{
"epoch": 10.004815151677278,
"grad_norm": 0.723887026309967,
"learning_rate": 0.0003360120922030482,
"loss": 1.4945015869140625,
"step": 187000
},
{
"epoch": 10.03156599432882,
"grad_norm": 0.760871946811676,
"learning_rate": 0.00033421265722563114,
"loss": 1.4913509521484376,
"step": 187500
},
{
"epoch": 10.058316836980365,
"grad_norm": 0.7596881985664368,
"learning_rate": 0.00033241322224821407,
"loss": 1.488368896484375,
"step": 188000
},
{
"epoch": 10.085067679631909,
"grad_norm": 0.7474902868270874,
"learning_rate": 0.000330613787270797,
"loss": 1.4905345458984376,
"step": 188500
},
{
"epoch": 10.111818522283452,
"grad_norm": 0.7005614638328552,
"learning_rate": 0.0003288143522933799,
"loss": 1.491130615234375,
"step": 189000
},
{
"epoch": 10.138569364934995,
"grad_norm": 0.7195891737937927,
"learning_rate": 0.00032701491731596276,
"loss": 1.4862049560546875,
"step": 189500
},
{
"epoch": 10.16532020758654,
"grad_norm": 0.6897690296173096,
"learning_rate": 0.0003252154823385457,
"loss": 1.4862669677734375,
"step": 190000
},
{
"epoch": 10.192071050238082,
"grad_norm": 0.7414577603340149,
"learning_rate": 0.0003234160473611286,
"loss": 1.488429931640625,
"step": 190500
},
{
"epoch": 10.218821892889626,
"grad_norm": 0.7738565802574158,
"learning_rate": 0.00032161661238371155,
"loss": 1.4845257568359376,
"step": 191000
},
{
"epoch": 10.24557273554117,
"grad_norm": 0.7129031419754028,
"learning_rate": 0.0003198171774062945,
"loss": 1.4905755615234375,
"step": 191500
},
{
"epoch": 10.272323578192713,
"grad_norm": 0.7828429937362671,
"learning_rate": 0.0003180177424288773,
"loss": 1.486052490234375,
"step": 192000
},
{
"epoch": 10.299074420844256,
"grad_norm": 0.7572054266929626,
"learning_rate": 0.00031621830745146023,
"loss": 1.48398974609375,
"step": 192500
},
{
"epoch": 10.3258252634958,
"grad_norm": 0.7738214731216431,
"learning_rate": 0.00031441887247404316,
"loss": 1.4865242919921875,
"step": 193000
},
{
"epoch": 10.352576106147344,
"grad_norm": 0.7070664167404175,
"learning_rate": 0.0003126194374966261,
"loss": 1.4872012939453125,
"step": 193500
},
{
"epoch": 10.379326948798887,
"grad_norm": 0.7296112775802612,
"learning_rate": 0.000310820002519209,
"loss": 1.48883984375,
"step": 194000
},
{
"epoch": 10.40607779145043,
"grad_norm": 0.7528848052024841,
"learning_rate": 0.00030902056754179185,
"loss": 1.482719482421875,
"step": 194500
},
{
"epoch": 10.432828634101973,
"grad_norm": 0.7360111474990845,
"learning_rate": 0.0003072211325643748,
"loss": 1.4778614501953125,
"step": 195000
},
{
"epoch": 10.459579476753518,
"grad_norm": 0.7553589940071106,
"learning_rate": 0.0003054216975869577,
"loss": 1.48245654296875,
"step": 195500
},
{
"epoch": 10.486330319405061,
"grad_norm": 0.7393176555633545,
"learning_rate": 0.00030362226260954064,
"loss": 1.4827337646484375,
"step": 196000
},
{
"epoch": 10.513081162056604,
"grad_norm": 0.7065622806549072,
"learning_rate": 0.0003018228276321235,
"loss": 1.4763675537109375,
"step": 196500
},
{
"epoch": 10.539832004708149,
"grad_norm": 0.7320563793182373,
"learning_rate": 0.0003000233926547064,
"loss": 1.4766748046875,
"step": 197000
},
{
"epoch": 10.566582847359692,
"grad_norm": 0.6942557692527771,
"learning_rate": 0.0002982239576772893,
"loss": 1.479320068359375,
"step": 197500
},
{
"epoch": 10.593333690011235,
"grad_norm": 0.7333496809005737,
"learning_rate": 0.00029642452269987226,
"loss": 1.48045166015625,
"step": 198000
},
{
"epoch": 10.620084532662778,
"grad_norm": 0.7259752750396729,
"learning_rate": 0.0002946250877224552,
