| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 1.9982692973347178, |
| "eval_steps": 500, |
| "global_step": 1444, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0013845621322256837, |
| "grad_norm": 49.28921127319336, |
| "learning_rate": 0.0, |
| "loss": 3.0795, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.006922810661128418, |
| "grad_norm": 18.114784240722656, |
| "learning_rate": 5.517241379310345e-06, |
| "loss": 2.9478, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.013845621322256836, |
| "grad_norm": 6.451106071472168, |
| "learning_rate": 1.2413793103448277e-05, |
| "loss": 2.3094, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.020768431983385256, |
| "grad_norm": 7.723776817321777, |
| "learning_rate": 1.9310344827586207e-05, |
| "loss": 1.9825, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.027691242644513673, |
| "grad_norm": 9.21464729309082, |
| "learning_rate": 2.620689655172414e-05, |
| "loss": 1.7586, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.034614053305642094, |
| "grad_norm": 7.424780368804932, |
| "learning_rate": 3.310344827586207e-05, |
| "loss": 1.5374, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.04153686396677051, |
| "grad_norm": 3.526257038116455, |
| "learning_rate": 4e-05, |
| "loss": 1.28, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.04845967462789893, |
| "grad_norm": 2.5668952465057373, |
| "learning_rate": 4.689655172413793e-05, |
| "loss": 1.2187, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.055382485289027346, |
| "grad_norm": 3.9412596225738525, |
| "learning_rate": 5.379310344827586e-05, |
| "loss": 1.1574, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.06230529595015576, |
| "grad_norm": 2.965135097503662, |
| "learning_rate": 6.068965517241379e-05, |
| "loss": 1.089, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.06922810661128419, |
| "grad_norm": 2.2038655281066895, |
| "learning_rate": 6.758620689655173e-05, |
| "loss": 1.0374, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.0761509172724126, |
| "grad_norm": 1.856432557106018, |
| "learning_rate": 7.448275862068966e-05, |
| "loss": 1.0215, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.08307372793354102, |
| "grad_norm": 4.251297473907471, |
| "learning_rate": 8.137931034482759e-05, |
| "loss": 1.1169, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.08999653859466944, |
| "grad_norm": 4.693065166473389, |
| "learning_rate": 8.827586206896552e-05, |
| "loss": 1.0188, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.09691934925579786, |
| "grad_norm": 2.24104642868042, |
| "learning_rate": 9.517241379310345e-05, |
| "loss": 1.0059, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.10384215991692627, |
| "grad_norm": 1.4557360410690308, |
| "learning_rate": 0.0001020689655172414, |
| "loss": 0.9839, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.11076497057805469, |
| "grad_norm": 1.4282116889953613, |
| "learning_rate": 0.00010896551724137931, |
| "loss": 1.01, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.11768778123918311, |
| "grad_norm": 1.2591500282287598, |
| "learning_rate": 0.00011586206896551725, |
| "loss": 1.0043, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.12461059190031153, |
| "grad_norm": 1.0611705780029297, |
| "learning_rate": 0.00012275862068965518, |
| "loss": 1.0127, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.13153340256143994, |
| "grad_norm": 1.1061965227127075, |
| "learning_rate": 0.0001296551724137931, |
| "loss": 0.9933, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.13845621322256838, |
| "grad_norm": 1.3926725387573242, |
| "learning_rate": 0.00013655172413793104, |
| "loss": 1.0056, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.14537902388369678, |
| "grad_norm": 0.986765444278717, |
| "learning_rate": 0.00014344827586206896, |
| "loss": 1.0034, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.1523018345448252, |
| "grad_norm": 0.9959716796875, |
| "learning_rate": 0.0001503448275862069, |
| "loss": 1.0215, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.1592246452059536, |
| "grad_norm": 0.9828895926475525, |
| "learning_rate": 0.00015724137931034485, |
| "loss": 1.0114, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.16614745586708204, |
| "grad_norm": 0.9268749356269836, |
| "learning_rate": 0.00016413793103448276, |
| "loss": 1.004, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.17307026652821045, |
| "grad_norm": 0.949165403842926, |
| "learning_rate": 0.0001710344827586207, |
| "loss": 1.0372, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.17999307718933888, |
| "grad_norm": 0.7588331699371338, |
| "learning_rate": 0.00017793103448275862, |
| "loss": 1.0159, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.18691588785046728, |
| "grad_norm": 0.846207857131958, |
| "learning_rate": 0.00018482758620689654, |
| "loss": 1.0207, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.19383869851159571, |
| "grad_norm": 0.9587262272834778, |
| "learning_rate": 0.0001917241379310345, |
| "loss": 1.0175, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.20076150917272412, |
| "grad_norm": 0.8179382681846619, |
| "learning_rate": 0.00019862068965517243, |
| "loss": 1.0181, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.20768431983385255, |
| "grad_norm": 0.8082414865493774, |
| "learning_rate": 0.0001999953208384105, |
| "loss": 1.0108, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.21460713049498095, |
| "grad_norm": 0.7612893581390381, |
| "learning_rate": 0.0001999763124949378, |
| "loss": 0.9901, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.22152994115610938, |
| "grad_norm": 0.8941105008125305, |
| "learning_rate": 0.0001999426852993044, |
| "loss": 0.9983, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.2284527518172378, |
| "grad_norm": 0.7896144390106201, |
| "learning_rate": 0.00019989444416858502, |
| "loss": 0.9707, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.23537556247836622, |
| "grad_norm": 0.735113799571991, |
| "learning_rate": 0.0001998315961567502, |
| "loss": 0.9955, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.24229837313949462, |
