{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.0, "eval_steps": 500.0, "global_step": 26534, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.003768749528906309, "grad_norm": 0.2906060218811035, "learning_rate": 9.999912386093533e-05, "loss": 0.5663, "step": 50 }, { "epoch": 0.007537499057812618, "grad_norm": 0.34499412775039673, "learning_rate": 9.999649547444612e-05, "loss": 0.5446, "step": 100 }, { "epoch": 0.011306248586718927, "grad_norm": 0.3074498772621155, "learning_rate": 9.999211493264562e-05, "loss": 0.5425, "step": 150 }, { "epoch": 0.015074998115625236, "grad_norm": 0.29462873935699463, "learning_rate": 9.998598238905239e-05, "loss": 0.5447, "step": 200 }, { "epoch": 0.018843747644531544, "grad_norm": 0.28769969940185547, "learning_rate": 9.99780980585849e-05, "loss": 0.5443, "step": 250 }, { "epoch": 0.022612497173437853, "grad_norm": 0.2771708369255066, "learning_rate": 9.996846221755392e-05, "loss": 0.543, "step": 300 }, { "epoch": 0.026381246702344163, "grad_norm": 0.27388525009155273, "learning_rate": 9.995707520365293e-05, "loss": 0.5437, "step": 350 }, { "epoch": 0.030149996231250472, "grad_norm": 0.29543155431747437, "learning_rate": 9.994393741594623e-05, "loss": 0.5428, "step": 400 }, { "epoch": 0.03391874576015678, "grad_norm": 0.2794398069381714, "learning_rate": 9.9929049314855e-05, "loss": 0.5404, "step": 450 }, { "epoch": 0.03768749528906309, "grad_norm": 0.2607417702674866, "learning_rate": 9.99124114221411e-05, "loss": 0.5451, "step": 500 }, { "epoch": 0.0414562448179694, "grad_norm": 0.24695731699466705, "learning_rate": 9.989402432088884e-05, "loss": 0.5435, "step": 550 }, { "epoch": 0.045224994346875706, "grad_norm": 0.2468685656785965, "learning_rate": 9.987388865548454e-05, "loss": 0.54, "step": 600 }, { "epoch": 0.048993743875782016, "grad_norm": 0.25657713413238525, "learning_rate": 9.985200513159391e-05, "loss": 0.5423, "step": 650 }, { "epoch": 0.052762493404688325, "grad_norm": 0.2581389546394348, "learning_rate": 9.982837451613738e-05, "loss": 0.5389, "step": 700 }, { "epoch": 0.056531242933594635, "grad_norm": 0.25357818603515625, "learning_rate": 9.980299763726313e-05, "loss": 0.5383, "step": 750 }, { "epoch": 0.060299992462500944, "grad_norm": 0.27078768610954285, "learning_rate": 9.977587538431816e-05, "loss": 0.5385, "step": 800 }, { "epoch": 0.06406874199140725, "grad_norm": 0.23457016050815582, "learning_rate": 9.974700870781712e-05, "loss": 0.5384, "step": 850 }, { "epoch": 0.06783749152031356, "grad_norm": 0.24491585791110992, "learning_rate": 9.971639861940889e-05, "loss": 0.5362, "step": 900 }, { "epoch": 0.07160624104921987, "grad_norm": 0.24196478724479675, "learning_rate": 9.968404619184125e-05, "loss": 0.5384, "step": 950 }, { "epoch": 0.07537499057812617, "grad_norm": 0.24577759206295013, "learning_rate": 9.964995255892323e-05, "loss": 0.536, "step": 1000 }, { "epoch": 0.07914374010703248, "grad_norm": 0.23471598327159882, "learning_rate": 9.961411891548538e-05, "loss": 0.536, "step": 1050 }, { "epoch": 0.0829124896359388, "grad_norm": 0.2284562885761261, "learning_rate": 9.957654651733788e-05, "loss": 0.5347, "step": 1100 }, { "epoch": 0.0866812391648451, "grad_norm": 0.267677366733551, "learning_rate": 9.953723668122656e-05, "loss": 0.5351, "step": 1150 }, { "epoch": 0.09044998869375141, "grad_norm": 0.23843041062355042, "learning_rate": 9.949619078478677e-05, "loss": 0.5331, "step": 1200 }, { "epoch": 0.09421873822265772, "grad_norm": 0.2461230456829071, "learning_rate": 9.945341026649501e-05, "loss": 0.5344, "step": 1250 }, { "epoch": 0.09798748775156403, "grad_norm": 0.22946181893348694, "learning_rate": 9.940889662561864e-05, "loss": 0.5328, "step": 1300 }, { "epoch": 0.10175623728047034, "grad_norm": 0.22417663037776947, "learning_rate": 9.93626514221632e-05, "loss": 0.5309, "step": 1350 }, { "epoch": 0.10552498680937665, "grad_norm": 0.24278171360492706, "learning_rate": 9.931467627681792e-05, "loss": 0.5317, "step": 1400 }, { "epoch": 0.10929373633828296, "grad_norm": 0.25257062911987305, "learning_rate": 9.926497287089871e-05, "loss": 0.5313, "step": 1450 }, { "epoch": 0.11306248586718927, "grad_norm": 0.2244357168674469, "learning_rate": 9.921354294628944e-05, "loss": 0.5318, "step": 1500 }, { "epoch": 0.11683123539609558, "grad_norm": 0.22987781465053558, "learning_rate": 9.916038830538071e-05, "loss": 0.53, "step": 1550 }, { "epoch": 0.12059998492500189, "grad_norm": 0.22749993205070496, "learning_rate": 9.910551081100684e-05, "loss": 0.5288, "step": 1600 }, { "epoch": 0.1243687344539082, "grad_norm": 0.23521047830581665, "learning_rate": 9.904891238638048e-05, "loss": 0.5286, "step": 1650 }, { "epoch": 0.1281374839828145, "grad_norm": 0.23071640729904175, "learning_rate": 9.899059501502526e-05, "loss": 0.5266, "step": 1700 }, { "epoch": 0.13190623351172082, "grad_norm": 0.21863527595996857, "learning_rate": 9.893056074070629e-05, "loss": 0.5271, "step": 1750 }, { "epoch": 0.1356749830406271, "grad_norm": 0.21325771510601044, "learning_rate": 9.886881166735846e-05, "loss": 0.5236, "step": 1800 }, { "epoch": 0.13944373256953344, "grad_norm": 0.2273271232843399, "learning_rate": 9.88053499590128e-05, "loss": 0.5246, "step": 1850 }, { "epoch": 0.14321248209843973, "grad_norm": 0.22810806334018707, "learning_rate": 9.874017783972058e-05, "loss": 0.5257, "step": 1900 }, { "epoch": 0.14698123162734605, "grad_norm": 0.21901291608810425, "learning_rate": 9.867329759347538e-05, "loss": 0.5245, "step": 1950 }, { "epoch": 0.15074998115625235, "grad_norm": 0.2476150393486023, "learning_rate": 9.860471156413309e-05, "loss": 0.5247, "step": 2000 }, { "epoch": 0.15451873068515867, "grad_norm": 0.22044631838798523, "learning_rate": 9.853442215532964e-05, "loss": 0.5259, "step": 2050 }, { "epoch": 0.15828748021406497, "grad_norm": 0.20892663300037384, "learning_rate": 9.846243183039694e-05, "loss": 0.5239, "step": 2100 }, { "epoch": 0.1620562297429713, "grad_norm": 0.2248477190732956, "learning_rate": 9.83887431122764e-05, "loss": 0.5215, "step": 2150 }, { "epoch": 0.1658249792718776, "grad_norm": 0.21989433467388153, "learning_rate": 9.831335858343064e-05, "loss": 0.5235, "step": 2200 }, { "epoch": 0.1695937288007839, "grad_norm": 0.2244018316268921, "learning_rate": 9.823628088575285e-05, "loss": 0.5226, "step": 2250 }, { "epoch": 0.1733624783296902, "grad_norm": 0.21466925740242004, "learning_rate": 9.815751272047434e-05, "loss": 0.5215, "step": 2300 }, { "epoch": 0.17713122785859653, "grad_norm": 0.21101678907871246, "learning_rate": 9.807705684806975e-05, "loss": 0.5207, "step": 2350 }, { "epoch": 0.18089997738750282, "grad_norm": 0.21319390833377838, "learning_rate": 9.79949160881604e-05, "loss": 0.5198, "step": 2400 }, { "epoch": 0.18466872691640915, "grad_norm": 0.2198377102613449, "learning_rate": 9.791109331941543e-05, "loss": 0.5233, "step": 2450 }, { "epoch": 0.18843747644531544, "grad_norm": 0.33185112476348877, "learning_rate": 9.782559147945094e-05, "loss": 0.5203, "step": 2500 }, { "epoch": 0.19220622597422174, "grad_norm": 0.22263136506080627, "learning_rate": 9.773841356472701e-05, "loss": 0.5193, "step": 2550 }, { "epoch": 0.19597497550312806, "grad_norm": 0.23071548342704773, "learning_rate": 9.76495626304427e-05, "loss": 0.5184, "step": 2600 }, { "epoch": 0.19974372503203436, "grad_norm": 0.21315498650074005, "learning_rate": 9.755904179042898e-05, "loss": 0.5175, "step": 2650 }, { "epoch": 0.20351247456094068, "grad_norm": 0.2224321961402893, "learning_rate": 9.746685421703961e-05, "loss": 0.5185, "step": 2700 }, { "epoch": 0.20728122408984698, "grad_norm": 0.20440195500850677, "learning_rate": 9.737300314104e-05, "loss": 0.5179, "step": 2750 }, { "epoch": 0.2110499736187533, "grad_norm": 0.220822274684906, "learning_rate": 9.727749185149388e-05, "loss": 0.5159, "step": 2800 }, { "epoch": 0.2148187231476596, "grad_norm": 0.2104455679655075, "learning_rate": 9.718032369564814e-05, "loss": 0.515, "step": 2850 }, { "epoch": 0.21858747267656592, "grad_norm": 0.21458739042282104, "learning_rate": 9.708150207881543e-05, "loss": 0.5146, "step": 2900 }, { "epoch": 0.22235622220547221, "grad_norm": 0.21516002714633942, "learning_rate": 9.698103046425496e-05, "loss": 0.5162, "step": 2950 }, { "epoch": 0.22612497173437854, "grad_norm": 0.22671422362327576, "learning_rate": 9.687891237305096e-05, "loss": 0.5132, "step": 3000 }, { "epoch": 0.22989372126328483, "grad_norm": 0.20153091847896576, "learning_rate": 9.677515138398938e-05, "loss": 0.5151, "step": 3050 }, { "epoch": 0.23366247079219116, "grad_norm": 0.21473319828510284, "learning_rate": 9.666975113343246e-05, "loss": 0.5113, "step": 3100 }, { "epoch": 0.23743122032109745, "grad_norm": 0.22579412162303925, "learning_rate": 9.65627153151913e-05, "loss": 0.5151, "step": 3150 }, { "epoch": 0.24119996985000378, "grad_norm": 0.21873536705970764, "learning_rate": 9.645404768039633e-05, "loss": 0.5135, "step": 3200 }, { "epoch": 0.24496871937891007, "grad_norm": 0.210534006357193, "learning_rate": 9.634375203736597e-05, "loss": 0.5125, "step": 3250 }, { "epoch": 0.2487374689078164, "grad_norm": 0.2117706835269928, "learning_rate": 