"loss": 1.477999267578125,
"step": 198500
},
{
"epoch": 10.646835375314323,
"grad_norm": 0.7029126286506653,
"learning_rate": 0.00029282565274503806,
"loss": 1.479612060546875,
"step": 199000
},
{
"epoch": 10.673586217965866,
"grad_norm": 0.7578304409980774,
"learning_rate": 0.00029102621776762094,
"loss": 1.4713853759765625,
"step": 199500
},
{
"epoch": 10.700337060617409,
"grad_norm": 0.7665888667106628,
"learning_rate": 0.00028922678279020387,
"loss": 1.4723492431640626,
"step": 200000
},
{
"epoch": 10.727087903268952,
"grad_norm": 0.7280429005622864,
"learning_rate": 0.0002874273478127868,
"loss": 1.4720491943359375,
"step": 200500
},
{
"epoch": 10.753838745920497,
"grad_norm": 0.7449855208396912,
"learning_rate": 0.00028562791283536973,
"loss": 1.46521875,
"step": 201000
},
{
"epoch": 10.78058958857204,
"grad_norm": 0.7010794878005981,
"learning_rate": 0.0002838284778579526,
"loss": 1.4695062255859375,
"step": 201500
},
{
"epoch": 10.807340431223583,
"grad_norm": 0.714584469795227,
"learning_rate": 0.0002820290428805355,
"loss": 1.473732177734375,
"step": 202000
},
{
"epoch": 10.834091273875128,
"grad_norm": 0.767084002494812,
"learning_rate": 0.0002802296079031184,
"loss": 1.470200439453125,
"step": 202500
},
{
"epoch": 10.86084211652667,
"grad_norm": 0.7632326483726501,
"learning_rate": 0.00027843017292570135,
"loss": 1.4696666259765625,
"step": 203000
},
{
"epoch": 10.887592959178214,
"grad_norm": 0.7717725038528442,
"learning_rate": 0.0002766307379482842,
"loss": 1.46488671875,
"step": 203500
},
{
"epoch": 10.914343801829757,
"grad_norm": 0.7862579226493835,
"learning_rate": 0.00027483130297086716,
"loss": 1.46486328125,
"step": 204000
},
{
"epoch": 10.941094644481302,
"grad_norm": 0.6994292736053467,
"learning_rate": 0.0002730318679934501,
"loss": 1.4673212890625,
"step": 204500
},
{
"epoch": 10.967845487132845,
"grad_norm": 0.7454450130462646,
"learning_rate": 0.00027123243301603296,
"loss": 1.4633814697265626,
"step": 205000
},
{
"epoch": 10.994596329784388,
"grad_norm": 0.7814640998840332,
"learning_rate": 0.0002694329980386159,
"loss": 1.4637462158203125,
"step": 205500
},
{
"epoch": 11.02134717243593,
"grad_norm": 0.7605470418930054,
"learning_rate": 0.00026763356306119877,
"loss": 1.4621773681640624,
"step": 206000
},
{
"epoch": 11.048098015087476,
"grad_norm": 0.7483647465705872,
"learning_rate": 0.0002658341280837817,
"loss": 1.4608890380859374,
"step": 206500
},
{
"epoch": 11.074848857739019,
"grad_norm": 0.7337815165519714,
"learning_rate": 0.00026403469310636463,
"loss": 1.458990966796875,
"step": 207000
},
{
"epoch": 11.101599700390562,
"grad_norm": 0.7331007719039917,
"learning_rate": 0.0002622352581289475,
"loss": 1.4553223876953125,
"step": 207500
},
{
"epoch": 11.128350543042107,
"grad_norm": 0.7788465619087219,
"learning_rate": 0.00026043582315153044,
"loss": 1.4683961181640626,
"step": 208000
},
{
"epoch": 11.15510138569365,
"grad_norm": 0.7576395273208618,
"learning_rate": 0.0002586363881741133,
"loss": 1.4613536376953125,
"step": 208500
},
{
"epoch": 11.181852228345193,
"grad_norm": 0.8308989405632019,
"learning_rate": 0.00025683695319669625,
"loss": 1.4643599853515625,
"step": 209000
},
{
"epoch": 11.208603070996736,
"grad_norm": 0.814745306968689,
"learning_rate": 0.0002550375182192792,
"loss": 1.4551763916015625,
"step": 209500
},
{
"epoch": 11.23535391364828,
"grad_norm": 0.8077753186225891,
"learning_rate": 0.00025323808324186206,
"loss": 1.4577049560546875,
"step": 210000
},
{
"epoch": 11.262104756299824,
"grad_norm": 0.725962221622467,