| "grad_norm": 0.7563541531562805, |
| "learning_rate": 0.00019975415045363467, |
| "loss": 0.9878, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.24922118380062305, |
| "grad_norm": 0.8806726932525635, |
| "learning_rate": 0.00019966211838359378, |
| "loss": 0.976, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.25614399446175146, |
| "grad_norm": 0.7841496467590332, |
| "learning_rate": 0.00019955551340384743, |
| "loss": 0.9606, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.2630668051228799, |
| "grad_norm": 0.5848103761672974, |
| "learning_rate": 0.00019943435110251248, |
| "loss": 0.9877, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.2699896157840083, |
| "grad_norm": 0.7116856575012207, |
| "learning_rate": 0.0001992986491963232, |
| "loss": 0.955, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.27691242644513675, |
| "grad_norm": 0.6569689512252808, |
| "learning_rate": 0.00019914842752804103, |
| "loss": 0.9524, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.2838352371062651, |
| "grad_norm": 0.7533493638038635, |
| "learning_rate": 0.00019898370806355263, |
| "loss": 0.9232, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.29075804776739356, |
| "grad_norm": 0.7606753706932068, |
| "learning_rate": 0.00019880451488865836, |
| "loss": 0.9854, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.297680858428522, |
| "grad_norm": 7.585168361663818, |
| "learning_rate": 0.00019861087420555018, |
| "loss": 1.1045, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.3046036690896504, |
| "grad_norm": 9.095848083496094, |
| "learning_rate": 0.0001984028143289803, |
| "loss": 1.2125, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.3115264797507788, |
| "grad_norm": 1.6645023822784424, |
| "learning_rate": 0.00019818036568212106, |
| "loss": 1.0607, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.3184492904119072, |
| "grad_norm": 1.3035401105880737, |
| "learning_rate": 0.00019794356079211604, |
| "loss": 0.9415, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.32537210107303566, |
| "grad_norm": 0.7202489376068115, |
| "learning_rate": 0.00019769243428532422, |
| "loss": 0.9706, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.3322949117341641, |
| "grad_norm": 0.8300184011459351, |
| "learning_rate": 0.00019742702288225652, |
| "loss": 0.926, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.33921772239529246, |
| "grad_norm": 0.6376895308494568, |
| "learning_rate": 0.00019714736539220648, |
| "loss": 0.9173, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.3461405330564209, |
| "grad_norm": 0.7818527221679688, |
| "learning_rate": 0.00019685350270757555, |
| "loss": 0.9413, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.3530633437175493, |
| "grad_norm": 0.6731443405151367, |
| "learning_rate": 0.0001965454777978936, |
| "loss": 0.954, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.35998615437867776, |
| "grad_norm": 3.000696897506714, |
| "learning_rate": 0.00019622333570353567, |
| "loss": 0.9432, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.36690896503980613, |
| "grad_norm": 0.7889404892921448, |
| "learning_rate": 0.00019588712352913625, |
| "loss": 0.9471, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.37383177570093457, |
| "grad_norm": 0.8370787501335144, |
| "learning_rate": 0.00019553689043670127, |
| "loss": 0.9521, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.380754586362063, |
| "grad_norm": 0.6203432679176331, |
| "learning_rate": 0.00019517268763841962, |
| "loss": 0.9263, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.38767739702319143, |
| "grad_norm": 0.5970066785812378, |
| "learning_rate": 0.00019479456838917473, |
| "loss": 0.9014, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.39460020768431986, |
| "grad_norm": 0.6280853748321533, |
| "learning_rate": 0.0001944025879787574, |
| "loss": 0.8933, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.40152301834544823, |
| "grad_norm": 0.8126713037490845, |
| "learning_rate": 0.0001939968037237812, |
| "loss": 0.9203, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.40844582900657667, |
| "grad_norm": 0.5656815767288208, |
| "learning_rate": 0.00019357727495930146, |
| "loss": 0.8971, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.4153686396677051, |
| "grad_norm": 0.6746362447738647, |
| "learning_rate": 0.00019314406303013904, |
| "loss": 0.8974, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.42229145032883353, |
| "grad_norm": 0.6263439655303955, |
| "learning_rate": 0.00019269723128191048, |
| "loss": 0.891, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.4292142609899619, |
| "grad_norm": 0.4856240451335907, |
| "learning_rate": 0.00019223684505176517, |
| "loss": 0.8999, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.43613707165109034, |
| "grad_norm": 0.5870519280433655, |
| "learning_rate": 0.00019176297165883165, |
| "loss": 0.9049, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.44305988231221877, |
| "grad_norm": 0.5149969458580017, |
| "learning_rate": 0.00019127568039437406, |
| "loss": 0.886, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.4499826929733472, |
| "grad_norm": 0.5945581793785095, |
| "learning_rate": 0.00019077504251166006, |
| "loss": 0.8818, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.4569055036344756, |
| "grad_norm": 0.5872354507446289, |
| "learning_rate": 0.00019026113121554204, |
| "loss": 0.8882, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.463828314295604, |
| "grad_norm": 0.5640285611152649, |
| "learning_rate": 0.00018973402165175268, |
| "loss": 0.8928, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.47075112495673244, |
| "grad_norm": 0.5472027659416199, |
| "learning_rate": 0.0001891937908959172, |
| "loss": 0.8807, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.47767393561786087, |
| "grad_norm": 0.5739114284515381, |
| "learning_rate": 0.00018864051794228282, |
| "loss": 0.8869, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.48459674627898924, |
| "grad_norm": 0.580653190612793, |
| "learning_rate": 0.00018807428369216822, |