9.623183225147308e-05, "loss": 0.5138, "step": 3300 }, { "epoch": 0.2525062184367227, "grad_norm": 0.20858237147331238, "learning_rate": 9.611829224500953e-05, "loss": 0.5107, "step": 3350 }, { "epoch": 0.256274967965629, "grad_norm": 0.1984904557466507, "learning_rate": 9.600313599704869e-05, "loss": 0.5086, "step": 3400 }, { "epoch": 0.2600437174945353, "grad_norm": 0.2197868824005127, "learning_rate": 9.588636754330609e-05, "loss": 0.511, "step": 3450 }, { "epoch": 0.26381246702344163, "grad_norm": 0.21637748181819916, "learning_rate": 9.576799097599786e-05, "loss": 0.5113, "step": 3500 }, { "epoch": 0.26758121655234796, "grad_norm": 0.2107384353876114, "learning_rate": 9.56480104436974e-05, "loss": 0.5133, "step": 3550 }, { "epoch": 0.2713499660812542, "grad_norm": 0.21115067601203918, "learning_rate": 9.552643015118998e-05, "loss": 0.5102, "step": 3600 }, { "epoch": 0.27511871561016055, "grad_norm": 0.2121039628982544, "learning_rate": 9.540325435932536e-05, "loss": 0.513, "step": 3650 }, { "epoch": 0.27888746513906687, "grad_norm": 0.20221100747585297, "learning_rate": 9.527848738486842e-05, "loss": 0.5073, "step": 3700 }, { "epoch": 0.2826562146679732, "grad_norm": 0.22263558208942413, "learning_rate": 9.5152133600348e-05, "loss": 0.509, "step": 3750 }, { "epoch": 0.28642496419687946, "grad_norm": 0.1968437135219574, "learning_rate": 9.502419743390357e-05, "loss": 0.5089, "step": 3800 }, { "epoch": 0.2901937137257858, "grad_norm": 0.21389682590961456, "learning_rate": 9.489468336913003e-05, "loss": 0.5095, "step": 3850 }, { "epoch": 0.2939624632546921, "grad_norm": 0.20469701290130615, "learning_rate": 9.476359594492068e-05, "loss": 0.5099, "step": 3900 }, { "epoch": 0.2977312127835984, "grad_norm": 0.20063601434230804, "learning_rate": 9.463093975530803e-05, "loss": 0.5054, "step": 3950 }, { "epoch": 0.3014999623125047, "grad_norm": 0.20724664628505707, "learning_rate": 9.449671944930288e-05, "loss": 0.507, "step": 4000 }, { "epoch": 0.305268711841411, "grad_norm": 0.19884862005710602, "learning_rate": 9.436093973073135e-05, "loss": 0.5063, "step": 4050 }, { "epoch": 0.30903746137031735, "grad_norm": 0.21703405678272247, "learning_rate": 9.422360535807009e-05, "loss": 0.5083, "step": 4100 }, { "epoch": 0.3128062108992236, "grad_norm": 0.20601245760917664, "learning_rate": 9.408472114427943e-05, "loss": 0.5049, "step": 4150 }, { "epoch": 0.31657496042812994, "grad_norm": 0.20570008456707, "learning_rate": 9.394429195663478e-05, "loss": 0.5083, "step": 4200 }, { "epoch": 0.32034370995703626, "grad_norm": 0.1968814581632614, "learning_rate": 9.380232271655603e-05, "loss": 0.5025, "step": 4250 }, { "epoch": 0.3241124594859426, "grad_norm": 0.20559944212436676, "learning_rate": 9.365881839943508e-05, "loss": 0.5045, "step": 4300 }, { "epoch": 0.32788120901484885, "grad_norm": 0.19153840839862823, "learning_rate": 9.351378403446144e-05, "loss": 0.5057, "step": 4350 }, { "epoch": 0.3316499585437552, "grad_norm": 0.21196983754634857, "learning_rate": 9.336722470444604e-05, "loss": 0.5052, "step": 4400 }, { "epoch": 0.3354187080726615, "grad_norm": 0.219445139169693, "learning_rate": 9.321914554564302e-05, "loss": 0.506, "step": 4450 }, { "epoch": 0.3391874576015678, "grad_norm": 0.19869840145111084, "learning_rate": 9.306955174756985e-05, "loss": 0.5032, "step": 4500 }, { "epoch": 0.3429562071304741, "grad_norm": 0.19720250368118286, "learning_rate": 9.291844855282532e-05, "loss": 0.5053, "step": 4550 }, { "epoch": 0.3467249566593804, "grad_norm": 0.1970975250005722, "learning_rate": 9.27658412569059e-05, "loss": 0.5027, "step": 4600 }, { "epoch": 0.35049370618828674, "grad_norm": 0.2193278670310974, "learning_rate": 9.261173520802014e-05, "loss": 0.5038, "step": 4650 }, { "epoch": 0.35426245571719306, "grad_norm": 0.2011794000864029, "learning_rate": 9.24561358069012e-05, "loss": 0.5039, "step": 4700 }, { "epoch": 0.3580312052460993, "grad_norm": 0.21378175914287567, "learning_rate": 9.229904850661766e-05, "loss": 0.5026, "step": 4750 }, { "epoch": 0.36179995477500565, "grad_norm": 0.20413589477539062, "learning_rate": 9.214047881238233e-05, "loss": 0.5001, "step": 4800 }, { "epoch": 0.365568704303912, "grad_norm": 0.20819853246212006, "learning_rate": 9.19804322813593e-05, "loss": 0.5011, "step": 4850 }, { "epoch": 0.3693374538328183, "grad_norm": 0.20473329722881317, "learning_rate": 9.181891452246937e-05, "loss": 0.5018, "step": 4900 }, { "epoch": 0.37310620336172456, "grad_norm": 0.1926797479391098, "learning_rate": 9.165593119619321e-05, "loss": 0.5021, "step": 4950 }, { "epoch": 0.3768749528906309, "grad_norm": 0.19186194241046906, "learning_rate": 9.149148801437321e-05, "loss": 0.499, "step": 5000 }, { "epoch": 0.3806437024195372, "grad_norm": 0.19856803119182587, "learning_rate": 9.132559074001319e-05, "loss": 0.5019, "step": 5050 }, { "epoch": 0.3844124519484435, "grad_norm": 0.20318450033664703, "learning_rate": 9.115824518707644e-05, "loss": 0.5023, "step": 5100 }, { "epoch": 0.3881812014773498, "grad_norm": 0.19259172677993774, "learning_rate": 9.098945722028203e-05, "loss": 0.4994, "step": 5150 }, { "epoch": 0.3919499510062561, "grad_norm": 0.19474363327026367, "learning_rate": 9.08192327548992e-05, "loss": 0.4992, "step": 5200 }, { "epoch": 0.39571870053516245, "grad_norm": 0.1978476643562317, "learning_rate": 9.064757775654011e-05, "loss": 0.4975, "step": 5250 }, { "epoch": 0.3994874500640687, "grad_norm": 0.19991905987262726, "learning_rate": 9.047449824095075e-05, "loss": 0.4939, "step": 5300 }, { "epoch": 0.40325619959297504, "grad_norm": 0.1946549117565155, "learning_rate": 9.030000027380012e-05, "loss": 0.4978, "step": 5350 }, { "epoch": 0.40702494912188136, "grad_norm": 0.18962378799915314, "learning_rate": 9.012408997046766e-05, "loss": 0.4978, "step": 5400 }, { "epoch": 0.4107936986507877, "grad_norm": 0.1853850781917572, "learning_rate": 8.994677349582888e-05, "loss": 0.4971, "step": 5450 }, { "epoch": 0.41456244817969395, "grad_norm": 0.19636128842830658, "learning_rate": 8.976805706403942e-05, "loss": 0.4979, "step": 5500 }, { "epoch": 0.4183311977086003, "grad_norm": 0.19317063689231873, "learning_rate": 8.958794693831716e-05, "loss": 0.4963, "step": 5550 }, { "epoch": 0.4220999472375066, "grad_norm": 0.1992931365966797, "learning_rate": 8.94064494307228e-05, "loss": 0.4947, "step": 5600 }, { "epoch": 0.4258686967664129, "grad_norm": 0.19183537364006042, "learning_rate": 8.922357090193858e-05, "loss": 0.4964, "step": 5650 }, { "epoch": 0.4296374462953192, "grad_norm": 0.20117546617984772, "learning_rate": 8.903931776104545e-05, "loss": 0.4949, "step": 5700 }, { "epoch": 0.4334061958242255, "grad_norm": 0.19739404320716858, "learning_rate": 8.885369646529838e-05, "loss": 0.4965, "step": 5750 }, { "epoch": 0.43717494535313184, "grad_norm": 0.2107592076063156, "learning_rate": 8.866671351990007e-05, "loss": 0.4957, "step": 5800 }, { "epoch": 0.44094369488203816, "grad_norm": 0.20069849491119385, "learning_rate": 8.847837547777312e-05, "loss": 0.4967, "step": 5850 }, { "epoch": 0.44471244441094443, "grad_norm": 0.19208498299121857, "learning_rate": 8.82886889393301e-05, "loss": 0.4934, "step": 5900 }, { "epoch": 0.44848119393985075, "grad_norm": 0.1858828216791153, "learning_rate": 8.80976605522425e-05, "loss": 0.4963, "step": 5950 }, { "epoch": 0.4522499434687571, "grad_norm": 0.1936417669057846, "learning_rate": 8.790529701120759e-05, "loss": 0.4943, "step": 6000 }, { "epoch": 0.4560186929976634, "grad_norm": 0.20013850927352905, "learning_rate": 8.771160505771391e-05, "loss": 0.4928, "step": 6050 }, { "epoch": 0.45978744252656967, "grad_norm": 0.1874343603849411, "learning_rate": 8.751659147980493e-05, "loss": 0.4929, "step": 6100 }, { "epoch": 0.463556192055476, "grad_norm": 0.1926513910293579, "learning_rate": 8.732026311184119e-05, "loss": 0.4928, "step": 6150 }, { "epoch": 0.4673249415843823, "grad_norm": 0.19659832119941711, "learning_rate": 8.712262683426082e-05, "loss": 0.4924, "step": 6200 }, { "epoch": 0.47109369111328864, "grad_norm": 0.1873038113117218, "learning_rate": 8.692368957333834e-05, "loss": 0.4888, "step": 6250 }, { "epoch": 0.4748624406421949, "grad_norm": 0.1897285133600235, "learning_rate": 8.672345830094199e-05, "loss": 0.4946, "step": 6300 }, { "epoch": 0.47863119017110123, "grad_norm": 0.18970079720020294, "learning_rate": 8.652194003428932e-05, "loss": 0.4909, "step": 6350 }, { "epoch": 0.48239993970000755, "grad_norm": 0.18571080267429352, "learning_rate": 8.631914183570143e-05, "loss": 0.4917, "step": 6400 }, { "epoch": 0.4861686892289138, "grad_norm": 0.19028496742248535, "learning_rate": 8.611507081235522e-05, "loss": 0.4927, "step": 6450 }, { "epoch": 0.48993743875782014, "grad_norm": 0.1985916942358017, "learning_rate": 8.590973411603452e-05, "loss": 0.4921, "step": 6500 }, { "epoch": 0.49370618828672647, "grad_norm": 0.19005028903484344, "learning_rate": 8.570313894287938e-05, "loss": 0.4897, "step": 6550 }, { "epoch": 0.4974749378156328, "grad_norm": 0.18022720515727997, "learning_rate": 8.549529253313386e-05, "loss": 0.4903, "step": 6600 }, { "epoch": 0.5012436873445391, "grad_norm": 0.20154112577438354, "learning_rate": 