| { | |
| "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": 56.73267364501953, | |
| "learning_rate": 0.0, | |
| "loss": 3.259, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006922810661128418, | |
| "grad_norm": 24.641572952270508, | |
| "learning_rate": 5.517241379310345e-06, | |
| "loss": 3.0823, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.013845621322256836, | |
| "grad_norm": 6.519947052001953, | |
| "learning_rate": 1.2413793103448277e-05, | |
| "loss": 2.2994, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.020768431983385256, | |
| "grad_norm": 8.049466133117676, | |
| "learning_rate": 1.9310344827586207e-05, | |
| "loss": 1.8455, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.027691242644513673, | |
| "grad_norm": 10.51276683807373, | |
| "learning_rate": 2.620689655172414e-05, | |
| "loss": 1.5757, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.034614053305642094, | |
| "grad_norm": 8.093124389648438, | |
| "learning_rate": 3.310344827586207e-05, | |
| "loss": 1.3246, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04153686396677051, | |
| "grad_norm": 4.431617259979248, | |
| "learning_rate": 4e-05, | |
| "loss": 1.0413, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.04845967462789893, | |
| "grad_norm": 2.7962491512298584, | |
| "learning_rate": 4.689655172413793e-05, | |
| "loss": 0.9326, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.055382485289027346, | |
| "grad_norm": 3.9973249435424805, | |
| "learning_rate": 5.379310344827586e-05, | |
| "loss": 0.8863, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06230529595015576, | |
| "grad_norm": 2.9854631423950195, | |
| "learning_rate": 6.068965517241379e-05, | |
| "loss": 0.8081, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.06922810661128419, | |
| "grad_norm": 2.09420108795166, | |
| "learning_rate": 6.758620689655173e-05, | |
| "loss": 0.6819, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0761509172724126, | |
| "grad_norm": 1.8408433198928833, | |
| "learning_rate": 7.448275862068966e-05, | |
| "loss": 0.6754, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.08307372793354102, | |
| "grad_norm": 3.799906015396118, | |
| "learning_rate": 8.137931034482759e-05, | |
| "loss": 0.6466, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.08999653859466944, | |
| "grad_norm": 1.5792860984802246, | |
| "learning_rate": 8.827586206896552e-05, | |
| "loss": 0.6622, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.09691934925579786, | |
| "grad_norm": 4.7252607345581055, | |
| "learning_rate": 9.517241379310345e-05, | |
| "loss": 0.6872, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.10384215991692627, | |
| "grad_norm": 1.8571418523788452, | |
| "learning_rate": 0.0001020689655172414, | |
| "loss": 0.669, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.11076497057805469, | |
| "grad_norm": 1.1482609510421753, | |
| "learning_rate": 0.00010896551724137931, | |
| "loss": 0.6514, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.11768778123918311, | |
| "grad_norm": 1.3610243797302246, | |
| "learning_rate": 0.00011586206896551725, | |
| "loss": 0.7626, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.12461059190031153, | |
| "grad_norm": 0.9764773845672607, | |
| "learning_rate": 0.00012275862068965518, | |
| "loss": 0.681, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.13153340256143994, | |
| "grad_norm": 0.8563634157180786, | |
| "learning_rate": 0.0001296551724137931, | |
| "loss": 0.6547, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.13845621322256838, | |
| "grad_norm": 0.8853511810302734, | |
| "learning_rate": 0.00013655172413793104, | |
| "loss": 0.6481, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.14537902388369678, | |
| "grad_norm": 1.5578516721725464, | |
| "learning_rate": 0.00014344827586206896, | |
| "loss": 0.7995, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.1523018345448252, | |
| "grad_norm": 2.2077314853668213, | |
| "learning_rate": 0.0001503448275862069, | |
| "loss": 0.7294, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1592246452059536, | |
| "grad_norm": 5.0542497634887695, | |
| "learning_rate": 0.00015724137931034485, | |
| "loss": 0.8053, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.16614745586708204, | |
| "grad_norm": 0.7379900217056274, | |
| "learning_rate": 0.00016413793103448276, | |
| "loss": 0.7604, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.17307026652821045, | |
| "grad_norm": 1.4255956411361694, | |
| "learning_rate": 0.0001710344827586207, | |
| "loss": 0.6901, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.17999307718933888, | |
| "grad_norm": 5.76345157623291, | |
| "learning_rate": 0.00017793103448275862, | |
| "loss": 0.6692, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.18691588785046728, | |
| "grad_norm": 1.020735502243042, | |
| "learning_rate": 0.00018482758620689654, | |
| "loss": 0.6952, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.19383869851159571, | |
| "grad_norm": 0.7500247359275818, | |
| "learning_rate": 0.0001917241379310345, | |
| "loss": 0.6808, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.20076150917272412, | |
| "grad_norm": 2.430368661880493, | |
| "learning_rate": 0.00019862068965517243, | |
| "loss": 0.9069, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.20768431983385255, | |
| "grad_norm": 2.4098498821258545, | |
| "learning_rate": 0.0001999953208384105, | |
| "loss": 0.8003, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.21460713049498095, | |
| "grad_norm": 1.0620583295822144, | |
| "learning_rate": 0.0001999763124949378, | |
| "loss": 0.9143, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.22152994115610938, | |
| "grad_norm": 0.8933396935462952, | |
| "learning_rate": 0.0001999426852993044, | |
| "loss": 0.677, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2284527518172378, | |
| "grad_norm": 0.822638213634491, | |
| "learning_rate": 0.00019989444416858502, | |
| "loss": 0.6956, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.23537556247836622, | |
| "grad_norm": 0.6508069634437561, | |
| "learning_rate": 0.0001998315961567502, | |
| "loss": 0.6825, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.24229837313949462, | |
| "grad_norm": 0.7333005666732788, | |
| "learning_rate": 0.00019975415045363467, | |
| "loss": 0.6708, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.24922118380062305, | |
| "grad_norm": 0.6714277863502502, | |
| "learning_rate": 0.00019966211838359378, | |