"learning_rate": 0.00025143864826444493,
"loss": 1.459432861328125,
"step": 210500
},
{
"epoch": 11.288855598951367,
"grad_norm": 0.750404953956604,
"learning_rate": 0.00024963921328702786,
"loss": 1.4516353759765626,
"step": 211000
},
{
"epoch": 11.31560644160291,
"grad_norm": 0.7517709732055664,
"learning_rate": 0.0002478397783096108,
"loss": 1.4558858642578125,
"step": 211500
},
{
"epoch": 11.342357284254454,
"grad_norm": 0.7569293975830078,
"learning_rate": 0.00024604034333219367,
"loss": 1.4541234130859375,
"step": 212000
},
{
"epoch": 11.369108126905997,
"grad_norm": 0.7466174960136414,
"learning_rate": 0.0002442409083547766,
"loss": 1.4536090087890625,
"step": 212500
},
{
"epoch": 11.39585896955754,
"grad_norm": 0.7441415190696716,
"learning_rate": 0.0002424414733773595,
"loss": 1.4533345947265626,
"step": 213000
},
{
"epoch": 11.422609812209085,
"grad_norm": 0.7417920231819153,
"learning_rate": 0.0002406420383999424,
"loss": 1.4490595703125,
"step": 213500
},
{
"epoch": 11.449360654860628,
"grad_norm": 0.7110759615898132,
"learning_rate": 0.00023884260342252534,
"loss": 1.4479295654296875,
"step": 214000
},
{
"epoch": 11.476111497512171,
"grad_norm": 0.7999290823936462,
"learning_rate": 0.00023704316844510824,
"loss": 1.44936962890625,
"step": 214500
},
{
"epoch": 11.502862340163714,
"grad_norm": 0.7483652234077454,
"learning_rate": 0.00023524373346769115,
"loss": 1.450311279296875,
"step": 215000
},
{
"epoch": 11.52961318281526,
"grad_norm": 0.7512121200561523,
"learning_rate": 0.00023344429849027408,
"loss": 1.4471968994140625,
"step": 215500
},
{
"epoch": 11.556364025466802,
"grad_norm": 0.7558231949806213,
"learning_rate": 0.00023164486351285696,
"loss": 1.445528564453125,
"step": 216000
},
{
"epoch": 11.583114868118345,
"grad_norm": 0.7120853662490845,
"learning_rate": 0.0002298454285354399,
"loss": 1.4474659423828125,
"step": 216500
},
{
"epoch": 11.609865710769888,
"grad_norm": 0.7217755913734436,
"learning_rate": 0.00022804599355802276,
"loss": 1.4469678955078125,
"step": 217000
},
{
"epoch": 11.636616553421433,
"grad_norm": 0.6941093802452087,
"learning_rate": 0.0002262465585806057,
"loss": 1.4444993896484375,
"step": 217500
},
{
"epoch": 11.663367396072976,
"grad_norm": 0.6859148740768433,
"learning_rate": 0.00022444712360318863,
"loss": 1.4369112548828125,
"step": 218000
},
{
"epoch": 11.69011823872452,
"grad_norm": 0.7878859639167786,
"learning_rate": 0.0002226476886257715,
"loss": 1.4470831298828124,
"step": 218500
},
{
"epoch": 11.716869081376064,
"grad_norm": 0.760720431804657,
"learning_rate": 0.00022084825364835443,
"loss": 1.440456787109375,
"step": 219000
},
{
"epoch": 11.743619924027607,
"grad_norm": 0.7022302150726318,
"learning_rate": 0.0002190488186709373,
"loss": 1.4445809326171875,
"step": 219500
},
{
"epoch": 11.77037076667915,
"grad_norm": 0.7376380562782288,
"learning_rate": 0.00021724938369352024,
"loss": 1.4409373779296875,
"step": 220000
},
{
"epoch": 11.797121609330693,
"grad_norm": 0.7675197720527649,
"learning_rate": 0.00021544994871610314,
"loss": 1.441703857421875,
"step": 220500
},
{
"epoch": 11.823872451982238,
"grad_norm": 0.7365001440048218,
"learning_rate": 0.00021365051373868605,
"loss": 1.434283935546875,
"step": 221000
},
{
"epoch": 11.850623294633781,
"grad_norm": 0.7452089190483093,
"learning_rate": 0.00021185107876126898,
"loss": 1.438060302734375,
"step": 221500
},
{
"epoch": 11.877374137285324,
"grad_norm": 0.7879006266593933,
"learning_rate": 0.00021005164378385188,
"loss": 1.4402716064453125,
"step": 222000
},
{
"epoch": 11.904124979936867,