8.528620217089236e-05, "loss": 0.4907, "step": 6650 }, { "epoch": 0.5050124368734454, "grad_norm": 0.1906825751066208, "learning_rate": 8.507587518384421e-05, "loss": 0.4895, "step": 6700 }, { "epoch": 0.5087811864023517, "grad_norm": 0.2176685482263565, "learning_rate": 8.486431894301704e-05, "loss": 0.4902, "step": 6750 }, { "epoch": 0.512549935931258, "grad_norm": 0.19985392689704895, "learning_rate": 8.465154086251828e-05, "loss": 0.4892, "step": 6800 }, { "epoch": 0.5163186854601644, "grad_norm": 0.2009229063987732, "learning_rate": 8.443754839927549e-05, "loss": 0.4876, "step": 6850 }, { "epoch": 0.5200874349890706, "grad_norm": 0.18816043436527252, "learning_rate": 8.422234905277495e-05, "loss": 0.4878, "step": 6900 }, { "epoch": 0.5238561845179769, "grad_norm": 0.1980554759502411, "learning_rate": 8.400595036479881e-05, "loss": 0.4876, "step": 6950 }, { "epoch": 0.5276249340468833, "grad_norm": 0.19587980210781097, "learning_rate": 8.378835991916083e-05, "loss": 0.4883, "step": 7000 }, { "epoch": 0.5313936835757895, "grad_norm": 0.19175052642822266, "learning_rate": 8.35695853414406e-05, "loss": 0.4879, "step": 7050 }, { "epoch": 0.5351624331046959, "grad_norm": 0.1843547224998474, "learning_rate": 8.334963429871627e-05, "loss": 0.4897, "step": 7100 }, { "epoch": 0.5389311826336022, "grad_norm": 0.18825556337833405, "learning_rate": 8.312851449929589e-05, "loss": 0.4884, "step": 7150 }, { "epoch": 0.5426999321625084, "grad_norm": 0.1980101317167282, "learning_rate": 8.290623369244721e-05, "loss": 0.4844, "step": 7200 }, { "epoch": 0.5464686816914148, "grad_norm": 0.2058933824300766, "learning_rate": 8.268279966812617e-05, "loss": 0.4854, "step": 7250 }, { "epoch": 0.5502374312203211, "grad_norm": 0.17871695756912231, "learning_rate": 8.245822025670384e-05, "loss": 0.4855, "step": 7300 }, { "epoch": 0.5540061807492274, "grad_norm": 0.18174917995929718, "learning_rate": 8.223250332869206e-05, "loss": 0.4861, "step": 7350 }, { "epoch": 0.5577749302781337, "grad_norm": 0.19006958603858948, "learning_rate": 8.200565679446753e-05, "loss": 0.4872, "step": 7400 }, { "epoch": 0.56154367980704, "grad_norm": 0.188937246799469, "learning_rate": 8.177768860399467e-05, "loss": 0.4832, "step": 7450 }, { "epoch": 0.5653124293359464, "grad_norm": 0.19133251905441284, "learning_rate": 8.154860674654698e-05, "loss": 0.4867, "step": 7500 }, { "epoch": 0.5690811788648527, "grad_norm": 0.18221163749694824, "learning_rate": 8.131841925042704e-05, "loss": 0.4846, "step": 7550 }, { "epoch": 0.5728499283937589, "grad_norm": 0.18946631252765656, "learning_rate": 8.108713418268514e-05, "loss": 0.4847, "step": 7600 }, { "epoch": 0.5766186779226653, "grad_norm": 0.18349218368530273, "learning_rate": 8.085475964883657e-05, "loss": 0.4839, "step": 7650 }, { "epoch": 0.5803874274515716, "grad_norm": 0.19051064550876617, "learning_rate": 8.062130379257764e-05, "loss": 0.485, "step": 7700 }, { "epoch": 0.5841561769804778, "grad_norm": 0.18272393941879272, "learning_rate": 8.038677479550016e-05, "loss": 0.4844, "step": 7750 }, { "epoch": 0.5879249265093842, "grad_norm": 0.17754477262496948, "learning_rate": 8.015118087680477e-05, "loss": 0.4838, "step": 7800 }, { "epoch": 0.5916936760382905, "grad_norm": 0.18516160547733307, "learning_rate": 7.99145302930129e-05, "loss": 0.4823, "step": 7850 }, { "epoch": 0.5954624255671968, "grad_norm": 0.18762782216072083, "learning_rate": 7.96768313376774e-05, "loss": 0.4819, "step": 7900 }, { "epoch": 0.5992311750961031, "grad_norm": 0.18815983831882477, "learning_rate": 7.943809234109185e-05, "loss": 0.4813, "step": 7950 }, { "epoch": 0.6029999246250094, "grad_norm": 0.19032306969165802, "learning_rate": 7.919832166999874e-05, "loss": 0.4798, "step": 8000 }, { "epoch": 0.6067686741539158, "grad_norm": 0.18525487184524536, "learning_rate": 7.895752772729612e-05, "loss": 0.4804, "step": 8050 }, { "epoch": 0.610537423682822, "grad_norm": 0.18313206732273102, "learning_rate": 7.871571895174316e-05, "loss": 0.4804, "step": 8100 }, { "epoch": 0.6143061732117283, "grad_norm": 0.19534072279930115, "learning_rate": 7.847290381766446e-05, "loss": 0.4807, "step": 8150 }, { "epoch": 0.6180749227406347, "grad_norm": 0.1794072687625885, "learning_rate": 7.822909083465298e-05, "loss": 0.481, "step": 8200 }, { "epoch": 0.621843672269541, "grad_norm": 0.18725860118865967, "learning_rate": 7.798428854727188e-05, "loss": 0.4803, "step": 8250 }, { "epoch": 0.6256124217984472, "grad_norm": 0.190611332654953, "learning_rate": 7.773850553475508e-05, "loss": 0.4802, "step": 8300 }, { "epoch": 0.6293811713273536, "grad_norm": 0.18728849291801453, "learning_rate": 7.749175041070647e-05, "loss": 0.4787, "step": 8350 }, { "epoch": 0.6331499208562599, "grad_norm": 0.19463524222373962, "learning_rate": 7.724403182279823e-05, "loss": 0.4791, "step": 8400 }, { "epoch": 0.6369186703851663, "grad_norm": 0.1864539533853531, "learning_rate": 7.699535845246761e-05, "loss": 0.4787, "step": 8450 }, { "epoch": 0.6406874199140725, "grad_norm": 0.18798550963401794, "learning_rate": 7.674573901461282e-05, "loss": 0.4773, "step": 8500 }, { "epoch": 0.6444561694429788, "grad_norm": 0.18946604430675507, "learning_rate": 7.649518225728745e-05, "loss": 0.4803, "step": 8550 }, { "epoch": 0.6482249189718852, "grad_norm": 0.19057536125183105, "learning_rate": 7.624369696139402e-05, "loss": 0.4785, "step": 8600 }, { "epoch": 0.6519936685007914, "grad_norm": 0.18735001981258392, "learning_rate": 7.599129194037624e-05, "loss": 0.4795, "step": 8650 }, { "epoch": 0.6557624180296977, "grad_norm": 0.19279134273529053, "learning_rate": 7.573797603991004e-05, "loss": 0.4775, "step": 8700 }, { "epoch": 0.6595311675586041, "grad_norm": 0.18735423684120178, "learning_rate": 7.548375813759366e-05, "loss": 0.478, "step": 8750 }, { "epoch": 0.6632999170875103, "grad_norm": 0.1862768828868866, "learning_rate": 7.522864714263655e-05, "loss": 0.478, "step": 8800 }, { "epoch": 0.6670686666164167, "grad_norm": 0.18440242111682892, "learning_rate": 7.497265199554699e-05, "loss": 0.4776, "step": 8850 }, { "epoch": 0.670837416145323, "grad_norm": 0.181020587682724, "learning_rate": 7.471578166781899e-05, "loss": 0.4738, "step": 8900 }, { "epoch": 0.6746061656742293, "grad_norm": 0.18536238372325897, "learning_rate": 7.445804516161768e-05, "loss": 0.4761, "step": 8950 }, { "epoch": 0.6783749152031356, "grad_norm": 0.1886385977268219, "learning_rate": 7.419945150946386e-05, "loss": 0.4773, "step": 9000 }, { "epoch": 0.6821436647320419, "grad_norm": 0.1746935397386551, "learning_rate": 7.394000977391764e-05, "loss": 0.4764, "step": 9050 }, { "epoch": 0.6859124142609482, "grad_norm": 0.18953968584537506, "learning_rate": 7.367972904726055e-05, "loss": 0.4766, "step": 9100 }, { "epoch": 0.6896811637898546, "grad_norm": 0.47858861088752747, "learning_rate": 7.341861845117708e-05, "loss": 0.4742, "step": 9150 }, { "epoch": 0.6934499133187608, "grad_norm": 0.19248232245445251, "learning_rate": 7.3156687136435e-05, "loss": 0.4753, "step": 9200 }, { "epoch": 0.6972186628476671, "grad_norm": 0.18838423490524292, "learning_rate": 7.289394428256456e-05, "loss": 0.4747, "step": 9250 }, { "epoch": 0.7009874123765735, "grad_norm": 0.18120819330215454, "learning_rate": 7.26303990975369e-05, "loss": 0.476, "step": 9300 }, { "epoch": 0.7047561619054797, "grad_norm": 0.18532796204090118, "learning_rate": 7.236606081744132e-05, "loss": 0.4766, "step": 9350 }, { "epoch": 0.7085249114343861, "grad_norm": 0.19351600110530853, "learning_rate": 7.210093870616155e-05, "loss": 0.4771, "step": 9400 }, { "epoch": 0.7122936609632924, "grad_norm": 0.18240147829055786, "learning_rate": 7.183504205505111e-05, "loss": 0.4747, "step": 9450 }, { "epoch": 0.7160624104921987, "grad_norm": 0.18519514799118042, "learning_rate": 7.156838018260776e-05, "loss": 0.4731, "step": 9500 }, { "epoch": 0.719831160021105, "grad_norm": 0.192172572016716, "learning_rate": 7.13009624341468e-05, "loss": 0.4724, "step": 9550 }, { "epoch": 0.7235999095500113, "grad_norm": 0.18007811903953552, "learning_rate": 7.103279818147371e-05, "loss": 0.4725, "step": 9600 }, { "epoch": 0.7273686590789176, "grad_norm": 0.18547901511192322, "learning_rate": 7.076389682255556e-05, "loss": 0.4739, "step": 9650 }, { "epoch": 0.731137408607824, "grad_norm": 0.18853244185447693, "learning_rate": 7.049426778119179e-05, "loss": 0.4748, "step": 9700 }, { "epoch": 0.7349061581367302, "grad_norm": 0.17960478365421295, "learning_rate": 7.022392050668381e-05, "loss": 0.4728, "step": 9750 }, { "epoch": 0.7386749076656366, "grad_norm": 0.19630731642246246, "learning_rate": 6.995286447350397e-05, "loss": 0.4713, "step": 9800 }, { "epoch": 0.7424436571945429, "grad_norm": 0.1827552616596222, "learning_rate": 6.968110918096342e-05, "loss": 0.4707, "step": 9850 }, { "epoch": 0.7462124067234491, "grad_norm": 0.18695256114006042, "learning_rate": 6.940866415287931e-05, "loss": 0.4704, "step": 9900 }, { "epoch": 0.7499811562523555, "grad_norm": 0.19157296419143677, "learning_rate": 6.91355389372409e-05, "loss": 0.4728, "step": 9950 }, { "epoch": 0.7537499057812618, "grad_norm": 0.18015547096729279, "learning_rate": 