| "loss": 0.6713, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.25614399446175146, | |
| "grad_norm": 0.6074556708335876, | |
| "learning_rate": 0.00019955551340384743, | |
| "loss": 0.6646, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.2630668051228799, | |
| "grad_norm": 0.6033316254615784, | |
| "learning_rate": 0.00019943435110251248, | |
| "loss": 0.651, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2699896157840083, | |
| "grad_norm": 0.6866341829299927, | |
| "learning_rate": 0.0001992986491963232, | |
| "loss": 0.6596, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.27691242644513675, | |
| "grad_norm": 0.5897858738899231, | |
| "learning_rate": 0.00019914842752804103, | |
| "loss": 0.6524, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2838352371062651, | |
| "grad_norm": 0.7324863076210022, | |
| "learning_rate": 0.00019898370806355263, | |
| "loss": 0.6391, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.29075804776739356, | |
| "grad_norm": 0.6330517530441284, | |
| "learning_rate": 0.00019880451488865836, | |
| "loss": 0.6644, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.297680858428522, | |
| "grad_norm": 0.5360478758811951, | |
| "learning_rate": 0.00019861087420555018, | |
| "loss": 0.6473, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.3046036690896504, | |
| "grad_norm": 0.5072117447853088, | |
| "learning_rate": 0.0001984028143289803, | |
| "loss": 0.6446, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3115264797507788, | |
| "grad_norm": 0.5131970047950745, | |
| "learning_rate": 0.00019818036568212106, | |
| "loss": 0.6189, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.3184492904119072, | |
| "grad_norm": 0.5710281133651733, | |
| "learning_rate": 0.00019794356079211604, | |
| "loss": 0.6583, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.32537210107303566, | |
| "grad_norm": 0.5091654062271118, | |
| "learning_rate": 0.00019769243428532422, | |
| "loss": 0.6252, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.3322949117341641, | |
| "grad_norm": 0.668511688709259, | |
| "learning_rate": 0.00019742702288225652, | |
| "loss": 0.6338, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.33921772239529246, | |
| "grad_norm": 0.4644317328929901, | |
| "learning_rate": 0.00019714736539220648, | |
| "loss": 0.6363, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.3461405330564209, | |
| "grad_norm": 0.5522793531417847, | |
| "learning_rate": 0.00019685350270757555, | |
| "loss": 0.6375, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.3530633437175493, | |
| "grad_norm": 0.5864453315734863, | |
| "learning_rate": 0.0001965454777978936, | |
| "loss": 0.6507, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.35998615437867776, | |
| "grad_norm": 0.6083868145942688, | |
| "learning_rate": 0.00019622333570353567, | |
| "loss": 0.6371, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.36690896503980613, | |
| "grad_norm": 0.5848725438117981, | |
| "learning_rate": 0.00019588712352913625, | |
| "loss": 0.6322, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.37383177570093457, | |
| "grad_norm": 0.5301589965820312, | |
| "learning_rate": 0.00019553689043670127, | |
| "loss": 0.6255, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.380754586362063, | |
| "grad_norm": 0.5005171895027161, | |
| "learning_rate": 0.00019517268763841962, | |
| "loss": 0.6178, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.38767739702319143, | |
| "grad_norm": 0.5215473771095276, | |
| "learning_rate": 0.00019479456838917473, | |
| "loss": 0.6241, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.39460020768431986, | |
| "grad_norm": 0.42020273208618164, | |
| "learning_rate": 0.0001944025879787574, | |
| "loss": 0.6463, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.40152301834544823, | |
| "grad_norm": 0.5007926225662231, | |
| "learning_rate": 0.0001939968037237812, | |
| "loss": 0.6213, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.40844582900657667, | |
| "grad_norm": 0.5666825771331787, | |
| "learning_rate": 0.00019357727495930146, | |
| "loss": 0.6247, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.4153686396677051, | |
| "grad_norm": 0.5147333741188049, | |
| "learning_rate": 0.00019314406303013904, | |
| "loss": 0.6188, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.42229145032883353, | |
| "grad_norm": 0.560345470905304, | |
| "learning_rate": 0.00019269723128191048, | |
| "loss": 0.6211, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.4292142609899619, | |
| "grad_norm": 0.471113383769989, | |
| "learning_rate": 0.00019223684505176517, | |
| "loss": 0.6153, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.43613707165109034, | |
| "grad_norm": 0.4909154772758484, | |
| "learning_rate": 0.00019176297165883165, | |
| "loss": 0.6229, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.44305988231221877, | |
| "grad_norm": 0.479992151260376, | |
| "learning_rate": 0.00019127568039437406, | |
| "loss": 0.6243, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.4499826929733472, | |
| "grad_norm": 0.412416934967041, | |
| "learning_rate": 0.00019077504251166006, | |
| "loss": 0.6237, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.4569055036344756, | |
| "grad_norm": 0.4314121603965759, | |
| "learning_rate": 0.00019026113121554204, | |
| "loss": 0.6085, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.463828314295604, | |
| "grad_norm": 0.44777366518974304, | |
| "learning_rate": 0.00018973402165175268, | |
| "loss": 0.6254, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.47075112495673244, | |
| "grad_norm": 0.4374144673347473, | |
| "learning_rate": 0.0001891937908959172, | |
| "loss": 0.628, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.47767393561786087, | |
| "grad_norm": 0.45410701632499695, | |
| "learning_rate": 0.00018864051794228282, | |
| "loss": 0.6226, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.48459674627898924, | |
| "grad_norm": 0.41423413157463074, | |
| "learning_rate": 0.00018807428369216822, | |
| "loss": 0.6085, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.4915195569401177, | |
| "grad_norm": 0.421219140291214, | |
| "learning_rate": 0.00018749517094213375, | |
| "loss": 0.6137, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.4984423676012461, | |
| "grad_norm": 0.4083799421787262, | |
| "learning_rate": 0.00018690326437187472, | |
| "loss": 0.6151, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5053651782623745, | |
| "grad_norm": 0.4138510525226593, | |