"grad_norm": 0.7459799647331238,
"learning_rate": 0.0002082522088064348,
"loss": 1.4389625244140625,
"step": 222500
},
{
"epoch": 11.930875822588412,
"grad_norm": 0.7374847531318665,
"learning_rate": 0.0002064527738290177,
"loss": 1.435837158203125,
"step": 223000
},
{
"epoch": 11.957626665239955,
"grad_norm": 0.7564308047294617,
"learning_rate": 0.0002046533388516006,
"loss": 1.432250732421875,
"step": 223500
},
{
"epoch": 11.984377507891498,
"grad_norm": 0.7205511331558228,
"learning_rate": 0.0002028539038741835,
"loss": 1.4373668212890625,
"step": 224000
},
{
"epoch": 12.011128350543043,
"grad_norm": 0.8186995983123779,
"learning_rate": 0.00020105446889676643,
"loss": 1.441746826171875,
"step": 224500
},
{
"epoch": 12.037879193194586,
"grad_norm": 0.729551374912262,
"learning_rate": 0.00019925503391934933,
"loss": 1.4410009765625,
"step": 225000
},
{
"epoch": 12.064630035846129,
"grad_norm": 0.7062949538230896,
"learning_rate": 0.00019745559894193224,
"loss": 1.436554443359375,
"step": 225500
},
{
"epoch": 12.091380878497672,
"grad_norm": 0.7159262895584106,
"learning_rate": 0.00019565616396451514,
"loss": 1.4290533447265625,
"step": 226000
},
{
"epoch": 12.118131721149217,
"grad_norm": 0.8138640522956848,
"learning_rate": 0.00019385672898709804,
"loss": 1.4325101318359375,
"step": 226500
},
{
"epoch": 12.14488256380076,
"grad_norm": 0.7465634346008301,
"learning_rate": 0.00019205729400968098,
"loss": 1.4300877685546876,
"step": 227000
},
{
"epoch": 12.171633406452303,
"grad_norm": 0.7509774565696716,
"learning_rate": 0.00019025785903226385,
"loss": 1.4303016357421876,
"step": 227500
},
{
"epoch": 12.198384249103846,
"grad_norm": 0.7163280844688416,
"learning_rate": 0.00018845842405484678,
"loss": 1.425113525390625,
"step": 228000
},
{
"epoch": 12.22513509175539,
"grad_norm": 0.8012059926986694,
"learning_rate": 0.00018665898907742971,
"loss": 1.4264906005859375,
"step": 228500
},
{
"epoch": 12.251885934406934,
"grad_norm": 0.7673007249832153,
"learning_rate": 0.0001848595541000126,
"loss": 1.4268021240234374,
"step": 229000
},
{
"epoch": 12.278636777058477,
"grad_norm": 0.7365991473197937,
"learning_rate": 0.00018306011912259552,
"loss": 1.421816162109375,
"step": 229500
},
{
"epoch": 12.305387619710022,
"grad_norm": 0.723831057548523,
"learning_rate": 0.0001812606841451784,
"loss": 1.4243193359375,
"step": 230000
},
{
"epoch": 12.332138462361565,
"grad_norm": 0.7596046924591064,
"learning_rate": 0.00017946124916776133,
"loss": 1.4229552001953125,
"step": 230500
},
{
"epoch": 12.358889305013108,
"grad_norm": 0.7320923209190369,
"learning_rate": 0.00017766181419034423,
"loss": 1.4240860595703124,
"step": 231000
},
{
"epoch": 12.38564014766465,
"grad_norm": 0.7441561222076416,
"learning_rate": 0.00017586237921292714,
"loss": 1.4169525146484374,
"step": 231500
},
{
"epoch": 12.412390990316196,
"grad_norm": 0.775891900062561,
"learning_rate": 0.00017406294423551007,
"loss": 1.42210546875,
"step": 232000
},
{
"epoch": 12.439141832967739,
"grad_norm": 0.7397918701171875,
"learning_rate": 0.00017226350925809294,
"loss": 1.426420166015625,
"step": 232500
},
{
"epoch": 12.465892675619282,
"grad_norm": 0.7503230571746826,
"learning_rate": 0.00017046407428067588,
"loss": 1.42242578125,
"step": 233000
},
{
"epoch": 12.492643518270825,
"grad_norm": 0.7026985883712769,
"learning_rate": 0.00016866463930325878,
"loss": 1.4258599853515626,
"step": 233500
},
{
"epoch": 12.51939436092237,
"grad_norm": 0.7166287302970886,
"learning_rate": 0.00016686520432584168,
"loss": 1.4213619384765626,