6.886174310587501e-05, "loss": 0.4732, "step": 10000 }, { "epoch": 0.757518655310168, "grad_norm": 0.19139304757118225, "learning_rate": 6.858728625411064e-05, "loss": 0.4713, "step": 10050 }, { "epoch": 0.7612874048390744, "grad_norm": 0.18222898244857788, "learning_rate": 6.831217800044252e-05, "loss": 0.4707, "step": 10100 }, { "epoch": 0.7650561543679807, "grad_norm": 0.196458101272583, "learning_rate": 6.803642798619418e-05, "loss": 0.4687, "step": 10150 }, { "epoch": 0.768824903896887, "grad_norm": 0.2024402916431427, "learning_rate": 6.776004587518001e-05, "loss": 0.4707, "step": 10200 }, { "epoch": 0.7725936534257933, "grad_norm": 0.18663807213306427, "learning_rate": 6.748304135336658e-05, "loss": 0.468, "step": 10250 }, { "epoch": 0.7763624029546996, "grad_norm": 0.18797928094863892, "learning_rate": 6.720542412853319e-05, "loss": 0.4701, "step": 10300 }, { "epoch": 0.780131152483606, "grad_norm": 0.18169479072093964, "learning_rate": 6.692720392993166e-05, "loss": 0.4666, "step": 10350 }, { "epoch": 0.7838999020125123, "grad_norm": 0.17825226485729218, "learning_rate": 6.66483905079454e-05, "loss": 0.4687, "step": 10400 }, { "epoch": 0.7876686515414185, "grad_norm": 0.18662355840206146, "learning_rate": 6.636899363374763e-05, "loss": 0.4687, "step": 10450 }, { "epoch": 0.7914374010703249, "grad_norm": 0.19263768196105957, "learning_rate": 6.608902309895895e-05, "loss": 0.4675, "step": 10500 }, { "epoch": 0.7952061505992312, "grad_norm": 0.1900280863046646, "learning_rate": 6.580848871530433e-05, "loss": 0.4694, "step": 10550 }, { "epoch": 0.7989749001281374, "grad_norm": 0.17986111342906952, "learning_rate": 6.552740031426902e-05, "loss": 0.4658, "step": 10600 }, { "epoch": 0.8027436496570438, "grad_norm": 0.1776350736618042, "learning_rate": 6.52457677467542e-05, "loss": 0.4696, "step": 10650 }, { "epoch": 0.8065123991859501, "grad_norm": 0.19290950894355774, "learning_rate": 6.496360088273161e-05, "loss": 0.4675, "step": 10700 }, { "epoch": 0.8102811487148565, "grad_norm": 0.17263682186603546, "learning_rate": 6.468090961089775e-05, "loss": 0.4672, "step": 10750 }, { "epoch": 0.8140498982437627, "grad_norm": 0.18622560799121857, "learning_rate": 6.439770383832732e-05, "loss": 0.4664, "step": 10800 }, { "epoch": 0.817818647772669, "grad_norm": 0.18446871638298035, "learning_rate": 6.41139934901259e-05, "loss": 0.4668, "step": 10850 }, { "epoch": 0.8215873973015754, "grad_norm": 0.1856282651424408, "learning_rate": 6.382978850908226e-05, "loss": 0.466, "step": 10900 }, { "epoch": 0.8253561468304816, "grad_norm": 0.17993173003196716, "learning_rate": 6.354509885531989e-05, "loss": 0.4653, "step": 10950 }, { "epoch": 0.8291248963593879, "grad_norm": 0.1821722388267517, "learning_rate": 6.325993450594782e-05, "loss": 0.4653, "step": 11000 }, { "epoch": 0.8328936458882943, "grad_norm": 0.18031588196754456, "learning_rate": 6.297430545471112e-05, "loss": 0.4651, "step": 11050 }, { "epoch": 0.8366623954172006, "grad_norm": 0.18513865768909454, "learning_rate": 6.26882217116406e-05, "loss": 0.4664, "step": 11100 }, { "epoch": 0.8404311449461069, "grad_norm": 0.18106934428215027, "learning_rate": 6.240169330270197e-05, "loss": 0.4636, "step": 11150 }, { "epoch": 0.8441998944750132, "grad_norm": 0.1788669228553772, "learning_rate": 6.211473026944452e-05, "loss": 0.4655, "step": 11200 }, { "epoch": 0.8479686440039195, "grad_norm": 0.19618336856365204, "learning_rate": 6.18273426686492e-05, "loss": 0.4663, "step": 11250 }, { "epoch": 0.8517373935328258, "grad_norm": 0.1805189996957779, "learning_rate": 6.153954057197612e-05, "loss": 0.4638, "step": 11300 }, { "epoch": 0.8555061430617321, "grad_norm": 0.17371240258216858, "learning_rate": 6.125133406561172e-05, "loss": 0.4631, "step": 11350 }, { "epoch": 0.8592748925906384, "grad_norm": 0.18736299872398376, "learning_rate": 6.0962733249915135e-05, "loss": 0.4626, "step": 11400 }, { "epoch": 0.8630436421195448, "grad_norm": 0.19724301993846893, "learning_rate": 6.067374823906431e-05, "loss": 0.462, "step": 11450 }, { "epoch": 0.866812391648451, "grad_norm": 0.18249565362930298, "learning_rate": 6.038438916070155e-05, "loss": 0.463, "step": 11500 }, { "epoch": 0.8705811411773573, "grad_norm": 0.17936116456985474, "learning_rate": 6.0094666155578525e-05, "loss": 0.4634, "step": 11550 }, { "epoch": 0.8743498907062637, "grad_norm": 0.18539419770240784, "learning_rate": 5.9804589377200946e-05, "loss": 0.4614, "step": 11600 }, { "epoch": 0.87811864023517, "grad_norm": 0.18874451518058777, "learning_rate": 5.951416899147273e-05, "loss": 0.4615, "step": 11650 }, { "epoch": 0.8818873897640763, "grad_norm": 0.19202075898647308, "learning_rate": 5.922341517633965e-05, "loss": 0.4647, "step": 11700 }, { "epoch": 0.8856561392929826, "grad_norm": 0.17386788129806519, "learning_rate": 5.893233812143274e-05, "loss": 0.4608, "step": 11750 }, { "epoch": 0.8894248888218889, "grad_norm": 0.1823577582836151, "learning_rate": 5.864094802771115e-05, "loss": 0.4609, "step": 11800 }, { "epoch": 0.8931936383507952, "grad_norm": 0.1865212470293045, "learning_rate": 5.834925510710465e-05, "loss": 0.4607, "step": 11850 }, { "epoch": 0.8969623878797015, "grad_norm": 0.18660889565944672, "learning_rate": 5.8057269582155735e-05, "loss": 0.4629, "step": 11900 }, { "epoch": 0.9007311374086078, "grad_norm": 0.17511461675167084, "learning_rate": 5.776500168566141e-05, "loss": 0.4591, "step": 11950 }, { "epoch": 0.9044998869375142, "grad_norm": 0.17838037014007568, "learning_rate": 5.7472461660314504e-05, "loss": 0.4609, "step": 12000 }, { "epoch": 0.9082686364664204, "grad_norm": 0.18015101552009583, "learning_rate": 5.717965975834483e-05, "loss": 0.4607, "step": 12050 }, { "epoch": 0.9120373859953268, "grad_norm": 0.169195756316185, "learning_rate": 5.6886606241159714e-05, "loss": 0.4582, "step": 12100 }, { "epoch": 0.9158061355242331, "grad_norm": 0.1795186996459961, "learning_rate": 5.659331137898458e-05, "loss": 0.461, "step": 12150 }, { "epoch": 0.9195748850531393, "grad_norm": 0.1830703467130661, "learning_rate": 5.6299785450502853e-05, "loss": 0.4594, "step": 12200 }, { "epoch": 0.9233436345820457, "grad_norm": 0.1865684539079666, "learning_rate": 5.600603874249585e-05, "loss": 0.4602, "step": 12250 }, { "epoch": 0.927112384110952, "grad_norm": 0.17610585689544678, "learning_rate": 5.571208154948218e-05, "loss": 0.4611, "step": 12300 }, { "epoch": 0.9308811336398582, "grad_norm": 0.18364217877388, "learning_rate": 5.541792417335707e-05, "loss": 0.4599, "step": 12350 }, { "epoch": 0.9346498831687646, "grad_norm": 0.1822074055671692, "learning_rate": 5.5123576923031253e-05, "loss": 0.4599, "step": 12400 }, { "epoch": 0.9384186326976709, "grad_norm": 0.18282079696655273, "learning_rate": 5.482905011406973e-05, "loss": 0.4568, "step": 12450 }, { "epoch": 0.9421873822265773, "grad_norm": 0.18633867800235748, "learning_rate": 5.453435406833017e-05, "loss": 0.457, "step": 12500 }, { "epoch": 0.9459561317554835, "grad_norm": 0.17714855074882507, "learning_rate": 5.4239499113601333e-05, "loss": 0.4558, "step": 12550 }, { "epoch": 0.9497248812843898, "grad_norm": 0.1829010248184204, "learning_rate": 5.3944495583240987e-05, "loss": 0.4591, "step": 12600 }, { "epoch": 0.9534936308132962, "grad_norm": 0.18278288841247559, "learning_rate": 5.364935381581377e-05, "loss": 0.4592, "step": 12650 }, { "epoch": 0.9572623803422025, "grad_norm": 0.18331314623355865, "learning_rate": 5.3354084154729034e-05, "loss": 0.4573, "step": 12700 }, { "epoch": 0.9610311298711087, "grad_norm": 0.19474105536937714, "learning_rate": 5.305869694787812e-05, "loss": 0.4557, "step": 12750 }, { "epoch": 0.9647998794000151, "grad_norm": 0.17656107246875763, "learning_rate": 5.276320254727187e-05, "loss": 0.4545, "step": 12800 }, { "epoch": 0.9685686289289214, "grad_norm": 0.17876358330249786, "learning_rate": 5.2467611308677836e-05, "loss": 0.4566, "step": 12850 }, { "epoch": 0.9723373784578276, "grad_norm": 0.1829279661178589, "learning_rate": 5.217193359125724e-05, "loss": 0.4562, "step": 12900 }, { "epoch": 0.976106127986734, "grad_norm": 0.19071832299232483, "learning_rate": 5.187617975720201e-05, "loss": 0.4555, "step": 12950 }, { "epoch": 0.9798748775156403, "grad_norm": 0.17582501471042633, "learning_rate": 5.15803601713717e-05, "loss": 0.4583, "step": 13000 }, { "epoch": 0.9836436270445467, "grad_norm": 0.18090075254440308, "learning_rate": 5.1284485200930085e-05, "loss": 0.4555, "step": 13050 }, { "epoch": 0.9874123765734529, "grad_norm": 0.18311740458011627, "learning_rate": 5.0988565214981976e-05, "loss": 0.4562, "step": 13100 }, { "epoch": 0.9911811261023592, "grad_norm": 0.18058288097381592, "learning_rate": 5.0692610584209754e-05, "loss": 0.4556, "step": 13150 }, { "epoch": 0.9949498756312656, "grad_norm": 0.18204964697360992, "learning_rate": 5.0396631680509945e-05, "loss": 0.4548, "step": 13200 }, { "epoch": 0.9987186251601718, "grad_norm": 0.18574148416519165, "learning_rate": 5.010063887662976e-05, "loss": 0.4514, "step": 13250 }, { "epoch": 1.0024873746890781, "grad_norm": 0.17709988355636597, "learning_rate": 4.9804642545803524e-05, "loss": 0.4307, "step": 