| "learning_rate": 0.00018629865053183911, | |
| "loss": 0.6114, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.5122879889235029, | |
| "grad_norm": 0.536926805973053, | |
| "learning_rate": 0.0001856814178305722, | |
| "loss": 0.6186, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5192107995846313, | |
| "grad_norm": 0.4805965721607208, | |
| "learning_rate": 0.00018505165652178893, | |
| "loss": 0.6262, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.5261336102457598, | |
| "grad_norm": 0.47122180461883545, | |
| "learning_rate": 0.00018440945869117675, | |
| "loss": 0.6025, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5330564209068882, | |
| "grad_norm": 0.43172767758369446, | |
| "learning_rate": 0.00018375491824293067, | |
| "loss": 0.6044, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.5399792315680166, | |
| "grad_norm": 0.403818815946579, | |
| "learning_rate": 0.00018308813088602227, | |
| "loss": 0.6218, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5469020422291451, | |
| "grad_norm": 0.4400286078453064, | |
| "learning_rate": 0.00018240919412020466, | |
| "loss": 0.6045, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.5538248528902735, | |
| "grad_norm": 0.40739211440086365, | |
| "learning_rate": 0.00018171820722175604, | |
| "loss": 0.6014, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5607476635514018, | |
| "grad_norm": 0.4094732701778412, | |
| "learning_rate": 0.00018101527122896293, | |
| "loss": 0.6075, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.5676704742125303, | |
| "grad_norm": 0.43222758173942566, | |
| "learning_rate": 0.0001803004889273463, | |
| "loss": 0.5989, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.5745932848736587, | |
| "grad_norm": 0.43636542558670044, | |
| "learning_rate": 0.0001795739648346317, | |
| "loss": 0.6083, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.5815160955347871, | |
| "grad_norm": 0.37528863549232483, | |
| "learning_rate": 0.00017883580518546647, | |
| "loss": 0.5777, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5884389061959155, | |
| "grad_norm": 0.3777056038379669, | |
| "learning_rate": 0.00017808611791588584, | |
| "loss": 0.6198, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.595361716857044, | |
| "grad_norm": 0.4150828421115875, | |
| "learning_rate": 0.00017732501264752996, | |
| "loss": 0.605, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6022845275181724, | |
| "grad_norm": 0.4109332263469696, | |
| "learning_rate": 0.00017655260067161503, | |
| "loss": 0.6124, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.6092073381793008, | |
| "grad_norm": 0.43217596411705017, | |
| "learning_rate": 0.00017576899493265954, | |
| "loss": 0.6128, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6161301488404292, | |
| "grad_norm": 0.4384288787841797, | |
| "learning_rate": 0.00017497431001196943, | |
| "loss": 0.5949, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.6230529595015576, | |
| "grad_norm": 0.4222124218940735, | |
| "learning_rate": 0.00017416866211088366, | |
| "loss": 0.6052, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.629975770162686, | |
| "grad_norm": 0.35749831795692444, | |
| "learning_rate": 0.00017335216903378267, | |
| "loss": 0.6032, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.6368985808238145, | |
| "grad_norm": 0.39720165729522705, | |
| "learning_rate": 0.00017252495017086296, | |
| "loss": 0.6051, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6438213914849429, | |
| "grad_norm": 0.4027459919452667, | |
| "learning_rate": 0.0001716871264806791, | |
| "loss": 0.5841, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.6507442021460713, | |
| "grad_norm": 0.4314420223236084, | |
| "learning_rate": 0.0001708388204724572, | |
| "loss": 0.6025, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.6576670128071997, | |
| "grad_norm": 0.4330042600631714, | |
| "learning_rate": 0.00016998015618818087, | |
| "loss": 0.5924, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.6645898234683282, | |
| "grad_norm": 0.4530496895313263, | |
| "learning_rate": 0.00016911125918445363, | |
| "loss": 0.6012, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.6715126341294566, | |
| "grad_norm": 0.3897707462310791, | |
| "learning_rate": 0.00016823225651413953, | |
| "loss": 0.6163, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.6784354447905849, | |
| "grad_norm": 0.39049839973449707, | |
| "learning_rate": 0.00016734327670778508, | |
| "loss": 0.5911, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.6853582554517134, | |
| "grad_norm": 0.4112708270549774, | |
| "learning_rate": 0.00016644444975482523, | |
| "loss": 0.6085, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.6922810661128418, | |
| "grad_norm": 0.3853406608104706, | |
| "learning_rate": 0.0001655359070845757, | |
| "loss": 0.6007, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6992038767739702, | |
| "grad_norm": 0.4462832510471344, | |
| "learning_rate": 0.0001646177815470151, | |
| "loss": 0.589, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.7061266874350987, | |
| "grad_norm": 0.41017070412635803, | |
| "learning_rate": 0.00016369020739335924, | |
| "loss": 0.5842, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.7130494980962271, | |
| "grad_norm": 0.38715627789497375, | |
| "learning_rate": 0.00016275332025643028, | |
| "loss": 0.5926, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.7199723087573555, | |
| "grad_norm": 0.41574394702911377, | |
| "learning_rate": 0.00016180725713082434, | |
| "loss": 0.5906, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.726895119418484, | |
| "grad_norm": 0.4102523624897003, | |
| "learning_rate": 0.0001608521563528795, | |
| "loss": 0.5984, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.7338179300796123, | |
| "grad_norm": 0.36808186769485474, | |
| "learning_rate": 0.00015988815758044792, | |
| "loss": 0.5802, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 0.39308398962020874, | |
| "learning_rate": 0.00015891540177247467, | |
| "loss": 0.6013, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.7476635514018691, | |
| "grad_norm": 0.41063180565834045, | |
| "learning_rate": 0.00015793403116838613, | |
| "loss": 0.581, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.7545863620629976, | |
| "grad_norm": 0.4476175904273987, | |
| "learning_rate": 0.00015694418926729146, | |
| "loss": 0.5974, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.761509172724126, | |
| "grad_norm": 0.4133051037788391, | |