"step": 234000
},
{
"epoch": 12.546145203573912,
"grad_norm": 0.6964495182037354,
"learning_rate": 0.0001650657693484246,
"loss": 1.420487060546875,
"step": 234500
},
{
"epoch": 12.572896046225456,
"grad_norm": 0.7205587029457092,
"learning_rate": 0.00016326633437100752,
"loss": 1.4192598876953124,
"step": 235000
},
{
"epoch": 12.599646888877,
"grad_norm": 0.7457663416862488,
"learning_rate": 0.00016146689939359042,
"loss": 1.419793701171875,
"step": 235500
},
{
"epoch": 12.626397731528543,
"grad_norm": 0.7040967345237732,
"learning_rate": 0.00015966746441617333,
"loss": 1.422240478515625,
"step": 236000
},
{
"epoch": 12.653148574180086,
"grad_norm": 0.7471552491188049,
"learning_rate": 0.00015786802943875623,
"loss": 1.4128699951171875,
"step": 236500
},
{
"epoch": 12.67989941683163,
"grad_norm": 0.7501447200775146,
"learning_rate": 0.00015606859446133913,
"loss": 1.4160101318359375,
"step": 237000
},
{
"epoch": 12.706650259483174,
"grad_norm": 0.7668195962905884,
"learning_rate": 0.00015426915948392206,
"loss": 1.4136405029296875,
"step": 237500
},
{
"epoch": 12.733401102134717,
"grad_norm": 0.7606184482574463,
"learning_rate": 0.00015246972450650494,
"loss": 1.409126708984375,
"step": 238000
},
{
"epoch": 12.76015194478626,
"grad_norm": 0.7456269264221191,
"learning_rate": 0.00015067028952908787,
"loss": 1.4066519775390625,
"step": 238500
},
{
"epoch": 12.786902787437803,
"grad_norm": 0.724746584892273,
"learning_rate": 0.00014887085455167078,
"loss": 1.413960205078125,
"step": 239000
},
{
"epoch": 12.813653630089348,
"grad_norm": 0.7239637970924377,
"learning_rate": 0.00014707141957425368,
"loss": 1.4140576171875,
"step": 239500
},
{
"epoch": 12.840404472740891,
"grad_norm": 0.7354808449745178,
"learning_rate": 0.0001452719845968366,
"loss": 1.406454833984375,
"step": 240000
},
{
"epoch": 12.867155315392434,
"grad_norm": 0.792609691619873,
"learning_rate": 0.0001434725496194195,
"loss": 1.405090576171875,
"step": 240500
},
{
"epoch": 12.893906158043979,
"grad_norm": 0.7580500245094299,
"learning_rate": 0.00014167311464200242,
"loss": 1.4115911865234374,
"step": 241000
},
{
"epoch": 12.920657000695522,
"grad_norm": 0.7074113488197327,
"learning_rate": 0.00013987367966458532,
"loss": 1.4088466796875,
"step": 241500
},
{
"epoch": 12.947407843347065,
"grad_norm": 0.7599959969520569,
"learning_rate": 0.00013807424468716823,
"loss": 1.4013341064453124,
"step": 242000
},
{
"epoch": 12.974158685998608,
"grad_norm": 0.7413616180419922,
"learning_rate": 0.00013627480970975116,
"loss": 1.402994384765625,
"step": 242500
},
{
"epoch": 13.000909528650153,
"grad_norm": 0.7284383177757263,
"learning_rate": 0.00013447537473233403,
"loss": 1.406735107421875,
"step": 243000
},
{
"epoch": 13.027660371301696,
"grad_norm": 0.7268626689910889,
"learning_rate": 0.00013267593975491696,
"loss": 1.40669970703125,
"step": 243500
},
{
"epoch": 13.05441121395324,
"grad_norm": 0.7634581923484802,
"learning_rate": 0.00013087650477749987,
"loss": 1.399744140625,
"step": 244000
},
{
"epoch": 13.081162056604784,
"grad_norm": 0.7139135003089905,
"learning_rate": 0.00012907706980008277,
"loss": 1.4086285400390626,
"step": 244500
},
{
"epoch": 13.107912899256327,
"grad_norm": 0.7512830495834351,
"learning_rate": 0.0001272776348226657,
"loss": 1.400561279296875,
"step": 245000
},
{
"epoch": 13.13466374190787,
"grad_norm": 0.7313770055770874,
"learning_rate": 0.0001254781998452486,
"loss": 1.4049873046875,
"step": 245500
},
{
"epoch": 13.161414584559413,
"grad_norm": 0.723584771156311,
"learning_rate": 0.0001236787648678315,