13300 }, { "epoch": 1.0062561242179844, "grad_norm": 0.18011409044265747, "learning_rate": 4.950865306138917e-05, "loss": 0.4178, "step": 13350 }, { "epoch": 1.0100248737468909, "grad_norm": 0.18502439558506012, "learning_rate": 4.9212680796504704e-05, "loss": 0.4173, "step": 13400 }, { "epoch": 1.0137936232757971, "grad_norm": 0.18953478336334229, "learning_rate": 4.8916736123664666e-05, "loss": 0.4197, "step": 13450 }, { "epoch": 1.0175623728047034, "grad_norm": 0.18180860579013824, "learning_rate": 4.8620829414416615e-05, "loss": 0.4164, "step": 13500 }, { "epoch": 1.0213311223336097, "grad_norm": 0.1739131063222885, "learning_rate": 4.832497103897762e-05, "loss": 0.4162, "step": 13550 }, { "epoch": 1.025099871862516, "grad_norm": 0.18233858048915863, "learning_rate": 4.8029171365870926e-05, "loss": 0.4152, "step": 13600 }, { "epoch": 1.0288686213914224, "grad_norm": 0.18319696187973022, "learning_rate": 4.7733440761562466e-05, "loss": 0.4187, "step": 13650 }, { "epoch": 1.0326373709203287, "grad_norm": 0.1849289834499359, "learning_rate": 4.743778959009766e-05, "loss": 0.4167, "step": 13700 }, { "epoch": 1.036406120449235, "grad_norm": 0.17826583981513977, "learning_rate": 4.7142228212738126e-05, "loss": 0.4178, "step": 13750 }, { "epoch": 1.0401748699781412, "grad_norm": 0.18805526196956635, "learning_rate": 4.684676698759864e-05, "loss": 0.4169, "step": 13800 }, { "epoch": 1.0439436195070475, "grad_norm": 0.18643540143966675, "learning_rate": 4.655141626928403e-05, "loss": 0.4176, "step": 13850 }, { "epoch": 1.0477123690359538, "grad_norm": 0.18345114588737488, "learning_rate": 4.62561864085264e-05, "loss": 0.4183, "step": 13900 }, { "epoch": 1.0514811185648603, "grad_norm": 0.18768061697483063, "learning_rate": 4.596108775182228e-05, "loss": 0.4165, "step": 13950 }, { "epoch": 1.0552498680937665, "grad_norm": 0.17966210842132568, "learning_rate": 4.566613064107015e-05, "loss": 0.4171, "step": 14000 }, { "epoch": 1.0590186176226728, "grad_norm": 0.1812521517276764, "learning_rate": 4.5371325413207885e-05, "loss": 0.4191, "step": 14050 }, { "epoch": 1.062787367151579, "grad_norm": 0.1929241418838501, "learning_rate": 4.507668239985055e-05, "loss": 0.4165, "step": 14100 }, { "epoch": 1.0665561166804853, "grad_norm": 0.17854027450084686, "learning_rate": 4.478221192692831e-05, "loss": 0.416, "step": 14150 }, { "epoch": 1.0703248662093918, "grad_norm": 0.18497681617736816, "learning_rate": 4.448792431432451e-05, "loss": 0.4145, "step": 14200 }, { "epoch": 1.074093615738298, "grad_norm": 0.18206162750720978, "learning_rate": 4.4193829875514176e-05, "loss": 0.4198, "step": 14250 }, { "epoch": 1.0778623652672044, "grad_norm": 0.17210817337036133, "learning_rate": 4.389993891720232e-05, "loss": 0.4159, "step": 14300 }, { "epoch": 1.0816311147961106, "grad_norm": 0.17640775442123413, "learning_rate": 4.360626173896293e-05, "loss": 0.4178, "step": 14350 }, { "epoch": 1.085399864325017, "grad_norm": 0.19096557796001434, "learning_rate": 4.3312808632877924e-05, "loss": 0.4178, "step": 14400 }, { "epoch": 1.0891686138539232, "grad_norm": 0.18599864840507507, "learning_rate": 4.301958988317651e-05, "loss": 0.4145, "step": 14450 }, { "epoch": 1.0929373633828297, "grad_norm": 0.17436467111110687, "learning_rate": 4.27266157658747e-05, "loss": 0.4168, "step": 14500 }, { "epoch": 1.096706112911736, "grad_norm": 0.17950861155986786, "learning_rate": 4.243389654841531e-05, "loss": 0.4169, "step": 14550 }, { "epoch": 1.1004748624406422, "grad_norm": 0.18077972531318665, "learning_rate": 4.214144248930797e-05, "loss": 0.4152, "step": 14600 }, { "epoch": 1.1042436119695485, "grad_norm": 0.18444353342056274, "learning_rate": 4.184926383776971e-05, "loss": 0.4154, "step": 14650 }, { "epoch": 1.1080123614984547, "grad_norm": 0.18004246056079865, "learning_rate": 4.155737083336575e-05, "loss": 0.4155, "step": 14700 }, { "epoch": 1.1117811110273612, "grad_norm": 0.18254701793193817, "learning_rate": 4.1265773705650646e-05, "loss": 0.4168, "step": 14750 }, { "epoch": 1.1155498605562675, "grad_norm": 0.18389485776424408, "learning_rate": 4.097448267380979e-05, "loss": 0.4185, "step": 14800 }, { "epoch": 1.1193186100851737, "grad_norm": 0.17165398597717285, "learning_rate": 4.0683507946301275e-05, "loss": 0.417, "step": 14850 }, { "epoch": 1.12308735961408, "grad_norm": 0.18042343854904175, "learning_rate": 4.03928597204981e-05, "loss": 0.4152, "step": 14900 }, { "epoch": 1.1268561091429863, "grad_norm": 0.1767844706773758, "learning_rate": 4.010254818233089e-05, "loss": 0.4164, "step": 14950 }, { "epoch": 1.1306248586718928, "grad_norm": 0.18272291123867035, "learning_rate": 3.9812583505930786e-05, "loss": 0.4161, "step": 15000 }, { "epoch": 1.134393608200799, "grad_norm": 0.18290477991104126, "learning_rate": 3.952297585327303e-05, "loss": 0.4173, "step": 15050 }, { "epoch": 1.1381623577297053, "grad_norm": 0.17782056331634521, "learning_rate": 3.923373537382074e-05, "loss": 0.4143, "step": 15100 }, { "epoch": 1.1419311072586116, "grad_norm": 0.1805420070886612, "learning_rate": 3.894487220416924e-05, "loss": 0.4146, "step": 15150 }, { "epoch": 1.1456998567875178, "grad_norm": 0.18226702511310577, "learning_rate": 3.86563964676908e-05, "loss": 0.413, "step": 15200 }, { "epoch": 1.1494686063164243, "grad_norm": 0.18122869729995728, "learning_rate": 3.836831827417992e-05, "loss": 0.4153, "step": 15250 }, { "epoch": 1.1532373558453306, "grad_norm": 0.17712771892547607, "learning_rate": 3.808064771949893e-05, "loss": 0.4159, "step": 15300 }, { "epoch": 1.1570061053742369, "grad_norm": 0.17394891381263733, "learning_rate": 3.77933948852243e-05, "loss": 0.4136, "step": 15350 }, { "epoch": 1.1607748549031431, "grad_norm": 0.1861298531293869, "learning_rate": 3.75065698382932e-05, "loss": 0.4146, "step": 15400 }, { "epoch": 1.1645436044320494, "grad_norm": 0.18379145860671997, "learning_rate": 3.722018263065075e-05, "loss": 0.4148, "step": 15450 }, { "epoch": 1.1683123539609557, "grad_norm": 0.1812155693769455, "learning_rate": 3.693424329889776e-05, "loss": 0.4147, "step": 15500 }, { "epoch": 1.172081103489862, "grad_norm": 0.1795729398727417, "learning_rate": 3.664876186393897e-05, "loss": 0.4158, "step": 15550 }, { "epoch": 1.1758498530187684, "grad_norm": 0.18769286572933197, "learning_rate": 3.636374833063191e-05, "loss": 0.4154, "step": 15600 }, { "epoch": 1.1796186025476747, "grad_norm": 0.18426291644573212, "learning_rate": 3.6079212687436176e-05, "loss": 0.4134, "step": 15650 }, { "epoch": 1.183387352076581, "grad_norm": 0.18709790706634521, "learning_rate": 3.579516490606346e-05, "loss": 0.4151, "step": 15700 }, { "epoch": 1.1871561016054872, "grad_norm": 0.18941350281238556, "learning_rate": 3.551161494112807e-05, "loss": 0.413, "step": 15750 }, { "epoch": 1.1909248511343935, "grad_norm": 0.1875937581062317, "learning_rate": 3.522857272979804e-05, "loss": 0.4135, "step": 15800 }, { "epoch": 1.1946936006633, "grad_norm": 0.18288429081439972, "learning_rate": 3.494604819144688e-05, "loss": 0.4125, "step": 15850 }, { "epoch": 1.1984623501922063, "grad_norm": 0.1818443238735199, "learning_rate": 3.4664051227306026e-05, "loss": 0.4114, "step": 15900 }, { "epoch": 1.2022310997211125, "grad_norm": 0.18611133098602295, "learning_rate": 3.438259172011772e-05, "loss": 0.4143, "step": 15950 }, { "epoch": 1.2059998492500188, "grad_norm": 0.18321532011032104, "learning_rate": 3.4101679533788734e-05, "loss": 0.4134, "step": 16000 }, { "epoch": 1.209768598778925, "grad_norm": 0.1766360104084015, "learning_rate": 3.382132451304465e-05, "loss": 0.4127, "step": 16050 }, { "epoch": 1.2135373483078316, "grad_norm": 0.1815403252840042, "learning_rate": 3.354153648308492e-05, "loss": 0.4118, "step": 16100 }, { "epoch": 1.2173060978367378, "grad_norm": 0.1798935979604721, "learning_rate": 3.3262325249238466e-05, "loss": 0.4114, "step": 16150 }, { "epoch": 1.221074847365644, "grad_norm": 0.1812141090631485, "learning_rate": 3.298370059662004e-05, "loss": 0.4131, "step": 16200 }, { "epoch": 1.2248435968945504, "grad_norm": 0.1809026598930359, "learning_rate": 3.270567228978736e-05, "loss": 0.4124, "step": 16250 }, { "epoch": 1.2286123464234566, "grad_norm": 0.18148508667945862, "learning_rate": 3.2428250072398846e-05, "loss": 0.4112, "step": 16300 }, { "epoch": 1.2323810959523631, "grad_norm": 0.18835850059986115, "learning_rate": 3.2151443666872166e-05, "loss": 0.4132, "step": 16350 }, { "epoch": 1.2361498454812694, "grad_norm": 0.18036408722400665, "learning_rate": 3.187526277404355e-05, "loss": 0.4128, "step": 16400 }, { "epoch": 1.2399185950101757, "grad_norm": 0.1809609979391098, "learning_rate": 3.159971707282776e-05, "loss": 0.413, "step": 16450 }, { "epoch": 1.243687344539082, "grad_norm": 0.19554051756858826, "learning_rate": 3.1324816219878903e-05, "loss": 0.4109, "step": 16500 }, { "epoch": 1.2474560940679882, "grad_norm": 0.1792004555463791, "learning_rate": 3.1050569849252044e-05, "loss": 0.4123, "step": 16550 }, { "epoch": 1.2512248435968947, "grad_norm": 0.19579829275608063, "learning_rate": 3.077698757206552e-05, "loss": 0.41, "step": 16600 }, { "epoch": 1.2549935931258007, "grad_norm": 