| "loss": 0.8543, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.4915195569401177, |
| "grad_norm": 0.5018455982208252, |
| "learning_rate": 0.00018749517094213375, |
| "loss": 0.8548, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.4984423676012461, |
| "grad_norm": 0.5900467038154602, |
| "learning_rate": 0.00018690326437187472, |
| "loss": 0.8685, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.5053651782623745, |
| "grad_norm": 0.5394186973571777, |
| "learning_rate": 0.00018629865053183911, |
| "loss": 0.8626, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.5122879889235029, |
| "grad_norm": 0.6365518569946289, |
| "learning_rate": 0.0001856814178305722, |
| "loss": 0.8474, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.5192107995846313, |
| "grad_norm": 0.5121780037879944, |
| "learning_rate": 0.00018505165652178893, |
| "loss": 0.8448, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.5261336102457598, |
| "grad_norm": 0.5004256367683411, |
| "learning_rate": 0.00018440945869117675, |
| "loss": 0.8681, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.5330564209068882, |
| "grad_norm": 0.5059666633605957, |
| "learning_rate": 0.00018375491824293067, |
| "loss": 0.8752, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.5399792315680166, |
| "grad_norm": 0.4677795171737671, |
| "learning_rate": 0.00018308813088602227, |
| "loss": 0.8695, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.5469020422291451, |
| "grad_norm": 0.5352007746696472, |
| "learning_rate": 0.00018240919412020466, |
| "loss": 0.8468, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.5538248528902735, |
| "grad_norm": 0.5106504559516907, |
| "learning_rate": 0.00018171820722175604, |
| "loss": 0.8699, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.5607476635514018, |
| "grad_norm": 0.4898536801338196, |
| "learning_rate": 0.00018101527122896293, |
| "loss": 0.8607, |
| "step": 405 |
| }, |
| { |
| "epoch": 0.5676704742125303, |
| "grad_norm": 0.49010294675827026, |
| "learning_rate": 0.0001803004889273463, |
| "loss": 0.8462, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.5745932848736587, |
| "grad_norm": 0.4630247950553894, |
| "learning_rate": 0.0001795739648346317, |
| "loss": 0.8243, |
| "step": 415 |
| }, |
| { |
| "epoch": 0.5815160955347871, |
| "grad_norm": 0.4984998404979706, |
| "learning_rate": 0.00017883580518546647, |
| "loss": 0.8262, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.5884389061959155, |
| "grad_norm": 0.5157510638237, |
| "learning_rate": 0.00017808611791588584, |
| "loss": 0.8397, |
| "step": 425 |
| }, |
| { |
| "epoch": 0.595361716857044, |
| "grad_norm": 0.48661330342292786, |
| "learning_rate": 0.00017732501264752996, |
| "loss": 0.8267, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.6022845275181724, |
| "grad_norm": 0.47257551550865173, |
| "learning_rate": 0.00017655260067161503, |
| "loss": 0.8519, |
| "step": 435 |
| }, |
| { |
| "epoch": 0.6092073381793008, |
| "grad_norm": 0.5127261877059937, |
| "learning_rate": 0.00017576899493265954, |
| "loss": 0.8355, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.6161301488404292, |
| "grad_norm": 0.45928096771240234, |
| "learning_rate": 0.00017497431001196943, |
| "loss": 0.8423, |
| "step": 445 |
| }, |
| { |
| "epoch": 0.6230529595015576, |
| "grad_norm": 0.45476192235946655, |
| "learning_rate": 0.00017416866211088366, |
| "loss": 0.847, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.629975770162686, |
| "grad_norm": 0.48962315917015076, |
| "learning_rate": 0.00017335216903378267, |
| "loss": 0.8317, |
| "step": 455 |
| }, |
| { |
| "epoch": 0.6368985808238145, |
| "grad_norm": 0.5388792753219604, |
| "learning_rate": 0.00017252495017086296, |
| "loss": 0.8223, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.6438213914849429, |
| "grad_norm": 0.4656524658203125, |
| "learning_rate": 0.0001716871264806791, |
| "loss": 0.8355, |
| "step": 465 |
| }, |
| { |
| "epoch": 0.6507442021460713, |
| "grad_norm": 0.5120430588722229, |
| "learning_rate": 0.0001708388204724572, |
| "loss": 0.8331, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.6576670128071997, |
| "grad_norm": 0.48045217990875244, |
| "learning_rate": 0.00016998015618818087, |
| "loss": 0.8478, |
| "step": 475 |
| }, |
| { |
| "epoch": 0.6645898234683282, |
| "grad_norm": 0.44614362716674805, |
| "learning_rate": 0.00016911125918445363, |
| "loss": 0.8286, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.6715126341294566, |
| "grad_norm": 0.44099363684654236, |
| "learning_rate": 0.00016823225651413953, |
| "loss": 0.8185, |
| "step": 485 |
| }, |
| { |
| "epoch": 0.6784354447905849, |
| "grad_norm": 0.45090341567993164, |
| "learning_rate": 0.00016734327670778508, |
| "loss": 0.8107, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.6853582554517134, |
| "grad_norm": 0.5897481441497803, |
| "learning_rate": 0.00016644444975482523, |
| "loss": 0.8067, |
| "step": 495 |
| }, |
| { |
| "epoch": 0.6922810661128418, |
| "grad_norm": 0.5060038566589355, |
| "learning_rate": 0.0001655359070845757, |
| "loss": 0.8121, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.6992038767739702, |
| "grad_norm": 0.4440653324127197, |
| "learning_rate": 0.0001646177815470151, |
| "loss": 0.8181, |
| "step": 505 |
| }, |
| { |
| "epoch": 0.7061266874350987, |
| "grad_norm": 0.4549269676208496, |
| "learning_rate": 0.00016369020739335924, |
| "loss": 0.7997, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.7130494980962271, |
| "grad_norm": 0.45471733808517456, |
| "learning_rate": 0.00016275332025643028, |
| "loss": 0.8154, |
| "step": 515 |
| }, |
| { |
| "epoch": 0.7199723087573555, |
| "grad_norm": 0.4292679727077484, |
| "learning_rate": 0.00016180725713082434, |
| "loss": 0.8234, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.726895119418484, |
| "grad_norm": 0.5121740698814392, |
| "learning_rate": 0.0001608521563528795, |
| "loss": 0.8044, |
| "step": 525 |
| }, |
| { |
| "epoch": 0.7338179300796123, |
| "grad_norm": 0.44355785846710205, |
| "learning_rate": 0.00015988815758044792, |
| "loss": 0.7996, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.7407407407407407, |