"loss": 1.3912607421875,
"step": 246000
},
{
"epoch": 13.188165427210958,
"grad_norm": 0.7714723348617554,
"learning_rate": 0.0001218793298904144,
"loss": 1.4012103271484375,
"step": 246500
},
{
"epoch": 13.214916269862501,
"grad_norm": 0.7506632208824158,
"learning_rate": 0.00012007989491299733,
"loss": 1.400626220703125,
"step": 247000
},
{
"epoch": 13.241667112514044,
"grad_norm": 0.6929410099983215,
"learning_rate": 0.00011828045993558023,
"loss": 1.397333984375,
"step": 247500
},
{
"epoch": 13.268417955165587,
"grad_norm": 0.7190459370613098,
"learning_rate": 0.00011648102495816314,
"loss": 1.3994888916015624,
"step": 248000
},
{
"epoch": 13.295168797817132,
"grad_norm": 0.7408356666564941,
"learning_rate": 0.00011468158998074604,
"loss": 1.401008056640625,
"step": 248500
},
{
"epoch": 13.321919640468675,
"grad_norm": 0.7640665769577026,
"learning_rate": 0.00011288215500332896,
"loss": 1.3962269287109375,
"step": 249000
},
{
"epoch": 13.348670483120218,
"grad_norm": 0.7460052371025085,
"learning_rate": 0.00011108272002591186,
"loss": 1.3960794677734376,
"step": 249500
},
{
"epoch": 13.375421325771761,
"grad_norm": 0.7272036075592041,
"learning_rate": 0.00010928328504849477,
"loss": 1.393092041015625,
"step": 250000
},
{
"epoch": 13.402172168423306,
"grad_norm": 0.7600502967834473,
"learning_rate": 0.00010748385007107768,
"loss": 1.393472412109375,
"step": 250500
},
{
"epoch": 13.428923011074849,
"grad_norm": 0.7245965003967285,
"learning_rate": 0.0001056844150936606,
"loss": 1.388460205078125,
"step": 251000
},
{
"epoch": 13.455673853726392,
"grad_norm": 0.7350643873214722,
"learning_rate": 0.0001038849801162435,
"loss": 1.38944189453125,
"step": 251500
},
{
"epoch": 13.482424696377937,
"grad_norm": 0.7101415991783142,
"learning_rate": 0.00010208554513882641,
"loss": 1.3916180419921875,
"step": 252000
},
{
"epoch": 13.50917553902948,
"grad_norm": 0.7949294447898865,
"learning_rate": 0.00010028611016140931,
"loss": 1.3905107421875,
"step": 252500
},
{
"epoch": 13.535926381681023,
"grad_norm": 0.7555908560752869,
"learning_rate": 9.848667518399222e-05,
"loss": 1.396496337890625,
"step": 253000
},
{
"epoch": 13.562677224332566,
"grad_norm": 0.7489547729492188,
"learning_rate": 9.668724020657515e-05,
"loss": 1.3956258544921876,
"step": 253500
},
{
"epoch": 13.58942806698411,
"grad_norm": 0.73517906665802,
"learning_rate": 9.488780522915805e-05,
"loss": 1.3915584716796876,
"step": 254000
},
{
"epoch": 13.616178909635654,
"grad_norm": 0.7375670671463013,
"learning_rate": 9.308837025174096e-05,
"loss": 1.3866414794921875,
"step": 254500
},
{
"epoch": 13.642929752287197,
"grad_norm": 0.7673133611679077,
"learning_rate": 9.128893527432386e-05,
"loss": 1.3906295166015625,
"step": 255000
},
{
"epoch": 13.66968059493874,
"grad_norm": 0.7347562313079834,
"learning_rate": 8.948950029690678e-05,
"loss": 1.3891064453125,
"step": 255500
},
{
"epoch": 13.696431437590284,
"grad_norm": 0.7317259907722473,
"learning_rate": 8.769006531948968e-05,
"loss": 1.3901058349609374,
"step": 256000
},
{
"epoch": 13.723182280241828,
"grad_norm": 0.7051880955696106,
"learning_rate": 8.589063034207258e-05,
"loss": 1.378251708984375,
"step": 256500
},
{
"epoch": 13.74993312289337,
"grad_norm": 0.7472794651985168,
"learning_rate": 8.40911953646555e-05,
"loss": 1.38393896484375,
"step": 257000
},
{
"epoch": 13.776683965544915,
"grad_norm": 0.7440924644470215,
"learning_rate": 8.229176038723842e-05,
"loss": 1.3849698486328126,
"step": 257500
},
{
"epoch": 13.803434808196458,
"grad_norm": 0.7583205699920654,
"learning_rate": 8.049232540982132e-05,
"loss": 1.38602783203125,
"step": 258000
},
{
"epoch": 13.830185650848001,
"grad_norm": 0.7302603721618652,
"learning_rate": 7.869289043240423e-05,
"loss": 1.3814085693359375,
"step": 258500
},
{
"epoch": 13.856936493499544,
"grad_norm": 0.75807785987854,
"learning_rate": 7.689345545498713e-05,
"loss": 1.386816162109375,
"step": 259000
},
{
"epoch": 13.88368733615109,
"grad_norm": 0.7766947150230408,
"learning_rate": 7.509402047757003e-05,
"loss": 1.3821641845703125,
"step": 259500
},
{
"epoch": 13.910438178802632,
"grad_norm": 0.7451296448707581,
"learning_rate": 7.329458550015295e-05,
"loss": 1.3785379638671875,
"step": 260000
},
{
"epoch": 13.937189021454175,
"grad_norm": 0.7101763486862183,
"learning_rate": 7.149515052273587e-05,
"loss": 1.3773973388671874,
"step": 260500
},
{
"epoch": 13.963939864105718,
"grad_norm": 0.7285464406013489,
"learning_rate": 6.969571554531877e-05,
"loss": 1.38761376953125,
"step": 261000
},
{
"epoch": 13.990690706757263,
"grad_norm": 0.7524569034576416,
"learning_rate": 6.789628056790168e-05,
"loss": 1.3822445068359375,
"step": 261500
},
{
"epoch": 14.017441549408806,
"grad_norm": 0.677680253982544,
"learning_rate": 6.60968455904846e-05,
"loss": 1.37884130859375,
"step": 262000
},
{
"epoch": 14.04419239206035,
"grad_norm": 0.7514787912368774,
"learning_rate": 6.42974106130675e-05,
"loss": 1.376386962890625,
"step": 262500
},
{
"epoch": 14.070943234711894,
"grad_norm": 0.7569826245307922,
"learning_rate": 6.24979756356504e-05,
"loss": 1.373482177734375,
"step": 263000
},
{
"epoch": 14.097694077363437,
"grad_norm": 0.7286829352378845,
"learning_rate": 6.069854065823332e-05,
"loss": 1.374998046875,
"step": 263500
},
{
"epoch": 14.12444492001498,
"grad_norm": 0.7757689952850342,
"learning_rate": 5.889910568081622e-05,
"loss": 1.38008154296875,
"step": 264000
},
{
"epoch": 14.151195762666523,
"grad_norm": 0.7480472922325134,
"learning_rate": 5.7099670703399134e-05,
"loss": 1.3774830322265625,
"step": 264500
},
{
"epoch": 14.177946605318068,
"grad_norm": 0.7332072257995605,
"learning_rate": 5.5300235725982044e-05,
"loss": 1.372183349609375,
"step": 265000
},
{
"epoch": 14.204697447969611,
"grad_norm": 0.7424035668373108,
"learning_rate": 5.350080074856495e-05,
"loss": 1.3759425048828124,
"step": 265500
},
{
"epoch": 14.231448290621154,
"grad_norm": 0.7623999714851379,
"learning_rate": 5.170136577114786e-05,
"loss": 1.36854345703125,
"step": 266000
},
{
"epoch": 14.258199133272699,
"grad_norm": 0.7361905574798584,
"learning_rate": 4.990193079373077e-05,
"loss": 1.370366943359375,
"step": 266500
},
{
"epoch": 14.284949975924242,
"grad_norm": 0.728047251701355,
"learning_rate": 4.810249581631368e-05,
"loss": 1.3704888916015625,
"step": 267000
},
{
"epoch": 14.311700818575785,
"grad_norm": 0.7034912705421448,
"learning_rate": 4.6303060838896584e-05,
"loss": 1.3723001708984375,
"step": 267500
},
{
"epoch": 14.338451661227328,
"grad_norm": 0.7248855829238892,
"learning_rate": 4.45036258614795e-05,
"loss": 1.37246142578125,
"step": 268000
},
{
"epoch": 14.365202503878873,
"grad_norm": 0.7469500303268433,
"learning_rate": 4.2704190884062405e-05,
"loss": 1.3720052490234376,
"step": 268500
},
{
"epoch": 14.391953346530416,
"grad_norm": 0.7975779175758362,
"learning_rate": 4.0904755906645315e-05,
"loss": 1.3671141357421874,
"step": 269000
},
{
"epoch": 14.418704189181959,
"grad_norm": 0.7334534525871277,
"learning_rate": 3.910532092922822e-05,
"loss": 1.371544189453125,
"step": 269500
},
{
"epoch": 14.445455031833502,