| "learning_rate": 0.00015594602080699956, | |
| "loss": 0.5854, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.7684319833852544, | |
| "grad_norm": 0.3670900762081146, | |
| "learning_rate": 0.00015493967174285512, | |
| "loss": 0.5942, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.7753547940463829, | |
| "grad_norm": 0.34682410955429077, | |
| "learning_rate": 0.00015392528922639662, | |
| "loss": 0.5875, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.7822776047075113, | |
| "grad_norm": 0.3749230206012726, | |
| "learning_rate": 0.00015290302158383923, | |
| "loss": 0.57, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.7892004153686397, | |
| "grad_norm": 0.34742382168769836, | |
| "learning_rate": 0.0001518730182943862, | |
| "loss": 0.5802, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.796123226029768, | |
| "grad_norm": 0.4164738655090332, | |
| "learning_rate": 0.0001508354299683715, | |
| "loss": 0.5788, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.8030460366908965, | |
| "grad_norm": 0.4050043523311615, | |
| "learning_rate": 0.0001497904083252371, | |
| "loss": 0.5794, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.8099688473520249, | |
| "grad_norm": 0.3878714144229889, | |
| "learning_rate": 0.00014873810617134806, | |
| "loss": 0.5936, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.8168916580131533, | |
| "grad_norm": 0.41179347038269043, | |
| "learning_rate": 0.00014767867737764876, | |
| "loss": 0.5978, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.8238144686742818, | |
| "grad_norm": 0.3596123158931732, | |
| "learning_rate": 0.0001466122768571633, | |
| "loss": 0.5665, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.8307372793354102, | |
| "grad_norm": 0.4059533178806305, | |
| "learning_rate": 0.0001455390605423437, | |
| "loss": 0.584, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8376600899965386, | |
| "grad_norm": 0.36474597454071045, | |
| "learning_rate": 0.00014445918536226904, | |
| "loss": 0.5634, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.8445829006576671, | |
| "grad_norm": 0.3671046197414398, | |
| "learning_rate": 0.00014337280921969858, | |
| "loss": 0.5996, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.8515057113187954, | |
| "grad_norm": 0.35552161931991577, | |
| "learning_rate": 0.0001422800909679829, | |
| "loss": 0.583, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.8584285219799238, | |
| "grad_norm": 0.4014418423175812, | |
| "learning_rate": 0.00014118119038783588, | |
| "loss": 0.5725, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.8653513326410522, | |
| "grad_norm": 0.4207477569580078, | |
| "learning_rate": 0.00014007626816397087, | |
| "loss": 0.5833, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.8722741433021807, | |
| "grad_norm": 0.4331388473510742, | |
| "learning_rate": 0.00013896548586160503, | |
| "loss": 0.6051, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.8791969539633091, | |
| "grad_norm": 0.41242560744285583, | |
| "learning_rate": 0.00013784900590283473, | |
| "loss": 0.5828, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.8861197646244375, | |
| "grad_norm": 0.39160752296447754, | |
| "learning_rate": 0.00013672699154288569, | |
| "loss": 0.5708, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.893042575285566, | |
| "grad_norm": 0.36544379591941833, | |
| "learning_rate": 0.00013559960684624112, | |
| "loss": 0.5789, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.8999653859466944, | |
| "grad_norm": 0.412986695766449, | |
| "learning_rate": 0.0001344670166626519, | |
| "loss": 0.5875, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.9068881966078228, | |
| "grad_norm": 0.38756757974624634, | |
| "learning_rate": 0.0001333293866030317, | |
| "loss": 0.5902, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.9138110072689511, | |
| "grad_norm": 0.33859777450561523, | |
| "learning_rate": 0.0001321868830152407, | |
| "loss": 0.5738, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.9207338179300796, | |
| "grad_norm": 0.3452150821685791, | |
| "learning_rate": 0.00013103967295976179, | |
| "loss": 0.5674, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.927656628591208, | |
| "grad_norm": 0.3356856405735016, | |
| "learning_rate": 0.0001298879241852724, | |
| "loss": 0.5751, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.9345794392523364, | |
| "grad_norm": 0.40660151839256287, | |
| "learning_rate": 0.0001287318051041159, | |
| "loss": 0.5652, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.9415022499134649, | |
| "grad_norm": 0.38893893361091614, | |
| "learning_rate": 0.00012757148476767553, | |
| "loss": 0.5683, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.9484250605745933, | |
| "grad_norm": 0.3641612231731415, | |
| "learning_rate": 0.0001264071328416555, | |
| "loss": 0.5664, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.9553478712357217, | |
| "grad_norm": 0.3656536936759949, | |
| "learning_rate": 0.00012523891958127168, | |
| "loss": 0.5906, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.9622706818968502, | |
| "grad_norm": 0.3721097409725189, | |
| "learning_rate": 0.0001240670158063565, | |
| "loss": 0.5774, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.9691934925579785, | |
| "grad_norm": 0.3411293923854828, | |
| "learning_rate": 0.00012289159287638105, | |
| "loss": 0.5708, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9761163032191069, | |
| "grad_norm": 0.31322574615478516, | |
| "learning_rate": 0.00012171282266539845, | |
| "loss": 0.5821, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.9830391138802353, | |
| "grad_norm": 0.38767164945602417, | |
| "learning_rate": 0.00012053087753691172, | |
| "loss": 0.5712, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.9899619245413638, | |
| "grad_norm": 0.33209285140037537, | |
| "learning_rate": 0.00011934593031867038, | |
| "loss": 0.5774, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.9968847352024922, | |
| "grad_norm": 0.3377987742424011, | |
| "learning_rate": 0.00011815815427739911, | |
| "loss": 0.5676, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.5649709105491638, | |
| "eval_runtime": 83.7527, | |
| "eval_samples_per_second": 24.704, | |
| "eval_steps_per_second": 1.552, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.0027691242644514, | |
| "grad_norm": 0.3240108788013458, | |
| "learning_rate": 0.00011696772309346182, | |
| "loss": 0.5161, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.0096919349255797, | |