| "grad_norm": 0.42263615131378174, |
| "learning_rate": 0.00015891540177247467, |
| "loss": 0.8219, |
| "step": 535 |
| }, |
| { |
| "epoch": 0.7476635514018691, |
| "grad_norm": 0.5358946919441223, |
| "learning_rate": 0.00015793403116838613, |
| "loss": 0.8089, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.7545863620629976, |
| "grad_norm": 0.47450023889541626, |
| "learning_rate": 0.00015694418926729146, |
| "loss": 0.7956, |
| "step": 545 |
| }, |
| { |
| "epoch": 0.761509172724126, |
| "grad_norm": 0.470217227935791, |
| "learning_rate": 0.00015594602080699956, |
| "loss": 0.8058, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.7684319833852544, |
| "grad_norm": 0.5145338773727417, |
| "learning_rate": 0.00015493967174285512, |
| "loss": 0.7972, |
| "step": 555 |
| }, |
| { |
| "epoch": 0.7753547940463829, |
| "grad_norm": 0.4233357310295105, |
| "learning_rate": 0.00015392528922639662, |
| "loss": 0.8084, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.7822776047075113, |
| "grad_norm": 0.41947290301322937, |
| "learning_rate": 0.00015290302158383923, |
| "loss": 0.7981, |
| "step": 565 |
| }, |
| { |
| "epoch": 0.7892004153686397, |
| "grad_norm": 0.46518903970718384, |
| "learning_rate": 0.0001518730182943862, |
| "loss": 0.7962, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.796123226029768, |
| "grad_norm": 0.41341090202331543, |
| "learning_rate": 0.0001508354299683715, |
| "loss": 0.7952, |
| "step": 575 |
| }, |
| { |
| "epoch": 0.8030460366908965, |
| "grad_norm": 0.45617249608039856, |
| "learning_rate": 0.0001497904083252371, |
| "loss": 0.8015, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.8099688473520249, |
| "grad_norm": 0.41003215312957764, |
| "learning_rate": 0.00014873810617134806, |
| "loss": 0.7894, |
| "step": 585 |
| }, |
| { |
| "epoch": 0.8168916580131533, |
| "grad_norm": 0.4336972236633301, |
| "learning_rate": 0.00014767867737764876, |
| "loss": 0.7886, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.8238144686742818, |
| "grad_norm": 0.43226006627082825, |
| "learning_rate": 0.0001466122768571633, |
| "loss": 0.7973, |
| "step": 595 |
| }, |
| { |
| "epoch": 0.8307372793354102, |
| "grad_norm": 0.4515872597694397, |
| "learning_rate": 0.0001455390605423437, |
| "loss": 0.8025, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.8376600899965386, |
| "grad_norm": 0.49443331360816956, |
| "learning_rate": 0.00014445918536226904, |
| "loss": 0.7883, |
| "step": 605 |
| }, |
| { |
| "epoch": 0.8445829006576671, |
| "grad_norm": 0.48309868574142456, |
| "learning_rate": 0.00014337280921969858, |
| "loss": 0.7824, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.8515057113187954, |
| "grad_norm": 0.4243367314338684, |
| "learning_rate": 0.0001422800909679829, |
| "loss": 0.7928, |
| "step": 615 |
| }, |
| { |
| "epoch": 0.8584285219799238, |
| "grad_norm": 0.49469149112701416, |
| "learning_rate": 0.00014118119038783588, |
| "loss": 0.785, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.8653513326410522, |
| "grad_norm": 0.4584200084209442, |
| "learning_rate": 0.00014007626816397087, |
| "loss": 0.804, |
| "step": 625 |
| }, |
| { |
| "epoch": 0.8722741433021807, |
| "grad_norm": 0.4225427210330963, |
| "learning_rate": 0.00013896548586160503, |
| "loss": 0.7888, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.8791969539633091, |
| "grad_norm": 0.45131561160087585, |
| "learning_rate": 0.00013784900590283473, |
| "loss": 0.7906, |
| "step": 635 |
| }, |
| { |
| "epoch": 0.8861197646244375, |
| "grad_norm": 0.4222111999988556, |
| "learning_rate": 0.00013672699154288569, |
| "loss": 0.7748, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.893042575285566, |
| "grad_norm": 0.4449770748615265, |
| "learning_rate": 0.00013559960684624112, |
| "loss": 0.7995, |
| "step": 645 |
| }, |
| { |
| "epoch": 0.8999653859466944, |
| "grad_norm": 0.4661027491092682, |
| "learning_rate": 0.0001344670166626519, |
| "loss": 0.7682, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.9068881966078228, |
| "grad_norm": 0.4581912159919739, |
| "learning_rate": 0.0001333293866030317, |
| "loss": 0.7728, |
| "step": 655 |
| }, |
| { |
| "epoch": 0.9138110072689511, |
| "grad_norm": 0.41657838225364685, |
| "learning_rate": 0.0001321868830152407, |
| "loss": 0.7557, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.9207338179300796, |
| "grad_norm": 0.4423748254776001, |
| "learning_rate": 0.00013103967295976179, |
| "loss": 0.7798, |
| "step": 665 |
| }, |
| { |
| "epoch": 0.927656628591208, |
| "grad_norm": 0.411111980676651, |
| "learning_rate": 0.0001298879241852724, |
| "loss": 0.7777, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.9345794392523364, |
| "grad_norm": 0.4396058917045593, |
| "learning_rate": 0.0001287318051041159, |
| "loss": 0.778, |
| "step": 675 |
| }, |
| { |
| "epoch": 0.9415022499134649, |
| "grad_norm": 0.4244561493396759, |
| "learning_rate": 0.00012757148476767553, |
| "loss": 0.7698, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.9484250605745933, |
| "grad_norm": 0.3860670328140259, |
| "learning_rate": 0.0001264071328416555, |
| "loss": 0.772, |
| "step": 685 |
| }, |
| { |
| "epoch": 0.9553478712357217, |
| "grad_norm": 0.38453128933906555, |
| "learning_rate": 0.00012523891958127168, |
| "loss": 0.7689, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.9622706818968502, |
| "grad_norm": 0.41507506370544434, |
| "learning_rate": 0.0001240670158063565, |
| "loss": 0.7662, |
| "step": 695 |
| }, |
| { |
| "epoch": 0.9691934925579785, |
| "grad_norm": 0.4023298919200897, |
| "learning_rate": 0.00012289159287638105, |
| "loss": 0.7721, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.9761163032191069, |
| "grad_norm": 0.41484305262565613, |
| "learning_rate": 0.00012171282266539845, |
| "loss": 0.7707, |
| "step": 705 |
| }, |
| { |
| "epoch": 0.9830391138802353, |
| "grad_norm": 0.4068526327610016, |
| "learning_rate": 0.00012053087753691172, |
| "loss": 0.7547, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.9899619245413638, |
| "grad_norm": 0.4563179910182953, |
| "learning_rate": 0.00011934593031867038, |
| "loss": 0.7691, |
| "step": 715 |