"grad_norm": 0.7450364828109741,
"learning_rate": 3.730588595181114e-05,
"loss": 1.3636947021484376,
"step": 270000
},
{
"epoch": 14.472205874485047,
"grad_norm": 0.7644858956336975,
"learning_rate": 3.550645097439404e-05,
"loss": 1.375625,
"step": 270500
},
{
"epoch": 14.49895671713659,
"grad_norm": 0.7200730443000793,
"learning_rate": 3.370701599697695e-05,
"loss": 1.3663231201171875,
"step": 271000
},
{
"epoch": 14.525707559788133,
"grad_norm": 0.8009911179542542,
"learning_rate": 3.190758101955986e-05,
"loss": 1.365982421875,
"step": 271500
},
{
"epoch": 14.552458402439676,
"grad_norm": 0.7420063018798828,
"learning_rate": 3.010814604214277e-05,
"loss": 1.3696416015625,
"step": 272000
},
{
"epoch": 14.57920924509122,
"grad_norm": 0.7348141074180603,
"learning_rate": 2.8308711064725676e-05,
"loss": 1.3625469970703126,
"step": 272500
},
{
"epoch": 14.605960087742764,
"grad_norm": 0.7384756207466125,
"learning_rate": 2.6509276087308587e-05,
"loss": 1.3638990478515625,
"step": 273000
},
{
"epoch": 14.632710930394307,
"grad_norm": 0.7639606595039368,
"learning_rate": 2.4709841109891494e-05,
"loss": 1.3587342529296875,
"step": 273500
},
{
"epoch": 14.659461773045852,
"grad_norm": 0.7207199931144714,
"learning_rate": 2.2910406132474404e-05,
"loss": 1.3648721923828124,
"step": 274000
},
{
"epoch": 14.686212615697395,
"grad_norm": 0.8077262043952942,
"learning_rate": 2.1110971155057315e-05,
"loss": 1.3647855224609375,
"step": 274500
},
{
"epoch": 14.712963458348938,
"grad_norm": 0.7645131945610046,
"learning_rate": 1.9311536177640222e-05,
"loss": 1.3604630126953126,
"step": 275000
},
{
"epoch": 14.73971430100048,
"grad_norm": 0.784151554107666,
"learning_rate": 1.751210120022313e-05,
"loss": 1.36273095703125,
"step": 275500
},
{
"epoch": 14.766465143652026,
"grad_norm": 0.7349718809127808,
"learning_rate": 1.5712666222806037e-05,
"loss": 1.36095654296875,
"step": 276000
},
{
"epoch": 14.793215986303569,
"grad_norm": 0.7056319713592529,
"learning_rate": 1.3913231245388949e-05,
"loss": 1.365312744140625,
"step": 276500
},
{
"epoch": 14.819966828955112,
"grad_norm": 0.7208203673362732,
"learning_rate": 1.2113796267971856e-05,
"loss": 1.362748046875,
"step": 277000
},
{
"epoch": 14.846717671606655,
"grad_norm": 0.7475512027740479,
"learning_rate": 1.0314361290554765e-05,
"loss": 1.3601500244140625,
"step": 277500
},
{
"epoch": 14.8734685142582,
"grad_norm": 0.68389892578125,
"learning_rate": 8.514926313137676e-06,
"loss": 1.361802490234375,
"step": 278000
},
{
"epoch": 14.900219356909743,
"grad_norm": 0.7252270579338074,
"learning_rate": 6.7154913357205836e-06,
"loss": 1.359617431640625,
"step": 278500
},
{
"epoch": 14.926970199561286,
"grad_norm": 0.7276490926742554,
"learning_rate": 4.916056358303493e-06,
"loss": 1.3629700927734374,
"step": 279000
},
{
"epoch": 14.95372104221283,
"grad_norm": 0.7710512280464172,
"learning_rate": 3.1166213808864018e-06,
"loss": 1.356470947265625,
"step": 279500
},
{
"epoch": 14.980471884864373,
"grad_norm": 0.6686839461326599,
"learning_rate": 1.3171864034693107e-06,
"loss": 1.3584473876953125,
"step": 280000
},
{
"epoch": 15.0,
"step": 280365,
"total_flos": 5.967779092275917e+17,
"train_loss": 1.679865282512584,
"train_runtime": 45612.3692,
"train_samples_per_second": 393.381,
"train_steps_per_second": 6.147
}
],
"logging_steps": 500,
"max_steps": 280365,
"num_input_tokens_seen": 0,
"num_train_epochs": 15,
"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": 5.967779092275917e+17,
"train_batch_size": 64,
"trial_name": null,
"trial_params": null
}