| "grad_norm": 0.36670801043510437, | |
| "learning_rate": 0.0001157748108354659, | |
| "loss": 0.4521, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.0166147455867083, | |
| "grad_norm": 0.33711522817611694, | |
| "learning_rate": 0.00011457959193480925, | |
| "loss": 0.4438, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.0235375562478366, | |
| "grad_norm": 0.3467736840248108, | |
| "learning_rate": 0.00011338224116017423, | |
| "loss": 0.4457, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.0304603669089651, | |
| "grad_norm": 0.3616083264350891, | |
| "learning_rate": 0.00011218293359197267, | |
| "loss": 0.4538, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.0373831775700935, | |
| "grad_norm": 0.35638511180877686, | |
| "learning_rate": 0.00011098184459674484, | |
| "loss": 0.4393, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.0443059882312218, | |
| "grad_norm": 0.34925246238708496, | |
| "learning_rate": 0.00010977914980151691, | |
| "loss": 0.4419, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.0512287988923503, | |
| "grad_norm": 0.35654428601264954, | |
| "learning_rate": 0.00010857502506812029, | |
| "loss": 0.42, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.0581516095534786, | |
| "grad_norm": 0.35899755358695984, | |
| "learning_rate": 0.00010736964646747629, | |
| "loss": 0.4179, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.0650744202146072, | |
| "grad_norm": 0.42657214403152466, | |
| "learning_rate": 0.00010616319025385089, | |
| "loss": 0.4338, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.0719972308757355, | |
| "grad_norm": 0.36595892906188965, | |
| "learning_rate": 0.00010495583283908177, | |
| "loss": 0.4431, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.078920041536864, | |
| "grad_norm": 0.31097301840782166, | |
| "learning_rate": 0.00010374775076678332, | |
| "loss": 0.4287, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.0858428521979924, | |
| "grad_norm": 0.3285866975784302, | |
| "learning_rate": 0.00010253912068653146, | |
| "loss": 0.4394, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.092765662859121, | |
| "grad_norm": 0.3313821852207184, | |
| "learning_rate": 0.00010133011932803378, | |
| "loss": 0.4487, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.0996884735202492, | |
| "grad_norm": 0.34151381254196167, | |
| "learning_rate": 0.00010012092347528705, | |
| "loss": 0.4448, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.1066112841813776, | |
| "grad_norm": 0.33228418231010437, | |
| "learning_rate": 9.891170994072793e-05, | |
| "loss": 0.4353, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.113534094842506, | |
| "grad_norm": 0.3602267801761627, | |
| "learning_rate": 9.770265553937831e-05, | |
| "loss": 0.433, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.1204569055036344, | |
| "grad_norm": 0.35099637508392334, | |
| "learning_rate": 9.649393706299127e-05, | |
| "loss": 0.4328, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.127379716164763, | |
| "grad_norm": 0.35660460591316223, | |
| "learning_rate": 9.52857312541998e-05, | |
| "loss": 0.4264, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.1343025268258913, | |
| "grad_norm": 0.31822940707206726, | |
| "learning_rate": 9.407821478067304e-05, | |
| "loss": 0.4273, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.1412253374870198, | |
| "grad_norm": 0.3330586552619934, | |
| "learning_rate": 9.287156420928342e-05, | |
| "loss": 0.4224, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.1481481481481481, | |
| "grad_norm": 0.3414239287376404, | |
| "learning_rate": 9.166595598028832e-05, | |
| "loss": 0.4453, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.1550709588092767, | |
| "grad_norm": 0.3497198224067688, | |
| "learning_rate": 9.046156638153056e-05, | |
| "loss": 0.4242, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.161993769470405, | |
| "grad_norm": 0.328546404838562, | |
| "learning_rate": 8.925857152266103e-05, | |
| "loss": 0.4186, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.1689165801315333, | |
| "grad_norm": 0.32587334513664246, | |
| "learning_rate": 8.805714730938728e-05, | |
| "loss": 0.4271, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.1758393907926619, | |
| "grad_norm": 0.3029276728630066, | |
| "learning_rate": 8.685746941775219e-05, | |
| "loss": 0.4271, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.1827622014537902, | |
| "grad_norm": 0.31646841764450073, | |
| "learning_rate": 8.565971326844584e-05, | |
| "loss": 0.4355, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.1896850121149187, | |
| "grad_norm": 0.312472939491272, | |
| "learning_rate": 8.44640540011553e-05, | |
| "loss": 0.4406, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.196607822776047, | |
| "grad_norm": 0.31264108419418335, | |
| "learning_rate": 8.327066644895469e-05, | |
| "loss": 0.4294, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.2035306334371756, | |
| "grad_norm": 0.33994224667549133, | |
| "learning_rate": 8.207972511274095e-05, | |
| "loss": 0.4425, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.210453444098304, | |
| "grad_norm": 0.318868488073349, | |
| "learning_rate": 8.089140413571747e-05, | |
| "loss": 0.4324, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.2173762547594325, | |
| "grad_norm": 0.3303157687187195, | |
| "learning_rate": 7.970587727793051e-05, | |
| "loss": 0.4224, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.2242990654205608, | |
| "grad_norm": 0.33076873421669006, | |
| "learning_rate": 7.852331789086135e-05, | |
| "loss": 0.4246, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.231221876081689, | |
| "grad_norm": 0.3209329843521118, | |
| "learning_rate": 7.73438988920784e-05, | |
| "loss": 0.4316, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.2381446867428176, | |
| "grad_norm": 0.3438841700553894, | |
| "learning_rate": 7.616779273995252e-05, | |
| "loss": 0.4346, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.245067497403946, | |
| "grad_norm": 0.31911325454711914, | |
| "learning_rate": 7.499517140843968e-05, | |
| "loss": 0.4337, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.2519903080650745, | |
| "grad_norm": 0.31535622477531433, | |
| "learning_rate": 7.382620636193438e-05, | |
| "loss": 0.418, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.2589131187262028, | |
| "grad_norm": 0.3386867046356201, | |
| "learning_rate": 7.266106853019759e-05, | |