0.18006202578544617, "learning_rate": 3.0504078976164118e-05, "loss": 0.4114, "step": 16650 }, { "epoch": 1.2587623426547072, "grad_norm": 0.18532243371009827, "learning_rate": 3.0231853625783163e-05, "loss": 0.4117, "step": 16700 }, { "epoch": 1.2625310921836135, "grad_norm": 0.17538198828697205, "learning_rate": 2.9960321061213193e-05, "loss": 0.4108, "step": 16750 }, { "epoch": 1.2662998417125197, "grad_norm": 0.1844860017299652, "learning_rate": 2.9689490798465698e-05, "loss": 0.4095, "step": 16800 }, { "epoch": 1.270068591241426, "grad_norm": 0.18097542226314545, "learning_rate": 2.941937232893959e-05, "loss": 0.4095, "step": 16850 }, { "epoch": 1.2738373407703323, "grad_norm": 0.1822078973054886, "learning_rate": 2.9149975119088596e-05, "loss": 0.411, "step": 16900 }, { "epoch": 1.2776060902992388, "grad_norm": 0.18340744078159332, "learning_rate": 2.8881308610089496e-05, "loss": 0.4125, "step": 16950 }, { "epoch": 1.281374839828145, "grad_norm": 0.18344102799892426, "learning_rate": 2.8613382217511265e-05, "loss": 0.4116, "step": 17000 }, { "epoch": 1.2851435893570513, "grad_norm": 0.18103405833244324, "learning_rate": 2.8346205330985053e-05, "loss": 0.4099, "step": 17050 }, { "epoch": 1.2889123388859576, "grad_norm": 0.18688762187957764, "learning_rate": 2.807978731387516e-05, "loss": 0.4113, "step": 17100 }, { "epoch": 1.2926810884148638, "grad_norm": 0.17642515897750854, "learning_rate": 2.7814137502950878e-05, "loss": 0.4102, "step": 17150 }, { "epoch": 1.2964498379437703, "grad_norm": 0.1789291650056839, "learning_rate": 2.754926520805925e-05, "loss": 0.409, "step": 17200 }, { "epoch": 1.3002185874726766, "grad_norm": 0.1897159218788147, "learning_rate": 2.728517971179892e-05, "loss": 0.4095, "step": 17250 }, { "epoch": 1.3039873370015829, "grad_norm": 0.1783190369606018, "learning_rate": 2.702189026919465e-05, "loss": 0.4094, "step": 17300 }, { "epoch": 1.3077560865304891, "grad_norm": 0.18828634917736053, "learning_rate": 2.675940610737307e-05, "loss": 0.4068, "step": 17350 }, { "epoch": 1.3115248360593954, "grad_norm": 0.19085575640201569, "learning_rate": 2.6497736425239315e-05, "loss": 0.4098, "step": 17400 }, { "epoch": 1.315293585588302, "grad_norm": 0.1795608252286911, "learning_rate": 2.623689039315458e-05, "loss": 0.4086, "step": 17450 }, { "epoch": 1.3190623351172082, "grad_norm": 0.17982016503810883, "learning_rate": 2.597687715261484e-05, "loss": 0.4077, "step": 17500 }, { "epoch": 1.3228310846461144, "grad_norm": 0.17843079566955566, "learning_rate": 2.5717705815930392e-05, "loss": 0.4097, "step": 17550 }, { "epoch": 1.3265998341750207, "grad_norm": 0.1862371265888214, "learning_rate": 2.5459385465906517e-05, "loss": 0.4078, "step": 17600 }, { "epoch": 1.330368583703927, "grad_norm": 0.18295156955718994, "learning_rate": 2.520192515552522e-05, "loss": 0.4093, "step": 17650 }, { "epoch": 1.3341373332328335, "grad_norm": 0.19080038368701935, "learning_rate": 2.4945333907627892e-05, "loss": 0.4067, "step": 17700 }, { "epoch": 1.3379060827617397, "grad_norm": 0.17629754543304443, "learning_rate": 2.468962071459922e-05, "loss": 0.4099, "step": 17750 }, { "epoch": 1.341674832290646, "grad_norm": 0.18612775206565857, "learning_rate": 2.443479453805189e-05, "loss": 0.4084, "step": 17800 }, { "epoch": 1.3454435818195523, "grad_norm": 0.17433632910251617, "learning_rate": 2.4180864308512624e-05, "loss": 0.407, "step": 17850 }, { "epoch": 1.3492123313484585, "grad_norm": 0.17874571681022644, "learning_rate": 2.392783892510917e-05, "loss": 0.4075, "step": 17900 }, { "epoch": 1.352981080877365, "grad_norm": 0.18273356556892395, "learning_rate": 2.367572725525844e-05, "loss": 0.4063, "step": 17950 }, { "epoch": 1.356749830406271, "grad_norm": 0.17750360071659088, "learning_rate": 2.3424538134355715e-05, "loss": 0.4075, "step": 18000 }, { "epoch": 1.3605185799351776, "grad_norm": 0.17407415807247162, "learning_rate": 2.3174280365465102e-05, "loss": 0.4062, "step": 18050 }, { "epoch": 1.3642873294640838, "grad_norm": 0.19176089763641357, "learning_rate": 2.2924962719010874e-05, "loss": 0.4068, "step": 18100 }, { "epoch": 1.36805607899299, "grad_norm": 0.18244892358779907, "learning_rate": 2.2676593932470213e-05, "loss": 0.4072, "step": 18150 }, { "epoch": 1.3718248285218964, "grad_norm": 0.18260513246059418, "learning_rate": 2.242918271006698e-05, "loss": 0.4079, "step": 18200 }, { "epoch": 1.3755935780508026, "grad_norm": 0.18689802289009094, "learning_rate": 2.2182737722466633e-05, "loss": 0.4045, "step": 18250 }, { "epoch": 1.3793623275797091, "grad_norm": 0.1868721842765808, "learning_rate": 2.193726760647245e-05, "loss": 0.4061, "step": 18300 }, { "epoch": 1.3831310771086154, "grad_norm": 0.18145039677619934, "learning_rate": 2.169278096472272e-05, "loss": 0.4062, "step": 18350 }, { "epoch": 1.3868998266375216, "grad_norm": 0.18034633994102478, "learning_rate": 2.1449286365389342e-05, "loss": 0.4051, "step": 18400 }, { "epoch": 1.390668576166428, "grad_norm": 0.1792653501033783, "learning_rate": 2.120679234187755e-05, "loss": 0.406, "step": 18450 }, { "epoch": 1.3944373256953342, "grad_norm": 0.17878781259059906, "learning_rate": 2.0965307392526818e-05, "loss": 0.4075, "step": 18500 }, { "epoch": 1.3982060752242407, "grad_norm": 0.18702946603298187, "learning_rate": 2.072483998031303e-05, "loss": 0.4067, "step": 18550 }, { "epoch": 1.401974824753147, "grad_norm": 0.1810487061738968, "learning_rate": 2.048539853255197e-05, "loss": 0.4045, "step": 18600 }, { "epoch": 1.4057435742820532, "grad_norm": 0.17860865592956543, "learning_rate": 2.0246991440603862e-05, "loss": 0.405, "step": 18650 }, { "epoch": 1.4095123238109595, "grad_norm": 0.1859513223171234, "learning_rate": 2.0009627059579372e-05, "loss": 0.4058, "step": 18700 }, { "epoch": 1.4132810733398657, "grad_norm": 0.18461273610591888, "learning_rate": 1.9773313708046758e-05, "loss": 0.4054, "step": 18750 }, { "epoch": 1.4170498228687722, "grad_norm": 0.18574093282222748, "learning_rate": 1.953805966774037e-05, "loss": 0.406, "step": 18800 }, { "epoch": 1.4208185723976785, "grad_norm": 0.1719568520784378, "learning_rate": 1.930387318327043e-05, "loss": 0.4056, "step": 18850 }, { "epoch": 1.4245873219265848, "grad_norm": 0.18191532790660858, "learning_rate": 1.9070762461834018e-05, "loss": 0.4045, "step": 18900 }, { "epoch": 1.428356071455491, "grad_norm": 0.1849079728126526, "learning_rate": 1.8838735672927516e-05, "loss": 0.4057, "step": 18950 }, { "epoch": 1.4321248209843973, "grad_norm": 0.18583734333515167, "learning_rate": 1.8607800948060266e-05, "loss": 0.404, "step": 19000 }, { "epoch": 1.4358935705133038, "grad_norm": 0.18024706840515137, "learning_rate": 1.837796638046962e-05, "loss": 0.4049, "step": 19050 }, { "epoch": 1.43966232004221, "grad_norm": 0.18181918561458588, "learning_rate": 1.8149240024837315e-05, "loss": 0.4048, "step": 19100 }, { "epoch": 1.4434310695711163, "grad_norm": 0.18205614387989044, "learning_rate": 1.792162989700717e-05, "loss": 0.4048, "step": 19150 }, { "epoch": 1.4471998191000226, "grad_norm": 0.18339186906814575, "learning_rate": 1.7695143973704143e-05, "loss": 0.4065, "step": 19200 }, { "epoch": 1.4509685686289289, "grad_norm": 0.17691025137901306, "learning_rate": 1.746979019225483e-05, "loss": 0.4042, "step": 19250 }, { "epoch": 1.4547373181578354, "grad_norm": 0.18054354190826416, "learning_rate": 1.7245576450309316e-05, "loss": 0.4044, "step": 19300 }, { "epoch": 1.4585060676867414, "grad_norm": 0.17796145379543304, "learning_rate": 1.7022510605564307e-05, "loss": 0.405, "step": 19350 }, { "epoch": 1.462274817215648, "grad_norm": 0.17998959124088287, "learning_rate": 1.6800600475487826e-05, "loss": 0.4033, "step": 19400 }, { "epoch": 1.4660435667445542, "grad_norm": 0.18134063482284546, "learning_rate": 1.657985383704521e-05, "loss": 0.4033, "step": 19450 }, { "epoch": 1.4698123162734604, "grad_norm": 0.18283304572105408, "learning_rate": 1.6360278426426624e-05, "loss": 0.4037, "step": 19500 }, { "epoch": 1.4735810658023667, "grad_norm": 0.17472119629383087, "learning_rate": 1.6141881938775848e-05, "loss": 0.4045, "step": 19550 }, { "epoch": 1.477349815331273, "grad_norm": 0.18283599615097046, "learning_rate": 1.5924672027920663e-05, "loss": 0.4036, "step": 19600 }, { "epoch": 1.4811185648601795, "grad_norm": 0.18116436898708344, "learning_rate": 1.570865630610457e-05, "loss": 0.4034, "step": 19650 }, { "epoch": 1.4848873143890857, "grad_norm": 0.1848943680524826, "learning_rate": 1.5493842343720104e-05, "loss": 0.4041, "step": 19700 }, { "epoch": 1.488656063917992, "grad_norm": 0.1797831803560257, "learning_rate": 1.528023766904341e-05, "loss": 0.4035, "step": 19750 }, { "epoch": 1.4924248134468983, "grad_norm": 0.17644710838794708, "learning_rate": 1.5067849767970488e-05, "loss": 0.4055, "step": 19800 }, { "epoch": 1.4961935629758045, "grad_norm": 0.18384455144405365, "learning_rate": 1.4856686083754817e-05, "loss": 0.4029, "step": 19850 }, { "epoch": 1.499962312504711, "grad_norm": 0.18589411675930023, "learning_rate": 1.4646754016746483e-05, "loss": 0.4025, "step": 19900 }, { "epoch": 1.5037310620336173, "grad_norm": 0.1780794858932495, "learning_rate": 