| }, |
| { |
| "epoch": 0.9968847352024922, |
| "grad_norm": 0.3979925513267517, |
| "learning_rate": 0.00011815815427739911, |
| "loss": 0.7706, |
| "step": 720 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_loss": 0.7720561027526855, |
| "eval_runtime": 59.9017, |
| "eval_samples_per_second": 34.54, |
| "eval_steps_per_second": 2.17, |
| "step": 723 |
| }, |
| { |
| "epoch": 1.0027691242644514, |
| "grad_norm": 0.42567870020866394, |
| "learning_rate": 0.00011696772309346182, |
| "loss": 0.6892, |
| "step": 725 |
| }, |
| { |
| "epoch": 1.0096919349255797, |
| "grad_norm": 0.45581841468811035, |
| "learning_rate": 0.0001157748108354659, |
| "loss": 0.5876, |
| "step": 730 |
| }, |
| { |
| "epoch": 1.0166147455867083, |
| "grad_norm": 0.3980908989906311, |
| "learning_rate": 0.00011457959193480925, |
| "loss": 0.5892, |
| "step": 735 |
| }, |
| { |
| "epoch": 1.0235375562478366, |
| "grad_norm": 0.41051357984542847, |
| "learning_rate": 0.00011338224116017423, |
| "loss": 0.5849, |
| "step": 740 |
| }, |
| { |
| "epoch": 1.0304603669089651, |
| "grad_norm": 0.4377879500389099, |
| "learning_rate": 0.00011218293359197267, |
| "loss": 0.5918, |
| "step": 745 |
| }, |
| { |
| "epoch": 1.0373831775700935, |
| "grad_norm": 0.495897114276886, |
| "learning_rate": 0.00011098184459674484, |
| "loss": 0.5653, |
| "step": 750 |
| }, |
| { |
| "epoch": 1.0443059882312218, |
| "grad_norm": 0.4108388125896454, |
| "learning_rate": 0.00010977914980151691, |
| "loss": 0.5915, |
| "step": 755 |
| }, |
| { |
| "epoch": 1.0512287988923503, |
| "grad_norm": 0.40004926919937134, |
| "learning_rate": 0.00010857502506812029, |
| "loss": 0.5689, |
| "step": 760 |
| }, |
| { |
| "epoch": 1.0581516095534786, |
| "grad_norm": 0.4072938561439514, |
| "learning_rate": 0.00010736964646747629, |
| "loss": 0.5727, |
| "step": 765 |
| }, |
| { |
| "epoch": 1.0650744202146072, |
| "grad_norm": 0.3785325586795807, |
| "learning_rate": 0.00010616319025385089, |
| "loss": 0.5846, |
| "step": 770 |
| }, |
| { |
| "epoch": 1.0719972308757355, |
| "grad_norm": 0.3935575485229492, |
| "learning_rate": 0.00010495583283908177, |
| "loss": 0.5753, |
| "step": 775 |
| }, |
| { |
| "epoch": 1.078920041536864, |
| "grad_norm": 0.41394945979118347, |
| "learning_rate": 0.00010374775076678332, |
| "loss": 0.5608, |
| "step": 780 |
| }, |
| { |
| "epoch": 1.0858428521979924, |
| "grad_norm": 0.4046129882335663, |
| "learning_rate": 0.00010253912068653146, |
| "loss": 0.5848, |
| "step": 785 |
| }, |
| { |
| "epoch": 1.092765662859121, |
| "grad_norm": 0.3733786940574646, |
| "learning_rate": 0.00010133011932803378, |
| "loss": 0.5717, |
| "step": 790 |
| }, |
| { |
| "epoch": 1.0996884735202492, |
| "grad_norm": 0.428487092256546, |
| "learning_rate": 0.00010012092347528705, |
| "loss": 0.5812, |
| "step": 795 |
| }, |
| { |
| "epoch": 1.1066112841813776, |
| "grad_norm": 0.41964441537857056, |
| "learning_rate": 9.891170994072793e-05, |
| "loss": 0.576, |
| "step": 800 |
| }, |
| { |
| "epoch": 1.113534094842506, |
| "grad_norm": 0.46450623869895935, |
| "learning_rate": 9.770265553937831e-05, |
| "loss": 0.5728, |
| "step": 805 |
| }, |
| { |
| "epoch": 1.1204569055036344, |
| "grad_norm": 0.4148964285850525, |
| "learning_rate": 9.649393706299127e-05, |
| "loss": 0.5569, |
| "step": 810 |
| }, |
| { |
| "epoch": 1.127379716164763, |
| "grad_norm": 0.42906996607780457, |
| "learning_rate": 9.52857312541998e-05, |
| "loss": 0.568, |
| "step": 815 |
| }, |
| { |
| "epoch": 1.1343025268258913, |
| "grad_norm": 0.4227571189403534, |
| "learning_rate": 9.407821478067304e-05, |
| "loss": 0.565, |
| "step": 820 |
| }, |
| { |
| "epoch": 1.1412253374870198, |
| "grad_norm": 0.40537649393081665, |
| "learning_rate": 9.287156420928342e-05, |
| "loss": 0.5624, |
| "step": 825 |
| }, |
| { |
| "epoch": 1.1481481481481481, |
| "grad_norm": 0.3608773648738861, |
| "learning_rate": 9.166595598028832e-05, |
| "loss": 0.581, |
| "step": 830 |
| }, |
| { |
| "epoch": 1.1550709588092767, |
| "grad_norm": 0.3971177339553833, |
| "learning_rate": 9.046156638153056e-05, |
| "loss": 0.5614, |
| "step": 835 |
| }, |
| { |
| "epoch": 1.161993769470405, |
| "grad_norm": 0.37738874554634094, |
| "learning_rate": 8.925857152266103e-05, |
| "loss": 0.5662, |
| "step": 840 |
| }, |
| { |
| "epoch": 1.1689165801315333, |
| "grad_norm": 0.409171462059021, |
| "learning_rate": 8.805714730938728e-05, |
| "loss": 0.5739, |
| "step": 845 |
| }, |
| { |
| "epoch": 1.1758393907926619, |
| "grad_norm": 0.3909130394458771, |
| "learning_rate": 8.685746941775219e-05, |
| "loss": 0.562, |
| "step": 850 |
| }, |
| { |
| "epoch": 1.1827622014537902, |
| "grad_norm": 0.41327953338623047, |
| "learning_rate": 8.565971326844584e-05, |
| "loss": 0.5651, |
| "step": 855 |
| }, |
| { |
| "epoch": 1.1896850121149187, |
| "grad_norm": 0.4001190662384033, |
| "learning_rate": 8.44640540011553e-05, |
| "loss": 0.5639, |
| "step": 860 |
| }, |
| { |
| "epoch": 1.196607822776047, |
| "grad_norm": 0.40711870789527893, |
| "learning_rate": 8.327066644895469e-05, |
| "loss": 0.5618, |
| "step": 865 |
| }, |
| { |
| "epoch": 1.2035306334371756, |
| "grad_norm": 0.43218299746513367, |
| "learning_rate": 8.207972511274095e-05, |
| "loss": 0.5585, |
| "step": 870 |
| }, |
| { |
| "epoch": 1.210453444098304, |
| "grad_norm": 0.38643211126327515, |
| "learning_rate": 8.089140413571747e-05, |
| "loss": 0.5578, |
| "step": 875 |
| }, |
| { |
| "epoch": 1.2173762547594325, |
| "grad_norm": 0.41631147265434265, |
| "learning_rate": 7.970587727793051e-05, |
| "loss": 0.5576, |
| "step": 880 |
| }, |
| { |
| "epoch": 1.2242990654205608, |
| "grad_norm": 0.3960113227367401, |
| "learning_rate": 7.852331789086135e-05, |
| "loss": 0.5586, |
| "step": 885 |
| }, |
| { |
| "epoch": 1.231221876081689, |
| "grad_norm": 0.4174184799194336, |
| "learning_rate": 7.73438988920784e-05, |
| "loss": 0.5607, |
| "step": 890 |
| }, |
| { |
| "epoch": 1.2381446867428176, |
| "grad_norm": 0.42617419362068176, |
| "learning_rate": 7.616779273995252e-05, |
| "loss": 0.5637, |
| "step": 895 |
| }, |
| { |
| "epoch": 1.245067497403946, |
| "grad_norm": 0.40726974606513977, |
| "learning_rate": 7.499517140843968e-05, |
| "loss": 0.5641, |
| "step": 900 |
| }, |
| { |
| "epoch": 1.2519903080650745, |
| "grad_norm": 0.417920857667923, |
| "learning_rate": 7.382620636193438e-05, |
| "loss": 0.5589, |
| "step": 905 |
| }, |
| { |
| "epoch": 1.2589131187262028, |
| "grad_norm": 0.43630263209342957, |
| "learning_rate": 7.266106853019759e-05, |
| "loss": 0.5577, |
| "step": 910 |
| }, |
| { |
| "epoch": 1.2658359293873311, |
| "grad_norm": 0.40452563762664795, |
| "learning_rate": 7.149992828336274e-05, |
| "loss": 0.5546, |
| "step": 915 |
| }, |
| { |
| "epoch": 1.2727587400484597, |
| "grad_norm": 0.44135504961013794, |
| "learning_rate": 7.034295540702397e-05, |
| "loss": 0.5525, |
| "step": 920 |
| }, |
| { |
| "epoch": 1.2796815507095882, |
| "grad_norm": 0.40712544322013855, |
| "learning_rate": 6.919031907740922e-05, |
| "loss": 0.5706, |
| "step": 925 |
| }, |
| { |
| "epoch": 1.2866043613707165, |
| "grad_norm": 0.42069798707962036, |
| "learning_rate": 6.804218783664288e-05, |
| "loss": 0.5657, |
| "step": 930 |
| }, |
| { |
| "epoch": 1.2935271720318449, |
| "grad_norm": 0.3738851547241211, |
| "learning_rate": 6.689872956810102e-05, |
| "loss": 0.5503, |
| "step": 935 |
| }, |
| { |
| "epoch": 1.3004499826929734, |
| "grad_norm": 0.4131317436695099, |
| "learning_rate": 6.5760111471863e-05, |
| "loss": 0.5502, |
| "step": 940 |
| }, |
| { |
| "epoch": 1.3073727933541017, |
| "grad_norm": 0.4052620530128479, |
| "learning_rate": 6.46265000402628e-05, |
| "loss": 0.5559, |
| "step": 945 |
| }, |
| { |
| "epoch": 1.3142956040152303, |
| "grad_norm": 0.3770054876804352, |
| "learning_rate": 6.349806103354417e-05, |
| "loss": 0.5505, |
| "step": 950 |
| }, |
| { |
| "epoch": 1.3212184146763586, |
| "grad_norm": 0.413860559463501, |
| "learning_rate": 6.23749594556226e-05, |
| "loss": 0.5476, |
| "step": 955 |
| }, |
| { |
| "epoch": 1.328141225337487, |
| "grad_norm": 0.39695823192596436, |
| "learning_rate": 6.125735952995775e-05, |
| "loss": 0.5519, |
| "step": 960 |
| }, |
| { |
| "epoch": 1.3350640359986155, |
| "grad_norm": 0.43241792917251587, |
| "learning_rate": 6.0145424675540394e-05, |
| "loss": 0.5495, |
| "step": 965 |
| }, |
| { |
| "epoch": 1.341986846659744, |
| "grad_norm": 0.4030551612377167, |
| "learning_rate": 5.9039317482996735e-05, |
| "loss": 0.5581, |
| "step": 970 |
| }, |
| { |
| "epoch": 1.3489096573208723, |
| "grad_norm": 0.3783753216266632, |
| "learning_rate": 5.793919969081374e-05, |
| "loss": 0.5403, |
| "step": 975 |
| }, |
| { |
| "epoch": 1.3558324679820006, |
| "grad_norm": 0.39069420099258423, |
| "learning_rate": 5.68452321616894e-05, |
| "loss": 0.5404, |
| "step": 980 |
| }, |
| { |
| "epoch": 1.3627552786431292, |
| "grad_norm": 0.38791337609291077, |
| "learning_rate": 5.5757574859010886e-05, |
| "loss": 0.5534, |
| "step": 985 |
| }, |
| { |
| "epoch": 1.3696780893042575, |
| "grad_norm": 0.42228081822395325, |
| "learning_rate": 5.467638682346403e-05, |
| "loss": 0.5383, |
| "step": 990 |
| }, |
| { |
| "epoch": 1.376600899965386, |
| "grad_norm": 0.3954121768474579, |
| "learning_rate": 5.3601826149777966e-05, |
| "loss": 0.5653, |
| "step": 995 |
| }, |
| { |
| "epoch": 1.3835237106265144, |
| "grad_norm": 0.3909887671470642, |
| "learning_rate": 5.25340499636082e-05, |
| "loss": 0.5446, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.3904465212876427, |
| "grad_norm": 0.41362643241882324, |
| "learning_rate": 5.147321439856091e-05, |
| "loss": 0.5466, |
| "step": 1005 |
| }, |
| { |
| "epoch": 1.3973693319487712, |
| "grad_norm": 0.3651193380355835, |
| "learning_rate": 5.041947457336274e-05, |
| "loss": 0.5494, |
| "step": 1010 |
| }, |
| { |
| "epoch": 1.4042921426098995, |
| "grad_norm": 0.36420342326164246, |
| "learning_rate": 4.937298456917889e-05, |
| "loss": 0.5333, |
| "step": 1015 |
| }, |
| { |
| "epoch": 1.411214953271028, |
| "grad_norm": 0.3928079605102539, |
| "learning_rate": 4.8333897407082896e-05, |
| "loss": 0.5445, |
| "step": 1020 |
| }, |
| { |
| "epoch": 1.4181377639321564, |
| "grad_norm": 0.36815693974494934, |
| "learning_rate": 4.7302365025681206e-05, |
| "loss": 0.5457, |
| "step": 1025 |
| }, |
| { |
| "epoch": 1.425060574593285, |
| "grad_norm": 0.3875131607055664, |
| "learning_rate": 4.627853825889642e-05, |
| "loss": 0.5296, |
| "step": 1030 |
| }, |
| { |
| "epoch": 1.4319833852544133, |
| "grad_norm": 0.4003137946128845, |
| "learning_rate": 4.526256681391192e-05, |
| "loss": 0.5455, |
| "step": 1035 |
| }, |
| { |
| "epoch": 1.4389061959155418, |
| "grad_norm": 0.40907323360443115, |
| "learning_rate": 4.4254599249281016e-05, |
| "loss": 0.5277, |
| "step": 1040 |
| }, |
| { |
| "epoch": 1.4458290065766701, |
| "grad_norm": 0.40487226843833923, |
| "learning_rate": 4.3254782953204375e-05, |
| "loss": 0.5337, |
| "step": 1045 |
| }, |
| { |
| "epoch": 1.4527518172377984, |
| "grad_norm": 0.3791086971759796, |
| "learning_rate": 4.2263264121978505e-05, |
| "loss": 0.5394, |
| "step": 1050 |
| }, |
| { |
| "epoch": 1.459674627898927, |
| "grad_norm": 0.4024101197719574, |
| "learning_rate": 4.12801877386183e-05, |
| "loss": 0.529, |
| "step": 1055 |
| }, |
| { |
| "epoch": 1.4665974385600553, |
| "grad_norm": 0.3829149305820465, |
| "learning_rate": 4.030569755165736e-05, |
| "loss": 0.5365, |
| "step": 1060 |
| }, |
| { |
| "epoch": 1.4735202492211839, |
| "grad_norm": 0.38531187176704407, |
| "learning_rate": 3.9339936054128466e-05, |
| "loss": 0.5367, |
| "step": 1065 |
| }, |
| { |
| "epoch": 1.4804430598823122, |
| "grad_norm": 0.39671218395233154, |
| "learning_rate": 3.8383044462727826e-05, |
| "loss": 0.5248, |
| "step": 1070 |
| }, |
| { |
| "epoch": 1.4873658705434407, |
| "grad_norm": 0.3982336223125458, |
| "learning_rate": 3.7435162697166107e-05, |
| "loss": 0.5469, |
| "step": 1075 |
| }, |
| { |
| "epoch": 1.494288681204569, |
| "grad_norm": 0.40101921558380127, |
| "learning_rate": 3.649642935970859e-05, |
| "loss": 0.5461, |
| "step": 1080 |
| }, |
| { |
| "epoch": 1.5012114918656976, |
| "grad_norm": 0.41018062829971313, |
| "learning_rate": 3.556698171490871e-05, |
| "loss": 0.53, |
| "step": 1085 |
| }, |
| { |
| "epoch": 1.508134302526826, |