| "loss": 0.4251, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.2658359293873311, | |
| "grad_norm": 0.33657196164131165, | |
| "learning_rate": 7.149992828336274e-05, | |
| "loss": 0.4344, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.2727587400484597, | |
| "grad_norm": 0.3212815225124359, | |
| "learning_rate": 7.034295540702397e-05, | |
| "loss": 0.4234, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.2796815507095882, | |
| "grad_norm": 0.3345280587673187, | |
| "learning_rate": 6.919031907740922e-05, | |
| "loss": 0.4313, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.2866043613707165, | |
| "grad_norm": 0.3357923626899719, | |
| "learning_rate": 6.804218783664288e-05, | |
| "loss": 0.4347, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.2935271720318449, | |
| "grad_norm": 0.33170652389526367, | |
| "learning_rate": 6.689872956810102e-05, | |
| "loss": 0.4284, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 1.3004499826929734, | |
| "grad_norm": 0.3165716826915741, | |
| "learning_rate": 6.5760111471863e-05, | |
| "loss": 0.4259, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.3073727933541017, | |
| "grad_norm": 0.31603774428367615, | |
| "learning_rate": 6.46265000402628e-05, | |
| "loss": 0.4293, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.3142956040152303, | |
| "grad_norm": 0.3546189069747925, | |
| "learning_rate": 6.349806103354417e-05, | |
| "loss": 0.4286, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.3212184146763586, | |
| "grad_norm": 0.32337120175361633, | |
| "learning_rate": 6.23749594556226e-05, | |
| "loss": 0.422, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.328141225337487, | |
| "grad_norm": 0.32635101675987244, | |
| "learning_rate": 6.125735952995775e-05, | |
| "loss": 0.4151, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.3350640359986155, | |
| "grad_norm": 0.3375142812728882, | |
| "learning_rate": 6.0145424675540394e-05, | |
| "loss": 0.4225, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 1.341986846659744, | |
| "grad_norm": 0.32950544357299805, | |
| "learning_rate": 5.9039317482996735e-05, | |
| "loss": 0.4299, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.3489096573208723, | |
| "grad_norm": 0.3233799934387207, | |
| "learning_rate": 5.793919969081374e-05, | |
| "loss": 0.4122, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.3558324679820006, | |
| "grad_norm": 0.3028440475463867, | |
| "learning_rate": 5.68452321616894e-05, | |
| "loss": 0.4137, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.3627552786431292, | |
| "grad_norm": 0.31653568148612976, | |
| "learning_rate": 5.5757574859010886e-05, | |
| "loss": 0.4187, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 1.3696780893042575, | |
| "grad_norm": 0.324060320854187, | |
| "learning_rate": 5.467638682346403e-05, | |
| "loss": 0.4033, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.376600899965386, | |
| "grad_norm": 0.32063546776771545, | |
| "learning_rate": 5.3601826149777966e-05, | |
| "loss": 0.4215, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 1.3835237106265144, | |
| "grad_norm": 0.3288467228412628, | |
| "learning_rate": 5.25340499636082e-05, | |
| "loss": 0.4137, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.3904465212876427, | |
| "grad_norm": 0.31894204020500183, | |
| "learning_rate": 5.147321439856091e-05, | |
| "loss": 0.42, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 1.3973693319487712, | |
| "grad_norm": 0.316155344247818, | |
| "learning_rate": 5.041947457336274e-05, | |
| "loss": 0.4327, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.4042921426098995, | |
| "grad_norm": 0.32577601075172424, | |
| "learning_rate": 4.937298456917889e-05, | |
| "loss": 0.4226, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 1.411214953271028, | |
| "grad_norm": 0.3630223870277405, | |
| "learning_rate": 4.8333897407082896e-05, | |
| "loss": 0.4174, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.4181377639321564, | |
| "grad_norm": 0.29172050952911377, | |
| "learning_rate": 4.7302365025681206e-05, | |
| "loss": 0.4204, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.425060574593285, | |
| "grad_norm": 0.32107263803482056, | |
| "learning_rate": 4.627853825889642e-05, | |
| "loss": 0.419, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.4319833852544133, | |
| "grad_norm": 0.32417726516723633, | |
| "learning_rate": 4.526256681391192e-05, | |
| "loss": 0.4096, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 1.4389061959155418, | |
| "grad_norm": 0.35364240407943726, | |
| "learning_rate": 4.4254599249281016e-05, | |
| "loss": 0.3992, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.4458290065766701, | |
| "grad_norm": 0.3062623143196106, | |
| "learning_rate": 4.3254782953204375e-05, | |
| "loss": 0.4127, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 1.4527518172377984, | |
| "grad_norm": 0.33005601167678833, | |
| "learning_rate": 4.2263264121978505e-05, | |
| "loss": 0.4262, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.459674627898927, | |
| "grad_norm": 0.3054872453212738, | |
| "learning_rate": 4.12801877386183e-05, | |
| "loss": 0.4031, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 1.4665974385600553, | |
| "grad_norm": 0.324898362159729, | |
| "learning_rate": 4.030569755165736e-05, | |
| "loss": 0.4181, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.4735202492211839, | |
| "grad_norm": 0.32239148020744324, | |
| "learning_rate": 3.9339936054128466e-05, | |
| "loss": 0.4048, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 1.4804430598823122, | |
| "grad_norm": 0.31783732771873474, | |
| "learning_rate": 3.8383044462727826e-05, | |
| "loss": 0.4034, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.4873658705434407, | |
| "grad_norm": 0.33856016397476196, | |
| "learning_rate": 3.7435162697166107e-05, | |
| "loss": 0.4142, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.494288681204569, | |
| "grad_norm": 0.3253874182701111, | |
| "learning_rate": 3.649642935970859e-05, | |
| "loss": 0.4172, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.5012114918656976, | |
| "grad_norm": 0.29950690269470215, | |
| "learning_rate": 3.556698171490871e-05, | |
| "loss": 0.4094, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 1.508134302526826, | |
| "grad_norm": 0.3280077278614044, | |
| "learning_rate": 3.464695566953644e-05, | |
| "loss": 0.4187, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.5150571131879542, | |
| "grad_norm": 0.32570454478263855, | |