1.4438060924132924e-05, "loss": 0.4027, "step": 19950 }, { "epoch": 1.5074998115625236, "grad_norm": 0.17796076834201813, "learning_rate": 1.4230614119680957e-05, "loss": 0.4005, "step": 20000 }, { "epoch": 1.5112685610914298, "grad_norm": 0.18531276285648346, "learning_rate": 1.4024420873480549e-05, "loss": 0.4043, "step": 20050 }, { "epoch": 1.515037310620336, "grad_norm": 0.17742446064949036, "learning_rate": 1.3819488411690018e-05, "loss": 0.4033, "step": 20100 }, { "epoch": 1.5188060601492426, "grad_norm": 0.17753294110298157, "learning_rate": 1.361582391628276e-05, "loss": 0.4019, "step": 20150 }, { "epoch": 1.5225748096781488, "grad_norm": 0.1751794070005417, "learning_rate": 1.3413434524795631e-05, "loss": 0.4012, "step": 20200 }, { "epoch": 1.5263435592070551, "grad_norm": 0.1775294840335846, "learning_rate": 1.32123273300787e-05, "loss": 0.4029, "step": 20250 }, { "epoch": 1.5301123087359614, "grad_norm": 0.18354074656963348, "learning_rate": 1.3012509380046745e-05, "loss": 0.4016, "step": 20300 }, { "epoch": 1.5338810582648676, "grad_norm": 0.19483138620853424, "learning_rate": 1.2813987677432242e-05, "loss": 0.4021, "step": 20350 }, { "epoch": 1.5376498077937741, "grad_norm": 0.1839035302400589, "learning_rate": 1.2616769179539944e-05, "loss": 0.4021, "step": 20400 }, { "epoch": 1.5414185573226802, "grad_norm": 0.1727752387523651, "learning_rate": 1.2420860798003047e-05, "loss": 0.4008, "step": 20450 }, { "epoch": 1.5451873068515867, "grad_norm": 0.1786087602376938, "learning_rate": 1.222626939854103e-05, "loss": 0.4008, "step": 20500 }, { "epoch": 1.548956056380493, "grad_norm": 0.17971089482307434, "learning_rate": 1.203300180071894e-05, "loss": 0.4003, "step": 20550 }, { "epoch": 1.5527248059093992, "grad_norm": 0.17970240116119385, "learning_rate": 1.1841064777708483e-05, "loss": 0.4025, "step": 20600 }, { "epoch": 1.5564935554383057, "grad_norm": 0.1815640926361084, "learning_rate": 1.1650465056050597e-05, "loss": 0.4029, "step": 20650 }, { "epoch": 1.5602623049672117, "grad_norm": 0.18187712132930756, "learning_rate": 1.1461209315419758e-05, "loss": 0.4014, "step": 20700 }, { "epoch": 1.5640310544961182, "grad_norm": 0.18395179510116577, "learning_rate": 1.1273304188389882e-05, "loss": 0.4022, "step": 20750 }, { "epoch": 1.5677998040250245, "grad_norm": 0.18109290301799774, "learning_rate": 1.1086756260201859e-05, "loss": 0.4001, "step": 20800 }, { "epoch": 1.5715685535539308, "grad_norm": 0.17941805720329285, "learning_rate": 1.0901572068532773e-05, "loss": 0.4017, "step": 20850 }, { "epoch": 1.5753373030828373, "grad_norm": 0.1812305897474289, "learning_rate": 1.0717758103266805e-05, "loss": 0.4003, "step": 20900 }, { "epoch": 1.5791060526117433, "grad_norm": 0.1804351657629013, "learning_rate": 1.0535320806267767e-05, "loss": 0.4007, "step": 20950 }, { "epoch": 1.5828748021406498, "grad_norm": 0.1748201847076416, "learning_rate": 1.0354266571153399e-05, "loss": 0.4004, "step": 21000 }, { "epoch": 1.586643551669556, "grad_norm": 0.18150129914283752, "learning_rate": 1.0174601743071205e-05, "loss": 0.402, "step": 21050 }, { "epoch": 1.5904123011984623, "grad_norm": 0.175798237323761, "learning_rate": 9.996332618476172e-06, "loss": 0.402, "step": 21100 }, { "epoch": 1.5941810507273688, "grad_norm": 0.1742759644985199, "learning_rate": 9.819465444910058e-06, "loss": 0.4001, "step": 21150 }, { "epoch": 1.5979498002562749, "grad_norm": 0.19138532876968384, "learning_rate": 9.644006420782476e-06, "loss": 0.4002, "step": 21200 }, { "epoch": 1.6017185497851814, "grad_norm": 0.18248885869979858, "learning_rate": 9.469961695153606e-06, "loss": 0.4018, "step": 21250 }, { "epoch": 1.6054872993140876, "grad_norm": 0.1761186271905899, "learning_rate": 9.29733736751881e-06, "loss": 0.4027, "step": 21300 }, { "epoch": 1.609256048842994, "grad_norm": 0.17524461448192596, "learning_rate": 9.126139487594749e-06, "loss": 0.4, "step": 21350 }, { "epoch": 1.6130247983719002, "grad_norm": 0.18030941486358643, "learning_rate": 8.956374055107442e-06, "loss": 0.4008, "step": 21400 }, { "epoch": 1.6167935479008064, "grad_norm": 0.17748509347438812, "learning_rate": 8.78804701958198e-06, "loss": 0.3987, "step": 21450 }, { "epoch": 1.620562297429713, "grad_norm": 0.180355504155159, "learning_rate": 8.621164280134004e-06, "loss": 0.4024, "step": 21500 }, { "epoch": 1.6243310469586192, "grad_norm": 0.1831391304731369, "learning_rate": 8.45573168526303e-06, "loss": 0.3987, "step": 21550 }, { "epoch": 1.6280997964875255, "grad_norm": 0.1812552660703659, "learning_rate": 8.291755032647402e-06, "loss": 0.3981, "step": 21600 }, { "epoch": 1.6318685460164317, "grad_norm": 0.18903885781764984, "learning_rate": 8.129240068941163e-06, "loss": 0.3995, "step": 21650 }, { "epoch": 1.635637295545338, "grad_norm": 0.18726922571659088, "learning_rate": 7.96819248957265e-06, "loss": 0.4003, "step": 21700 }, { "epoch": 1.6394060450742445, "grad_norm": 0.18688932061195374, "learning_rate": 7.808617938544866e-06, "loss": 0.3972, "step": 21750 }, { "epoch": 1.6431747946031505, "grad_norm": 0.17523853480815887, "learning_rate": 7.650522008237754e-06, "loss": 0.4016, "step": 21800 }, { "epoch": 1.646943544132057, "grad_norm": 0.1828313022851944, "learning_rate": 7.493910239212121e-06, "loss": 0.4014, "step": 21850 }, { "epoch": 1.6507122936609633, "grad_norm": 0.18200884759426117, "learning_rate": 7.338788120015522e-06, "loss": 0.3977, "step": 21900 }, { "epoch": 1.6544810431898695, "grad_norm": 0.1785590499639511, "learning_rate": 7.185161086989883e-06, "loss": 0.3981, "step": 21950 }, { "epoch": 1.658249792718776, "grad_norm": 0.17938192188739777, "learning_rate": 7.033034524081023e-06, "loss": 0.3986, "step": 22000 }, { "epoch": 1.662018542247682, "grad_norm": 0.18355585634708405, "learning_rate": 6.882413762649914e-06, "loss": 0.3996, "step": 22050 }, { "epoch": 1.6657872917765886, "grad_norm": 0.17709852755069733, "learning_rate": 6.733304081285874e-06, "loss": 0.3994, "step": 22100 }, { "epoch": 1.6695560413054948, "grad_norm": 0.1820652186870575, "learning_rate": 6.585710705621584e-06, "loss": 0.3991, "step": 22150 }, { "epoch": 1.673324790834401, "grad_norm": 0.17669349908828735, "learning_rate": 6.439638808149923e-06, "loss": 0.4008, "step": 22200 }, { "epoch": 1.6770935403633076, "grad_norm": 0.17596611380577087, "learning_rate": 6.295093508042716e-06, "loss": 0.3995, "step": 22250 }, { "epoch": 1.6808622898922136, "grad_norm": 0.18253453075885773, "learning_rate": 6.152079870971311e-06, "loss": 0.3971, "step": 22300 }, { "epoch": 1.6846310394211201, "grad_norm": 0.1786227822303772, "learning_rate": 6.0106029089291025e-06, "loss": 0.3983, "step": 22350 }, { "epoch": 1.6883997889500264, "grad_norm": 0.18454194068908691, "learning_rate": 5.870667580055805e-06, "loss": 0.3999, "step": 22400 }, { "epoch": 1.6921685384789327, "grad_norm": 0.17887701094150543, "learning_rate": 5.732278788463746e-06, "loss": 0.3971, "step": 22450 }, { "epoch": 1.6959372880078392, "grad_norm": 0.18452878296375275, "learning_rate": 5.595441384065986e-06, "loss": 0.3975, "step": 22500 }, { "epoch": 1.6997060375367452, "grad_norm": 0.17868489027023315, "learning_rate": 5.460160162406336e-06, "loss": 0.4006, "step": 22550 }, { "epoch": 1.7034747870656517, "grad_norm": 0.17659594118595123, "learning_rate": 5.3264398644913114e-06, "loss": 0.3974, "step": 22600 }, { "epoch": 1.707243536594558, "grad_norm": 0.17893588542938232, "learning_rate": 5.194285176624003e-06, "loss": 0.3983, "step": 22650 }, { "epoch": 1.7110122861234642, "grad_norm": 0.17637696862220764, "learning_rate": 5.063700730239784e-06, "loss": 0.3977, "step": 22700 }, { "epoch": 1.7147810356523705, "grad_norm": 0.17973153293132782, "learning_rate": 4.934691101744043e-06, "loss": 0.3977, "step": 22750 }, { "epoch": 1.7185497851812768, "grad_norm": 0.1788349598646164, "learning_rate": 4.807260812351793e-06, "loss": 0.3967, "step": 22800 }, { "epoch": 1.7223185347101833, "grad_norm": 0.1776108741760254, "learning_rate": 4.681414327929201e-06, "loss": 0.4005, "step": 22850 }, { "epoch": 1.7260872842390895, "grad_norm": 0.18279901146888733, "learning_rate": 4.557156058837137e-06, "loss": 0.3968, "step": 22900 }, { "epoch": 1.7298560337679958, "grad_norm": 0.18151997029781342, "learning_rate": 4.4344903597765386e-06, "loss": 0.4007, "step": 22950 }, { "epoch": 1.733624783296902, "grad_norm": 0.17980162799358368, "learning_rate": 4.31342152963583e-06, "loss": 0.3978, "step": 23000 }, { "epoch": 1.7373935328258083, "grad_norm": 0.1825917512178421, "learning_rate": 4.19395381134029e-06, "loss": 0.3984, "step": 23050 }, { "epoch": 1.7411622823547148, "grad_norm": 0.17702624201774597, "learning_rate": 4.076091391703302e-06, "loss": 0.3996, "step": 23100 }, { "epoch": 1.7449310318836209, "grad_norm": 0.18847475945949554, "learning_rate": 3.9598384012796755e-06, "loss": 0.3985, "step": 23150 }, { "epoch": 1.7486997814125274, "grad_norm": 0.17738161981105804, "learning_rate": 3.845198914220871e-06, "loss": 0.396, "step": 23200 }, { "epoch": 1.7524685309414336, "grad_norm": 0.17686378955841064, "learning_rate": 3.732176948132199e-06, "loss": 0.3968, "step": 23250 }, { "epoch": 1.75623728047034, "grad_norm": 0.17942722141742706, "learning_rate": 3.6207764639320462e-06, "loss": 0.3973, "step": 23300 }, { "epoch": 1.7600060299992464, "grad_norm": 0.17456026375293732, "learning_rate": 3.511001365713057e-06, "loss": 0.3966, "step": 23350 }, { "epoch": 1.7637747795281524, "grad_norm": 0.17901940643787384, "learning_rate": 3.4028555006052953e-06, "loss": 0.3958, "step": 23400 }, { "epoch": 1.767543529057059, "grad_norm": 0.1767757087945938, "learning_rate": 3.2963426586414657e-06, "loss": 0.3976, "step": 23450 }, { "epoch": 1.7713122785859652, "grad_norm": 0.18666401505470276, "learning_rate": 3.191466572624019e-06, "loss": 0.3952, "step": 23500 }, { "epoch": 1.7750810281148715, "grad_norm": 0.17169831693172455, "learning_rate": 3.088230917994406e-06, "loss": 0.3964, "step": 23550 }, { "epoch": 1.778849777643778, "grad_norm": 0.18045750260353088, "learning_rate": 2.986639312704209e-06, "loss": 0.3949, "step": 23600 }, { "epoch": 1.782618527172684, "grad_norm": 0.17623819410800934, "learning_rate": 2.8866953170883948e-06, "loss": 0.396, "step": 23650 }, { "epoch": 1.7863872767015905, "grad_norm": 0.1810029298067093, "learning_rate": 2.788402433740517e-06, "loss": 0.3973, "step": 23700 }, { "epoch": 1.7901560262304967, "grad_norm": 0.17791134119033813, "learning_rate": 2.691764107389977e-06, "loss": 0.3959, "step": 23750 }, { "epoch": 1.793924775759403, "grad_norm": 0.1789197474718094, "learning_rate": 2.596783724781282e-06, "loss": 0.3974, "step": 23800 }, { "epoch": 1.7976935252883095, "grad_norm": 0.17759652435779572, "learning_rate": 2.5034646145553788e-06, "loss": 0.3974, "step": 23850 }, { "epoch": 1.8014622748172155, "grad_norm": 0.1805141419172287, "learning_rate": 2.4118100471329787e-06, "loss": 0.3959, "step": 23900 }, { "epoch": 1.805231024346122, "grad_norm": 0.18057376146316528, "learning_rate": 2.3218232345999625e-06, "loss": 0.3969, "step": 23950 }, { "epoch": 1.8089997738750283, "grad_norm": 0.1702955961227417, "learning_rate": 2.2335073305948086e-06, "loss": 0.394, "step": 24000 }, { "epoch": 1.8127685234039346, "grad_norm": 0.1743067502975464, "learning_rate": 2.1468654301980527e-06, "loss": 0.397, "step": 24050 }, { "epoch": 1.8165372729328408, "grad_norm": 0.17483128607273102, "learning_rate": 2.0619005698238437e-06, "loss": 0.3975, "step": 24100 }, { "epoch": 1.820306022461747, "grad_norm": 0.18305592238903046, "learning_rate": 1.9786157271135032e-06, "loss": 0.3971, "step": 24150 }, { "epoch": 1.8240747719906536, "grad_norm": 0.17903833091259003, "learning_rate": 1.8970138208311949e-06, "loss": 0.3973, "step": 24200 }, { "epoch": 1.8278435215195596, "grad_norm": 0.18153904378414154, "learning_rate": 1.817097710761656e-06, "loss": 0.3944, "step": 24250 }, { "epoch": 1.8316122710484661, "grad_norm": 0.1754598170518875, "learning_rate": 1.7388701976099041e-06, "loss": 0.3953, "step": 24300 }, { "epoch": 1.8353810205773724, "grad_norm": 0.18127521872520447, "learning_rate": 1.6623340229031503e-06, "loss": 0.3973, "step": 24350 }, { "epoch": 1.8391497701062787, "grad_norm": 0.17995725572109222, "learning_rate": 1.5874918688946972e-06, "loss": 0.3975, "step": 24400 }, { "epoch": 1.8429185196351852, "grad_norm": 0.18092754483222961, "learning_rate": 1.5143463584699424e-06, "loss": 0.3955, "step": 24450 }, { "epoch": 1.8466872691640912, "grad_norm": 0.1810247302055359, "learning_rate": 1.4429000550544414e-06, "loss": 0.3948, "step": 24500 }, { "epoch": 1.8504560186929977, "grad_norm": 0.18044689297676086, "learning_rate": 1.3731554625241006e-06, "loss": 0.3974, "step": 24550 }, { "epoch": 1.854224768221904, "grad_norm": 0.17675849795341492, "learning_rate": 1.305115025117387e-06, "loss": 0.3968, "step": 24600 }, { "epoch": 1.8579935177508102, "grad_norm": 0.17403309047222137, "learning_rate": 1.2387811273497252e-06, "loss": 0.3973, "step": 24650 }, { "epoch": 1.8617622672797167, "grad_norm": 0.17401446402072906, "learning_rate": 1.1741560939298791e-06, "loss": 0.3956, "step": 24700 }, { "epoch": 1.8655310168086228, "grad_norm": 0.18209025263786316, "learning_rate": 1.1112421896784853e-06, "loss": 0.3959, "step": 24750 }, { "epoch": 1.8692997663375293, "grad_norm": 0.17801640927791595, "learning_rate": 1.0500416194487384e-06, "loss": 0.3966, "step": 24800 }, { "epoch": 1.8730685158664355, "grad_norm": 0.17677666246891022, "learning_rate": 9.905565280490302e-07, "loss": 0.3983, "step": 24850 }, { "epoch": 1.8768372653953418, "grad_norm": 0.17637313902378082, "learning_rate": 9.327890001678719e-07, "loss": 0.3974, "step": 24900 }, { "epoch": 1.8806060149242483, "grad_norm": 0.17628493905067444, "learning_rate": 8.76741060300762e-07, "loss": 0.3967, "step": 24950 }, { "epoch": 1.8843747644531543, "grad_norm": 0.17734892666339874, "learning_rate": 8.224146726792947e-07, "loss": 0.3936, "step": 25000 }, { "epoch": 1.8881435139820608, "grad_norm": 0.17440645396709442, "learning_rate": 7.698117412022864e-07, "loss": 0.3964, "step": 25050 }, { "epoch": 1.891912263510967, "grad_norm": 0.1781737059354782, "learning_rate": 7.189341093690627e-07, "loss": 0.3961, "step": 25100 }, { "epoch": 1.8956810130398734, "grad_norm": 0.17803654074668884, "learning_rate": 6.697835602148483e-07, "loss": 0.3943, "step": 25150 }, { "epoch": 1.8994497625687798, "grad_norm": 0.17455802857875824, "learning_rate": 6.223618162483014e-07, "loss": 0.3958, "step": 25200 }, { "epoch": 1.9032185120976859, "grad_norm": 0.17685271799564362, "learning_rate": 5.766705393911165e-07, "loss": 0.3968, "step": 25250 }, { "epoch": 1.9069872616265924, "grad_norm": 0.18048785626888275, "learning_rate": 5.327113309197828e-07, "loss": 0.3944, "step": 25300 }, { "epoch": 1.9107560111554986, "grad_norm": 0.17648503184318542, "learning_rate": 4.904857314095068e-07, "loss": 0.3954, "step": 25350 }, { "epoch": 1.914524760684405, "grad_norm": 0.17615610361099243, "learning_rate": 4.4999522068017164e-07, "loss": 0.3963, "step": 25400 }, { "epoch": 1.9182935102133112, "grad_norm": 0.17669633030891418, "learning_rate": 4.112412177445124e-07, "loss": 0.3965, "step": 25450 }, { "epoch": 1.9220622597422174, "grad_norm": 0.17653539776802063, "learning_rate": 3.7422508075835583e-07, "loss": 0.397, "step": 25500 }, { "epoch": 1.925831009271124, "grad_norm": 0.17871299386024475, "learning_rate": 3.3894810697305267e-07, "loss": 0.3952, "step": 25550 }, { "epoch": 1.92959975880003, "grad_norm": 0.18262889981269836, "learning_rate": 3.05411532689992e-07, "loss": 0.3979, "step": 25600 }, { "epoch": 1.9333685083289365, "grad_norm": 0.1723639965057373, "learning_rate": 2.7361653321729133e-07, "loss": 0.3959, "step": 25650 }, { "epoch": 1.9371372578578427, "grad_norm": 0.1799677610397339, "learning_rate": 2.435642228285906e-07, "loss": 0.3959, "step": 25700 }, { "epoch": 1.940906007386749, "grad_norm": 0.18162404000759125, "learning_rate": 2.15255654724017e-07, "loss": 0.3957, "step": 25750 }, { "epoch": 1.9446747569156555, "grad_norm": 0.17320673167705536, "learning_rate": 1.886918209932642e-07, "loss": 0.395, "step": 25800 }, { "epoch": 1.9484435064445615, "grad_norm": 0.1870715469121933, "learning_rate": 1.63873652580826e-07, "loss": 0.3966, "step": 25850 }, { "epoch": 1.952212255973468, "grad_norm": 0.1799531877040863, "learning_rate": 1.4080201925338322e-07, "loss": 0.397, "step": 25900 }, { "epoch": 1.9559810055023743, "grad_norm": 0.18043088912963867, "learning_rate": 1.1947772956930615e-07, "loss": 0.3956, "step": 25950 }, { "epoch": 1.9597497550312806, "grad_norm": 0.18167664110660553, "learning_rate": 9.99015308503215e-08, "loss": 0.3967, "step": 26000 }, { "epoch": 1.963518504560187, "grad_norm": 0.1858936846256256, "learning_rate": 8.207410915532232e-08, "loss": 0.3968, "step": 26050 }, { "epoch": 1.967287254089093, "grad_norm": 0.17515628039836884, "learning_rate": 6.599608925633715e-08, "loss": 0.3951, "step": 26100 }, { "epoch": 1.9710560036179996, "grad_norm": 0.17847982048988342, "learning_rate": 5.166803461661429e-08, "loss": 0.3944, "step": 26150 }, { "epoch": 1.9748247531469059, "grad_norm": 0.17742793262004852, "learning_rate": 3.909044737089307e-08, "loss": 0.3963, "step": 26200 }, { "epoch": 1.9785935026758121, "grad_norm": 0.17997916042804718, "learning_rate": 2.826376830779576e-08, "loss": 0.3948, "step": 26250 }, { "epoch": 1.9823622522047186, "grad_norm": 0.1779303252696991, "learning_rate": 1.9188376854373246e-08, "loss": 0.396, "step": 26300 }, { "epoch": 1.9861310017336247, "grad_norm": 0.18305301666259766, "learning_rate": 1.1864591062832331e-08, "loss": 0.397, "step": 26350 }, { "epoch": 1.9898997512625312, "grad_norm": 0.1764865517616272, "learning_rate": 6.292667599366864e-09, "loss": 0.3953, "step": 26400 }, { "epoch": 1.9936685007914374, "grad_norm": 0.17203836143016815, "learning_rate": 2.4728017351649534e-09, "loss": 0.3968, "step": 26450 }, { "epoch": 1.9974372503203437, "grad_norm": 0.17117449641227722, "learning_rate": 4.0512733956998837e-10, "loss": 0.3977, "step": 26500 } ], "logging_steps": 50, "max_steps": 26534, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1.017526223018121e+20, "train_batch_size": 2, "trial_name": null, "trial_params": null }