| "grad_norm": 0.39131489396095276, |
| "learning_rate": 3.464695566953644e-05, |
| "loss": 0.5287, |
| "step": 1090 |
| }, |
| { |
| "epoch": 1.5150571131879542, |
| "grad_norm": 0.3957297205924988, |
| "learning_rate": 3.373648575270576e-05, |
| "loss": 0.5301, |
| "step": 1095 |
| }, |
| { |
| "epoch": 1.5219799238490828, |
| "grad_norm": 0.3676046133041382, |
| "learning_rate": 3.283570509620344e-05, |
| "loss": 0.5332, |
| "step": 1100 |
| }, |
| { |
| "epoch": 1.5289027345102113, |
| "grad_norm": 0.3769654333591461, |
| "learning_rate": 3.1944745415021915e-05, |
| "loss": 0.5352, |
| "step": 1105 |
| }, |
| { |
| "epoch": 1.5358255451713396, |
| "grad_norm": 0.3649960160255432, |
| "learning_rate": 3.106373698809958e-05, |
| "loss": 0.5449, |
| "step": 1110 |
| }, |
| { |
| "epoch": 1.542748355832468, |
| "grad_norm": 0.3833390772342682, |
| "learning_rate": 3.0192808639271065e-05, |
| "loss": 0.5462, |
| "step": 1115 |
| }, |
| { |
| "epoch": 1.5496711664935963, |
| "grad_norm": 0.38520780205726624, |
| "learning_rate": 2.933208771843018e-05, |
| "loss": 0.5326, |
| "step": 1120 |
| }, |
| { |
| "epoch": 1.5565939771547248, |
| "grad_norm": 0.39242011308670044, |
| "learning_rate": 2.8481700082908268e-05, |
| "loss": 0.5275, |
| "step": 1125 |
| }, |
| { |
| "epoch": 1.5635167878158533, |
| "grad_norm": 0.38760125637054443, |
| "learning_rate": 2.7641770079071127e-05, |
| "loss": 0.5294, |
| "step": 1130 |
| }, |
| { |
| "epoch": 1.5704395984769817, |
| "grad_norm": 0.37061071395874023, |
| "learning_rate": 2.681242052413656e-05, |
| "loss": 0.5173, |
| "step": 1135 |
| }, |
| { |
| "epoch": 1.57736240913811, |
| "grad_norm": 0.40858691930770874, |
| "learning_rate": 2.59937726882158e-05, |
| "loss": 0.5149, |
| "step": 1140 |
| }, |
| { |
| "epoch": 1.5842852197992385, |
| "grad_norm": 0.4115959107875824, |
| "learning_rate": 2.5185946276580918e-05, |
| "loss": 0.5395, |
| "step": 1145 |
| }, |
| { |
| "epoch": 1.591208030460367, |
| "grad_norm": 0.3848041296005249, |
| "learning_rate": 2.4389059412161087e-05, |
| "loss": 0.5317, |
| "step": 1150 |
| }, |
| { |
| "epoch": 1.5981308411214954, |
| "grad_norm": 0.38582009077072144, |
| "learning_rate": 2.3603228618270356e-05, |
| "loss": 0.5278, |
| "step": 1155 |
| }, |
| { |
| "epoch": 1.6050536517826237, |
| "grad_norm": 0.38087934255599976, |
| "learning_rate": 2.2828568801569283e-05, |
| "loss": 0.5087, |
| "step": 1160 |
| }, |
| { |
| "epoch": 1.611976462443752, |
| "grad_norm": 0.36079418659210205, |
| "learning_rate": 2.2065193235262782e-05, |
| "loss": 0.5052, |
| "step": 1165 |
| }, |
| { |
| "epoch": 1.6188992731048806, |
| "grad_norm": 0.38548552989959717, |
| "learning_rate": 2.1313213542537058e-05, |
| "loss": 0.5248, |
| "step": 1170 |
| }, |
| { |
| "epoch": 1.6258220837660091, |
| "grad_norm": 0.41799798607826233, |
| "learning_rate": 2.0572739680237717e-05, |
| "loss": 0.5357, |
| "step": 1175 |
| }, |
| { |
| "epoch": 1.6327448944271374, |
| "grad_norm": 0.37560001015663147, |
| "learning_rate": 1.984387992279153e-05, |
| "loss": 0.5364, |
| "step": 1180 |
| }, |
| { |
| "epoch": 1.6396677050882658, |
| "grad_norm": 0.39734479784965515, |
| "learning_rate": 1.912674084637409e-05, |
| "loss": 0.5323, |
| "step": 1185 |
| }, |
| { |
| "epoch": 1.6465905157493943, |
| "grad_norm": 0.3698159158229828, |
| "learning_rate": 1.8421427313326046e-05, |
| "loss": 0.5149, |
| "step": 1190 |
| }, |
| { |
| "epoch": 1.6535133264105226, |
| "grad_norm": 0.3833644986152649, |
| "learning_rate": 1.7728042456819704e-05, |
| "loss": 0.53, |
| "step": 1195 |
| }, |
| { |
| "epoch": 1.6604361370716512, |
| "grad_norm": 0.37530168890953064, |
| "learning_rate": 1.7046687665778717e-05, |
| "loss": 0.5147, |
| "step": 1200 |
| }, |
| { |
| "epoch": 1.6673589477327795, |
| "grad_norm": 0.37578684091567993, |
| "learning_rate": 1.6377462570052438e-05, |
| "loss": 0.521, |
| "step": 1205 |
| }, |
| { |
| "epoch": 1.6742817583939078, |
| "grad_norm": 0.4046749770641327, |
| "learning_rate": 1.5720465025847987e-05, |
| "loss": 0.5208, |
| "step": 1210 |
| }, |
| { |
| "epoch": 1.6812045690550363, |
| "grad_norm": 0.3772706091403961, |
| "learning_rate": 1.5075791101421166e-05, |
| "loss": 0.5204, |
| "step": 1215 |
| }, |
| { |
| "epoch": 1.6881273797161649, |
| "grad_norm": 0.37441906332969666, |
| "learning_rate": 1.4443535063029279e-05, |
| "loss": 0.5341, |
| "step": 1220 |
| }, |
| { |
| "epoch": 1.6950501903772932, |
| "grad_norm": 0.39095550775527954, |
| "learning_rate": 1.3823789361147044e-05, |
| "loss": 0.5111, |
| "step": 1225 |
| }, |
| { |
| "epoch": 1.7019730010384215, |
| "grad_norm": 0.4048159122467041, |
| "learning_rate": 1.3216644616948282e-05, |
| "loss": 0.5183, |
| "step": 1230 |
| }, |
| { |
| "epoch": 1.70889581169955, |
| "grad_norm": 0.3774990141391754, |
| "learning_rate": 1.2622189609054979e-05, |
| "loss": 0.5142, |
| "step": 1235 |
| }, |
| { |
| "epoch": 1.7158186223606784, |
| "grad_norm": 0.4018307030200958, |
| "learning_rate": 1.2040511260555831e-05, |
| "loss": 0.5179, |
| "step": 1240 |
| }, |
| { |
| "epoch": 1.722741433021807, |
| "grad_norm": 0.37508299946784973, |
| "learning_rate": 1.1471694626296049e-05, |
| "loss": 0.514, |
| "step": 1245 |
| }, |
| { |
| "epoch": 1.7296642436829353, |
| "grad_norm": 0.3805769979953766, |
| "learning_rate": 1.0915822880440308e-05, |
| "loss": 0.516, |
| "step": 1250 |
| }, |
| { |
| "epoch": 1.7365870543440636, |
| "grad_norm": 0.37996870279312134, |
| "learning_rate": 1.037297730431085e-05, |
| "loss": 0.5105, |
| "step": 1255 |
| }, |
| { |
| "epoch": 1.7435098650051921, |
| "grad_norm": 0.3757014572620392, |
| "learning_rate": 9.843237274502282e-06, |
| "loss": 0.5176, |
| "step": 1260 |
| }, |
| { |
| "epoch": 1.7504326756663207, |
| "grad_norm": 0.39681464433670044, |
| "learning_rate": 9.326680251274777e-06, |
| "loss": 0.527, |
| "step": 1265 |
| }, |
| { |
| "epoch": 1.757355486327449, |
| "grad_norm": 0.3785605728626251, |
| "learning_rate": 8.823381767227667e-06, |
| "loss": 0.5223, |
| "step": 1270 |
| }, |
| { |
| "epoch": 1.7642782969885773, |
| "grad_norm": 0.38478976488113403, |
| "learning_rate": 8.333415416254831e-06, |
| "loss": 0.5255, |
| "step": 1275 |
| }, |
| { |
| "epoch": 1.7712011076497056, |
| "grad_norm": 0.37452688813209534, |
| "learning_rate": 7.856852842783547e-06, |
| "loss": 0.5141, |
| "step": 1280 |
| }, |
| { |
| "epoch": 1.7781239183108342, |
| "grad_norm": 0.3667319118976593, |
| "learning_rate": 7.393763731298387e-06, |
| "loss": 0.5056, |
| "step": 1285 |
| }, |
| { |
| "epoch": 1.7850467289719627, |
| "grad_norm": 0.36586126685142517, |
| "learning_rate": 6.944215796151765e-06, |
| "loss": 0.5108, |
| "step": 1290 |
| }, |
| { |
| "epoch": 1.791969539633091, |
| "grad_norm": 0.407480388879776, |
| "learning_rate": 6.5082747716625255e-06, |
| "loss": 0.5168, |
| "step": 1295 |
| }, |
| { |
| "epoch": 1.7988923502942193, |
| "grad_norm": 0.3544744551181793, |
| "learning_rate": 6.086004402504098e-06, |
| "loss": 0.5044, |
| "step": 1300 |
| }, |
| { |
| "epoch": 1.8058151609553479, |
| "grad_norm": 0.39020371437072754, |
| "learning_rate": 5.6774664343834495e-06, |
| "loss": 0.5248, |
| "step": 1305 |
| }, |
| { |
| "epoch": 1.8127379716164764, |
| "grad_norm": 0.38111376762390137, |
| "learning_rate": 5.2827206050125585e-06, |
| "loss": 0.5151, |
| "step": 1310 |
| }, |
| { |
| "epoch": 1.8196607822776047, |
| "grad_norm": 0.38476815819740295, |
| "learning_rate": 4.9018246353732755e-06, |
| "loss": 0.533, |
| "step": 1315 |
| }, |
| { |
| "epoch": 1.826583592938733, |
| "grad_norm": 0.3769124746322632, |
| "learning_rate": 4.534834221277262e-06, |
| "loss": 0.5047, |
| "step": 1320 |
| }, |
| { |
| "epoch": 1.8335064035998614, |
| "grad_norm": 0.3822435140609741, |
| "learning_rate": 4.181803025221898e-06, |
| "loss": 0.5212, |
| "step": 1325 |
| }, |
| { |
| "epoch": 1.84042921426099, |
| "grad_norm": 0.3835265338420868, |
| "learning_rate": 3.842782668543599e-06, |
| "loss": 0.515, |
| "step": 1330 |
| }, |
| { |
| "epoch": 1.8473520249221185, |
| "grad_norm": 0.34277990460395813, |
| "learning_rate": 3.5178227238696347e-06, |
| "loss": 0.5097, |
| "step": 1335 |
| }, |
| { |
| "epoch": 1.8542748355832468, |
| "grad_norm": 0.41254210472106934, |
| "learning_rate": 3.2069707078694057e-06, |
| "loss": 0.4997, |
| "step": 1340 |
| }, |
| { |
| "epoch": 1.8611976462443751, |
| "grad_norm": 0.38151687383651733, |
| "learning_rate": 2.9102720743064527e-06, |
| "loss": 0.5253, |
| "step": 1345 |
| }, |
| { |
| "epoch": 1.8681204569055037, |
| "grad_norm": 0.3614235520362854, |
| "learning_rate": 2.6277702073919997e-06, |
| "loss": 0.5022, |
| "step": 1350 |
| }, |
| { |
| "epoch": 1.8750432675666322, |
| "grad_norm": 0.37769678235054016, |
| "learning_rate": 2.3595064154412374e-06, |
| "loss": 0.5083, |
| "step": 1355 |
| }, |
| { |
| "epoch": 1.8819660782277605, |
| "grad_norm": 0.3952181041240692, |
| "learning_rate": 2.1055199248330546e-06, |
| "loss": 0.5093, |
| "step": 1360 |
| }, |
| { |
| "epoch": 1.8888888888888888, |
| "grad_norm": 0.3974086344242096, |
| "learning_rate": 1.8658478742742625e-06, |
| "loss": 0.5087, |
| "step": 1365 |
| }, |
| { |
| "epoch": 1.8958116995500172, |
| "grad_norm": 0.38178950548171997, |
| "learning_rate": 1.6405253093690343e-06, |
| "loss": 0.5072, |
| "step": 1370 |
| }, |
| { |
| "epoch": 1.9027345102111457, |
| "grad_norm": 0.39221471548080444, |
| "learning_rate": 1.429585177494419e-06, |
| "loss": 0.504, |
| "step": 1375 |
| }, |
| { |
| "epoch": 1.9096573208722742, |
| "grad_norm": 0.37083545327186584, |
| "learning_rate": 1.2330583229827053e-06, |
| "loss": 0.5139, |
| "step": 1380 |
| }, |
| { |
| "epoch": 1.9165801315334026, |
| "grad_norm": 0.3701615631580353, |
| "learning_rate": 1.0509734826112394e-06, |
| "loss": 0.5023, |
| "step": 1385 |
| }, |
| { |
| "epoch": 1.9235029421945309, |
| "grad_norm": 0.40432795882225037, |
| "learning_rate": 8.833572814004321e-07, |
| "loss": 0.5026, |
| "step": 1390 |
| }, |
| { |
| "epoch": 1.9304257528556594, |
| "grad_norm": 0.3880690634250641, |
| "learning_rate": 7.302342287205943e-07, |
| "loss": 0.5042, |
| "step": 1395 |
| }, |
| { |
| "epoch": 1.937348563516788, |
| "grad_norm": 0.38099151849746704, |
| "learning_rate": 5.916267147080934e-07, |
| "loss": 0.5098, |
| "step": 1400 |
| }, |
| { |
| "epoch": 1.9442713741779163, |
| "grad_norm": 0.3906520903110504, |
| "learning_rate": 4.6755500699138253e-07, |
| "loss": 0.5039, |
| "step": 1405 |
| }, |
| { |
| "epoch": 1.9511941848390446, |
| "grad_norm": 0.39573198556900024, |
| "learning_rate": 3.5803724772740523e-07, |
| "loss": 0.505, |
| "step": 1410 |
| }, |
| { |
| "epoch": 1.958116995500173, |
| "grad_norm": 0.3726860284805298, |
| "learning_rate": 2.630894509488058e-07, |
| "loss": 0.5088, |
| "step": 1415 |
| }, |
| { |
| "epoch": 1.9650398061613015, |
| "grad_norm": 0.35756033658981323, |
| "learning_rate": 1.8272550022230272e-07, |
| "loss": 0.506, |
| "step": 1420 |
| }, |
| { |
| "epoch": 1.97196261682243, |
| "grad_norm": 0.3828778862953186, |
| "learning_rate": 1.1695714661856816e-07, |
| "loss": 0.494, |
| "step": 1425 |
| }, |
| { |
| "epoch": 1.9788854274835583, |
| "grad_norm": 0.3731125593185425, |
| "learning_rate": 6.579400699397998e-08, |
| "loss": 0.5106, |
| "step": 1430 |
| }, |
| { |
| "epoch": 1.9858082381446867, |
| "grad_norm": 0.3635178804397583, |
| "learning_rate": 2.9243562584402305e-08, |
| "loss": 0.517, |
| "step": 1435 |
| }, |
| { |
| "epoch": 1.9927310488058152, |
| "grad_norm": 0.38713252544403076, |
| "learning_rate": 7.311157911249478e-09, |
| "loss": 0.5221, |
| "step": 1440 |
| }, |
| { |
| "epoch": 1.9982692973347178, |
| "eval_loss": 0.7301178574562073, |
| "eval_runtime": 60.039, |
| "eval_samples_per_second": 34.461, |
| "eval_steps_per_second": 2.165, |
| "step": 1444 |
| }, |
| { |
| "epoch": 1.9982692973347178, |
| "step": 1444, |
| "total_flos": 4.774796461726499e+18, |
| "train_loss": 0.735994421048838, |
| "train_runtime": 11345.701, |
| "train_samples_per_second": 8.147, |
| "train_steps_per_second": 0.127 |
| } |
| ], |
| "logging_steps": 5, |
| "max_steps": 1444, |
| "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": false, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 4.774796461726499e+18, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