| "learning_rate": 3.373648575270576e-05, | |
| "loss": 0.4152, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 1.5219799238490828, | |
| "grad_norm": 0.32040858268737793, | |
| "learning_rate": 3.283570509620344e-05, | |
| "loss": 0.4182, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.5289027345102113, | |
| "grad_norm": 0.3151017725467682, | |
| "learning_rate": 3.1944745415021915e-05, | |
| "loss": 0.4107, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 1.5358255451713396, | |
| "grad_norm": 0.31537941098213196, | |
| "learning_rate": 3.106373698809958e-05, | |
| "loss": 0.407, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.542748355832468, | |
| "grad_norm": 0.3275078237056732, | |
| "learning_rate": 3.0192808639271065e-05, | |
| "loss": 0.4145, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 1.5496711664935963, | |
| "grad_norm": 0.30740946531295776, | |
| "learning_rate": 2.933208771843018e-05, | |
| "loss": 0.4177, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.5565939771547248, | |
| "grad_norm": 0.30290839076042175, | |
| "learning_rate": 2.8481700082908268e-05, | |
| "loss": 0.4058, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.5635167878158533, | |
| "grad_norm": 0.28324180841445923, | |
| "learning_rate": 2.7641770079071127e-05, | |
| "loss": 0.4004, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.5704395984769817, | |
| "grad_norm": 0.3258994519710541, | |
| "learning_rate": 2.681242052413656e-05, | |
| "loss": 0.4032, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 1.57736240913811, | |
| "grad_norm": 0.32113397121429443, | |
| "learning_rate": 2.59937726882158e-05, | |
| "loss": 0.4053, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.5842852197992385, | |
| "grad_norm": 0.3457752764225006, | |
| "learning_rate": 2.5185946276580918e-05, | |
| "loss": 0.4064, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 1.591208030460367, | |
| "grad_norm": 0.32053738832473755, | |
| "learning_rate": 2.4389059412161087e-05, | |
| "loss": 0.4082, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.5981308411214954, | |
| "grad_norm": 0.3150237798690796, | |
| "learning_rate": 2.3603228618270356e-05, | |
| "loss": 0.4131, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 1.6050536517826237, | |
| "grad_norm": 0.33137986063957214, | |
| "learning_rate": 2.2828568801569283e-05, | |
| "loss": 0.3979, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.611976462443752, | |
| "grad_norm": 0.3342689871788025, | |
| "learning_rate": 2.2065193235262782e-05, | |
| "loss": 0.4067, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 1.6188992731048806, | |
| "grad_norm": 0.319580614566803, | |
| "learning_rate": 2.1313213542537058e-05, | |
| "loss": 0.3997, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.6258220837660091, | |
| "grad_norm": 0.32372474670410156, | |
| "learning_rate": 2.0572739680237717e-05, | |
| "loss": 0.4109, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.6327448944271374, | |
| "grad_norm": 0.30622854828834534, | |
| "learning_rate": 1.984387992279153e-05, | |
| "loss": 0.4118, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.6396677050882658, | |
| "grad_norm": 0.31043940782546997, | |
| "learning_rate": 1.912674084637409e-05, | |
| "loss": 0.396, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 1.6465905157493943, | |
| "grad_norm": 0.33528950810432434, | |
| "learning_rate": 1.8421427313326046e-05, | |
| "loss": 0.4081, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.6535133264105226, | |
| "grad_norm": 0.32642996311187744, | |
| "learning_rate": 1.7728042456819704e-05, | |
| "loss": 0.3967, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 1.6604361370716512, | |
| "grad_norm": 0.31457340717315674, | |
| "learning_rate": 1.7046687665778717e-05, | |
| "loss": 0.4001, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.6673589477327795, | |
| "grad_norm": 0.3099569082260132, | |
| "learning_rate": 1.6377462570052438e-05, | |
| "loss": 0.3995, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 1.6742817583939078, | |
| "grad_norm": 0.3081655204296112, | |
| "learning_rate": 1.5720465025847987e-05, | |
| "loss": 0.4033, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.6812045690550363, | |
| "grad_norm": 0.313600093126297, | |
| "learning_rate": 1.5075791101421166e-05, | |
| "loss": 0.3984, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 1.6881273797161649, | |
| "grad_norm": 0.31717297434806824, | |
| "learning_rate": 1.4443535063029279e-05, | |
| "loss": 0.3999, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.6950501903772932, | |
| "grad_norm": 0.31446799635887146, | |
| "learning_rate": 1.3823789361147044e-05, | |
| "loss": 0.4051, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.7019730010384215, | |
| "grad_norm": 0.3024658262729645, | |
| "learning_rate": 1.3216644616948282e-05, | |
| "loss": 0.4014, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.70889581169955, | |
| "grad_norm": 0.2988710403442383, | |
| "learning_rate": 1.2622189609054979e-05, | |
| "loss": 0.3962, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 1.7158186223606784, | |
| "grad_norm": 0.31273016333580017, | |
| "learning_rate": 1.2040511260555831e-05, | |
| "loss": 0.3929, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.722741433021807, | |
| "grad_norm": 0.3140827715396881, | |
| "learning_rate": 1.1471694626296049e-05, | |
| "loss": 0.3957, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 1.7296642436829353, | |
| "grad_norm": 0.30374085903167725, | |
| "learning_rate": 1.0915822880440308e-05, | |
| "loss": 0.3928, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.7365870543440636, | |
| "grad_norm": 0.3040621280670166, | |
| "learning_rate": 1.037297730431085e-05, | |
| "loss": 0.3835, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 1.7435098650051921, | |
| "grad_norm": 0.307586133480072, | |
| "learning_rate": 9.843237274502282e-06, | |
| "loss": 0.4083, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.7504326756663207, | |
| "grad_norm": 0.3173294961452484, | |
| "learning_rate": 9.326680251274777e-06, | |
| "loss": 0.3944, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 1.757355486327449, | |
| "grad_norm": 0.3213316798210144, | |
| "learning_rate": 8.823381767227667e-06, | |
| "loss": 0.3978, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.7642782969885773, | |
| "grad_norm": 0.3075086772441864, | |
| "learning_rate": 8.333415416254831e-06, | |
| "loss": 0.3966, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 1.7712011076497056, | |
| "grad_norm": 0.2939542531967163, | |
| "learning_rate": 7.856852842783547e-06, | |
| "loss": 0.4033, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.7781239183108342, | |
| "grad_norm": 0.29994168877601624, | |
| "learning_rate": 7.393763731298387e-06, | |
| "loss": 0.4127, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 1.7850467289719627, | |
| "grad_norm": 0.3047325313091278, | |
| "learning_rate": 6.944215796151765e-06, | |
| "loss": 0.3925, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.791969539633091, | |
| "grad_norm": 0.32491716742515564, | |
| "learning_rate": 6.5082747716625255e-06, | |
| "loss": 0.3887, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 1.7988923502942193, | |
| "grad_norm": 0.30582690238952637, | |
| "learning_rate": 6.086004402504098e-06, | |
| "loss": 0.402, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.8058151609553479, | |
| "grad_norm": 0.30637919902801514, | |
| "learning_rate": 5.6774664343834495e-06, | |
| "loss": 0.402, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 1.8127379716164764, | |
| "grad_norm": 0.3132084310054779, | |
| "learning_rate": 5.2827206050125585e-06, | |
| "loss": 0.4095, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.8196607822776047, | |
| "grad_norm": 0.3189055621623993, | |
| "learning_rate": 4.9018246353732755e-06, | |
| "loss": 0.4076, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 1.826583592938733, | |
| "grad_norm": 0.30169403553009033, | |
| "learning_rate": 4.534834221277262e-06, | |
| "loss": 0.395, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.8335064035998614, | |
| "grad_norm": 0.29050448536872864, | |
| "learning_rate": 4.181803025221898e-06, | |
| "loss": 0.4005, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 1.84042921426099, | |
| "grad_norm": 0.31095075607299805, | |
| "learning_rate": 3.842782668543599e-06, | |
| "loss": 0.3919, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.8473520249221185, | |
| "grad_norm": 0.2732907235622406, | |
| "learning_rate": 3.5178227238696347e-06, | |
| "loss": 0.4076, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 1.8542748355832468, | |
| "grad_norm": 0.2910282611846924, | |
| "learning_rate": 3.2069707078694057e-06, | |
| "loss": 0.3896, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.8611976462443751, | |
| "grad_norm": 0.2946375608444214, | |
| "learning_rate": 2.9102720743064527e-06, | |
| "loss": 0.3889, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 1.8681204569055037, | |
| "grad_norm": 0.31166285276412964, | |
| "learning_rate": 2.6277702073919997e-06, | |
| "loss": 0.3938, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.8750432675666322, | |
| "grad_norm": 0.3089560866355896, | |
| "learning_rate": 2.3595064154412374e-06, | |
| "loss": 0.3973, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 1.8819660782277605, | |
| "grad_norm": 0.3016653060913086, | |
| "learning_rate": 2.1055199248330546e-06, | |
| "loss": 0.3873, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.8888888888888888, | |
| "grad_norm": 0.3156551420688629, | |
| "learning_rate": 1.8658478742742625e-06, | |
| "loss": 0.3971, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 1.8958116995500172, | |
| "grad_norm": 0.3043115437030792, | |
| "learning_rate": 1.6405253093690343e-06, | |
| "loss": 0.4001, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.9027345102111457, | |
| "grad_norm": 0.31930527091026306, | |
| "learning_rate": 1.429585177494419e-06, | |
| "loss": 0.3966, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 1.9096573208722742, | |
| "grad_norm": 0.3331674337387085, | |
| "learning_rate": 1.2330583229827053e-06, | |
| "loss": 0.4058, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.9165801315334026, | |
| "grad_norm": 0.30298393964767456, | |
| "learning_rate": 1.0509734826112394e-06, | |
| "loss": 0.3881, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 1.9235029421945309, | |
| "grad_norm": 0.32788029313087463, | |
| "learning_rate": 8.833572814004321e-07, | |
| "loss": 0.3874, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.9304257528556594, | |
| "grad_norm": 0.2930392622947693, | |
| "learning_rate": 7.302342287205943e-07, | |
| "loss": 0.3862, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 1.937348563516788, | |
| "grad_norm": 0.2866605222225189, | |
| "learning_rate": 5.916267147080934e-07, | |
| "loss": 0.4039, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.9442713741779163, | |
| "grad_norm": 0.3281935453414917, | |
| "learning_rate": 4.6755500699138253e-07, | |
| "loss": 0.3972, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 1.9511941848390446, | |
| "grad_norm": 0.2974799871444702, | |
| "learning_rate": 3.5803724772740523e-07, | |
| "loss": 0.3914, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.958116995500173, | |
| "grad_norm": 0.29594045877456665, | |
| "learning_rate": 2.630894509488058e-07, | |
| "loss": 0.4055, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 1.9650398061613015, | |
| "grad_norm": 0.29101622104644775, | |
| "learning_rate": 1.8272550022230272e-07, | |
| "loss": 0.3963, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.97196261682243, | |
| "grad_norm": 0.31727471947669983, | |
| "learning_rate": 1.1695714661856816e-07, | |
| "loss": 0.3967, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 1.9788854274835583, | |
| "grad_norm": 0.3088945746421814, | |
| "learning_rate": 6.579400699397998e-08, | |
| "loss": 0.3956, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.9858082381446867, | |
| "grad_norm": 0.31148701906204224, | |
| "learning_rate": 2.9243562584402305e-08, | |
| "loss": 0.3919, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 1.9927310488058152, | |
| "grad_norm": 0.297936350107193, | |
| "learning_rate": 7.311157911249478e-09, | |
| "loss": 0.3962, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.9982692973347178, | |
| "eval_loss": 0.549924910068512, | |
| "eval_runtime": 83.7401, | |
| "eval_samples_per_second": 24.707, | |
| "eval_steps_per_second": 1.552, | |
| "step": 1444 | |
| }, | |
| { | |
| "epoch": 1.9982692973347178, | |
| "step": 1444, | |
| "total_flos": 6.937079953921933e+18, | |
| "train_loss": 0.5484044306661283, | |
| "train_runtime": 15358.5626, | |
| "train_samples_per_second": 6.019, | |
| "train_steps_per_second": 0.094 | |
| } | |
| ], | |
| "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": 6.937079953921933e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |