| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 3665, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.013642564802182811, | |
| "grad_norm": 5.18676233291626, | |
| "learning_rate": 7.999880967748602e-05, | |
| "loss": 2.6446, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.027285129604365622, | |
| "grad_norm": 5.224745273590088, | |
| "learning_rate": 7.999469507332807e-05, | |
| "loss": 2.5833, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.040927694406548434, | |
| "grad_norm": 5.343475341796875, | |
| "learning_rate": 7.998764179444167e-05, | |
| "loss": 2.3323, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.054570259208731244, | |
| "grad_norm": 3.960702657699585, | |
| "learning_rate": 7.997765035907784e-05, | |
| "loss": 2.2764, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06821282401091405, | |
| "grad_norm": 5.80894136428833, | |
| "learning_rate": 7.99661465457664e-05, | |
| "loss": 2.3653, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08185538881309687, | |
| "grad_norm": 3.9454762935638428, | |
| "learning_rate": 7.995057481402883e-05, | |
| "loss": 2.3655, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09549795361527967, | |
| "grad_norm": 4.658463478088379, | |
| "learning_rate": 7.993206764937005e-05, | |
| "loss": 2.2008, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.10914051841746249, | |
| "grad_norm": 7.062836647033691, | |
| "learning_rate": 7.99106264116338e-05, | |
| "loss": 2.4619, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.12278308321964529, | |
| "grad_norm": 5.339938640594482, | |
| "learning_rate": 7.988625267624962e-05, | |
| "loss": 2.4662, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1364256480218281, | |
| "grad_norm": 6.298858165740967, | |
| "learning_rate": 7.9858948234117e-05, | |
| "loss": 2.3909, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15006821282401092, | |
| "grad_norm": 5.622934818267822, | |
| "learning_rate": 7.98287150914739e-05, | |
| "loss": 2.4736, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.16371077762619374, | |
| "grad_norm": 3.229802370071411, | |
| "learning_rate": 7.979555546974922e-05, | |
| "loss": 2.3796, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.17735334242837653, | |
| "grad_norm": 5.506136894226074, | |
| "learning_rate": 7.975947180539966e-05, | |
| "loss": 2.2357, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.19099590723055934, | |
| "grad_norm": 5.04240608215332, | |
| "learning_rate": 7.972046674973065e-05, | |
| "loss": 2.3435, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.20463847203274216, | |
| "grad_norm": 6.470039367675781, | |
| "learning_rate": 7.967854316870156e-05, | |
| "loss": 2.4652, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.21828103683492497, | |
| "grad_norm": 3.5564420223236084, | |
| "learning_rate": 7.963370414271514e-05, | |
| "loss": 2.2549, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.23192360163710776, | |
| "grad_norm": 3.998383045196533, | |
| "learning_rate": 7.958595296639115e-05, | |
| "loss": 2.3102, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.24556616643929058, | |
| "grad_norm": 4.615527629852295, | |
| "learning_rate": 7.953529314832426e-05, | |
| "loss": 2.2444, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2592087312414734, | |
| "grad_norm": 6.488475322723389, | |
| "learning_rate": 7.948172841082635e-05, | |
| "loss": 2.2928, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2728512960436562, | |
| "grad_norm": 4.134075164794922, | |
| "learning_rate": 7.942526268965287e-05, | |
| "loss": 2.2113, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.286493860845839, | |
| "grad_norm": 5.373613357543945, | |
| "learning_rate": 7.936590013371378e-05, | |
| "loss": 2.0795, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.30013642564802184, | |
| "grad_norm": 5.965483665466309, | |
| "learning_rate": 7.93036451047686e-05, | |
| "loss": 2.3991, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.31377899045020463, | |
| "grad_norm": 4.945356369018555, | |
| "learning_rate": 7.923850217710604e-05, | |
| "loss": 2.3399, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3274215552523875, | |
| "grad_norm": 5.210105895996094, | |
| "learning_rate": 7.917047613720773e-05, | |
| "loss": 2.3305, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.34106412005457026, | |
| "grad_norm": 3.3301730155944824, | |
| "learning_rate": 7.909957198339669e-05, | |
| "loss": 2.2404, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.35470668485675305, | |
| "grad_norm": 3.685528039932251, | |
| "learning_rate": 7.902579492546998e-05, | |
| "loss": 2.1853, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3683492496589359, | |
| "grad_norm": 5.30728006362915, | |
| "learning_rate": 7.894915038431595e-05, | |
| "loss": 2.3177, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.3819918144611187, | |
| "grad_norm": 2.644773006439209, | |
| "learning_rate": 7.886964399151586e-05, | |
| "loss": 2.3621, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3956343792633015, | |
| "grad_norm": 4.921834468841553, | |
| "learning_rate": 7.878728158893015e-05, | |
| "loss": 2.3687, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.4092769440654843, | |
| "grad_norm": 3.9288289546966553, | |
| "learning_rate": 7.870206922826921e-05, | |
| "loss": 2.2021, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4229195088676671, | |
| "grad_norm": 6.143318176269531, | |
| "learning_rate": 7.862294656001264e-05, | |
| "loss": 2.0236, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.43656207366984995, | |
| "grad_norm": 4.3533806800842285, | |
| "learning_rate": 7.853233670108533e-05, | |
| "loss": 2.4181, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.45020463847203274, | |
| "grad_norm": 3.7804017066955566, | |
| "learning_rate": 7.843889561656962e-05, | |
| "loss": 2.2359, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4638472032742155, | |
| "grad_norm": 2.7896721363067627, | |
| "learning_rate": 7.83426301721999e-05, | |
| "loss": 2.2156, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.47748976807639837, | |
| "grad_norm": 4.168877601623535, | |
| "learning_rate": 7.824354744123498e-05, | |
| "loss": 2.256, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.49113233287858116, | |
| "grad_norm": 7.003695011138916, | |
| "learning_rate": 7.814165470393832e-05, | |
| "loss": 2.2627, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.504774897680764, | |
| "grad_norm": 9.064947128295898, | |
| "learning_rate": 7.803695944704313e-05, | |
| "loss": 2.1303, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5184174624829468, | |
| "grad_norm": 6.936222553253174, | |
| "learning_rate": 7.79294693632023e-05, | |
| "loss": 2.307, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5320600272851296, | |
| "grad_norm": 4.441155433654785, | |
| "learning_rate": 7.781919235042309e-05, | |
| "loss": 2.2182, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5457025920873124, | |
| "grad_norm": 7.236617088317871, | |
| "learning_rate": 7.770613651148688e-05, | |
| "loss": 2.2309, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5593451568894953, | |
| "grad_norm": 4.184124946594238, | |
| "learning_rate": 7.75903101533538e-05, | |
| "loss": 2.0949, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.572987721691678, | |
| "grad_norm": 2.9721992015838623, | |
| "learning_rate": 7.747172178655232e-05, | |
| "loss": 2.0541, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5866302864938608, | |
| "grad_norm": 2.36011004447937, | |
| "learning_rate": 7.735038012455398e-05, | |
| "loss": 2.0429, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6002728512960437, | |
| "grad_norm": 5.892980575561523, | |
| "learning_rate": 7.722629408313309e-05, | |
| "loss": 2.1867, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6139154160982264, | |
| "grad_norm": 5.935312747955322, | |
| "learning_rate": 7.709947277971168e-05, | |
| "loss": 2.2556, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6275579809004093, | |
| "grad_norm": 5.781426906585693, | |
| "learning_rate": 7.69699255326896e-05, | |
| "loss": 2.3728, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6412005457025921, | |
| "grad_norm": 6.5159149169921875, | |
| "learning_rate": 7.683766186075973e-05, | |
| "loss": 2.4319, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.654843110504775, | |
| "grad_norm": 3.946726083755493, | |
| "learning_rate": 7.670269148220874e-05, | |
| "loss": 2.2417, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.6684856753069577, | |
| "grad_norm": 5.711483001708984, | |
| "learning_rate": 7.656502431420286e-05, | |
| "loss": 2.1867, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.6821282401091405, | |
| "grad_norm": 1.9001113176345825, | |
| "learning_rate": 7.64246704720593e-05, | |
| "loss": 2.2426, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6957708049113234, | |
| "grad_norm": 6.293882846832275, | |
| "learning_rate": 7.628164026850302e-05, | |
| "loss": 2.1057, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.7094133697135061, | |
| "grad_norm": 2.980394124984741, | |
| "learning_rate": 7.613594421290888e-05, | |
| "loss": 2.0724, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.723055934515689, | |
| "grad_norm": 7.354122161865234, | |
| "learning_rate": 7.59875930105296e-05, | |
| "loss": 2.1513, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.7366984993178718, | |
| "grad_norm": 3.451127290725708, | |
| "learning_rate": 7.583659756170904e-05, | |
| "loss": 1.996, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.7503410641200545, | |
| "grad_norm": 4.8892951011657715, | |
| "learning_rate": 7.568296896108135e-05, | |
| "loss": 2.2927, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.7639836289222374, | |
| "grad_norm": 5.78281831741333, | |
| "learning_rate": 7.552671849675571e-05, | |
| "loss": 2.2445, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.7776261937244202, | |
| "grad_norm": 3.0643134117126465, | |
| "learning_rate": 7.536785764948705e-05, | |
| "loss": 2.1723, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.791268758526603, | |
| "grad_norm": 5.0908732414245605, | |
| "learning_rate": 7.520639809183234e-05, | |
| "loss": 2.1945, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.8049113233287858, | |
| "grad_norm": 4.066890716552734, | |
| "learning_rate": 7.5042351687293e-05, | |
| "loss": 2.385, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.8185538881309686, | |
| "grad_norm": 6.034420967102051, | |
| "learning_rate": 7.487573048944318e-05, | |
| "loss": 1.8776, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8321964529331515, | |
| "grad_norm": 2.5823731422424316, | |
| "learning_rate": 7.470654674104415e-05, | |
| "loss": 2.1586, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.8458390177353342, | |
| "grad_norm": 4.781266212463379, | |
| "learning_rate": 7.453481287314469e-05, | |
| "loss": 2.1229, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.859481582537517, | |
| "grad_norm": 4.008852481842041, | |
| "learning_rate": 7.436054150416777e-05, | |
| "loss": 2.1562, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.8731241473396999, | |
| "grad_norm": 7.812346458435059, | |
| "learning_rate": 7.418374543898328e-05, | |
| "loss": 2.2563, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.8867667121418826, | |
| "grad_norm": 5.588919639587402, | |
| "learning_rate": 7.400443766796728e-05, | |
| "loss": 2.0738, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.9004092769440655, | |
| "grad_norm": 3.1923656463623047, | |
| "learning_rate": 7.382263136604744e-05, | |
| "loss": 1.9363, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.9140518417462483, | |
| "grad_norm": 6.451751708984375, | |
| "learning_rate": 7.363833989173504e-05, | |
| "loss": 2.3014, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.927694406548431, | |
| "grad_norm": 4.7484517097473145, | |
| "learning_rate": 7.34515767861434e-05, | |
| "loss": 2.1957, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.9413369713506139, | |
| "grad_norm": 4.468317031860352, | |
| "learning_rate": 7.326235577199292e-05, | |
| "loss": 2.4058, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.9549795361527967, | |
| "grad_norm": 4.4900078773498535, | |
| "learning_rate": 7.307069075260283e-05, | |
| "loss": 2.0714, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9686221009549796, | |
| "grad_norm": 4.51573371887207, | |
| "learning_rate": 7.287659581086957e-05, | |
| "loss": 2.0244, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.9822646657571623, | |
| "grad_norm": 2.892160654067993, | |
| "learning_rate": 7.268008520823203e-05, | |
| "loss": 2.1891, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.9959072305593452, | |
| "grad_norm": 5.765126705169678, | |
| "learning_rate": 7.248117338362371e-05, | |
| "loss": 2.0008, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.009549795361528, | |
| "grad_norm": 4.076498031616211, | |
| "learning_rate": 7.227987495241174e-05, | |
| "loss": 1.7823, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.0231923601637107, | |
| "grad_norm": 5.407172203063965, | |
| "learning_rate": 7.207620470532302e-05, | |
| "loss": 1.7272, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.0368349249658937, | |
| "grad_norm": 2.3942642211914062, | |
| "learning_rate": 7.187017760735748e-05, | |
| "loss": 1.7809, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.0504774897680764, | |
| "grad_norm": 6.462409019470215, | |
| "learning_rate": 7.166180879668843e-05, | |
| "loss": 1.5226, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.0641200545702592, | |
| "grad_norm": 10.609193801879883, | |
| "learning_rate": 7.145111358355031e-05, | |
| "loss": 1.6743, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.077762619372442, | |
| "grad_norm": 5.816223621368408, | |
| "learning_rate": 7.123810744911376e-05, | |
| "loss": 1.3974, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.0914051841746248, | |
| "grad_norm": 8.911855697631836, | |
| "learning_rate": 7.102280604434805e-05, | |
| "loss": 1.4882, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.1050477489768076, | |
| "grad_norm": 8.37972640991211, | |
| "learning_rate": 7.080522518887116e-05, | |
| "loss": 1.8217, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.1186903137789905, | |
| "grad_norm": 5.856902599334717, | |
| "learning_rate": 7.058538086978738e-05, | |
| "loss": 1.5443, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.1323328785811733, | |
| "grad_norm": 8.661952018737793, | |
| "learning_rate": 7.036328924051266e-05, | |
| "loss": 1.426, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.145975443383356, | |
| "grad_norm": 7.178607940673828, | |
| "learning_rate": 7.013896661958766e-05, | |
| "loss": 1.6086, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.159618008185539, | |
| "grad_norm": 6.399686813354492, | |
| "learning_rate": 6.991242948947879e-05, | |
| "loss": 1.64, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.1732605729877217, | |
| "grad_norm": 5.770147800445557, | |
| "learning_rate": 6.968369449536705e-05, | |
| "loss": 1.6026, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.1869031377899044, | |
| "grad_norm": 7.86605167388916, | |
| "learning_rate": 6.945277844392504e-05, | |
| "loss": 1.5026, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.2005457025920874, | |
| "grad_norm": 3.953367233276367, | |
| "learning_rate": 6.921969830208212e-05, | |
| "loss": 1.6705, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.21418826739427, | |
| "grad_norm": 10.93819808959961, | |
| "learning_rate": 6.898447119577764e-05, | |
| "loss": 1.6982, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.2278308321964528, | |
| "grad_norm": 4.714999675750732, | |
| "learning_rate": 6.874711440870264e-05, | |
| "loss": 1.2631, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.2414733969986358, | |
| "grad_norm": 4.0259480476379395, | |
| "learning_rate": 6.85076453810299e-05, | |
| "loss": 1.5005, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.2551159618008185, | |
| "grad_norm": 7.3578643798828125, | |
| "learning_rate": 6.826608170813243e-05, | |
| "loss": 1.6076, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.2687585266030013, | |
| "grad_norm": 5.368468284606934, | |
| "learning_rate": 6.802244113929075e-05, | |
| "loss": 1.7727, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.2824010914051842, | |
| "grad_norm": 7.0559916496276855, | |
| "learning_rate": 6.777674157638862e-05, | |
| "loss": 1.6125, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.296043656207367, | |
| "grad_norm": 6.349229335784912, | |
| "learning_rate": 6.75290010725977e-05, | |
| "loss": 1.4289, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.30968622100955, | |
| "grad_norm": 4.133677959442139, | |
| "learning_rate": 6.727923783105111e-05, | |
| "loss": 1.6043, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.3233287858117326, | |
| "grad_norm": 5.688299655914307, | |
| "learning_rate": 6.702747020350586e-05, | |
| "loss": 1.6637, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.3369713506139154, | |
| "grad_norm": 8.215998649597168, | |
| "learning_rate": 6.677371668899448e-05, | |
| "loss": 1.717, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.350613915416098, | |
| "grad_norm": 6.267200946807861, | |
| "learning_rate": 6.651799593246569e-05, | |
| "loss": 1.572, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.364256480218281, | |
| "grad_norm": 3.5424532890319824, | |
| "learning_rate": 6.626032672341458e-05, | |
| "loss": 1.6264, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.3778990450204638, | |
| "grad_norm": 6.2543206214904785, | |
| "learning_rate": 6.600072799450186e-05, | |
| "loss": 1.5057, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.3915416098226467, | |
| "grad_norm": 5.060370445251465, | |
| "learning_rate": 6.573921882016284e-05, | |
| "loss": 1.6963, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.4051841746248295, | |
| "grad_norm": 7.101504325866699, | |
| "learning_rate": 6.547581841520589e-05, | |
| "loss": 1.3956, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.4188267394270122, | |
| "grad_norm": 7.740041732788086, | |
| "learning_rate": 6.521054613340064e-05, | |
| "loss": 1.469, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.4324693042291952, | |
| "grad_norm": 7.723188877105713, | |
| "learning_rate": 6.494342146605581e-05, | |
| "loss": 1.5518, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.446111869031378, | |
| "grad_norm": 5.523282527923584, | |
| "learning_rate": 6.467446404058722e-05, | |
| "loss": 1.5123, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.4597544338335606, | |
| "grad_norm": 10.065855026245117, | |
| "learning_rate": 6.44036936190755e-05, | |
| "loss": 1.5065, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.4733969986357436, | |
| "grad_norm": 8.757994651794434, | |
| "learning_rate": 6.413113009681411e-05, | |
| "loss": 1.5362, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.4870395634379263, | |
| "grad_norm": 10.307364463806152, | |
| "learning_rate": 6.385679350084743e-05, | |
| "loss": 1.7138, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.500682128240109, | |
| "grad_norm": 7.119322299957275, | |
| "learning_rate": 6.358070398849938e-05, | |
| "loss": 1.6659, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.514324693042292, | |
| "grad_norm": 12.579729080200195, | |
| "learning_rate": 6.330288184589216e-05, | |
| "loss": 1.6374, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.5279672578444747, | |
| "grad_norm": 8.232572555541992, | |
| "learning_rate": 6.30233474864558e-05, | |
| "loss": 1.6296, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.5416098226466577, | |
| "grad_norm": 10.136465072631836, | |
| "learning_rate": 6.274212144942824e-05, | |
| "loss": 1.6491, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.5552523874488404, | |
| "grad_norm": 5.127601623535156, | |
| "learning_rate": 6.245922439834612e-05, | |
| "loss": 1.6309, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.5688949522510232, | |
| "grad_norm": 6.231171607971191, | |
| "learning_rate": 6.217467711952658e-05, | |
| "loss": 1.5365, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.5825375170532059, | |
| "grad_norm": 7.796588897705078, | |
| "learning_rate": 6.188850052053985e-05, | |
| "loss": 1.493, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.5961800818553888, | |
| "grad_norm": 8.250170707702637, | |
| "learning_rate": 6.160071562867316e-05, | |
| "loss": 1.6771, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.6098226466575716, | |
| "grad_norm": 4.563484191894531, | |
| "learning_rate": 6.131134358938559e-05, | |
| "loss": 1.5381, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.6234652114597545, | |
| "grad_norm": 7.776663303375244, | |
| "learning_rate": 6.1020405664754455e-05, | |
| "loss": 1.4791, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.6371077762619373, | |
| "grad_norm": 6.122801780700684, | |
| "learning_rate": 6.0727923231913035e-05, | |
| "loss": 1.5296, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.65075034106412, | |
| "grad_norm": 5.5693278312683105, | |
| "learning_rate": 6.04339177814798e-05, | |
| "loss": 1.4165, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.6643929058663027, | |
| "grad_norm": 8.107104301452637, | |
| "learning_rate": 6.013841091597947e-05, | |
| "loss": 1.5694, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.6780354706684857, | |
| "grad_norm": 12.389050483703613, | |
| "learning_rate": 5.9841424348255596e-05, | |
| "loss": 1.6024, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.6916780354706686, | |
| "grad_norm": 6.175624370574951, | |
| "learning_rate": 5.954297989987526e-05, | |
| "loss": 1.499, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.7053206002728514, | |
| "grad_norm": 4.447775840759277, | |
| "learning_rate": 5.924309949952565e-05, | |
| "loss": 1.5177, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.718963165075034, | |
| "grad_norm": 6.788479804992676, | |
| "learning_rate": 5.8941805181402886e-05, | |
| "loss": 1.4564, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.7326057298772168, | |
| "grad_norm": 6.172934532165527, | |
| "learning_rate": 5.8639119083592954e-05, | |
| "loss": 1.3263, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.7462482946793996, | |
| "grad_norm": 7.947272777557373, | |
| "learning_rate": 5.833506344644507e-05, | |
| "loss": 1.6518, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.7598908594815825, | |
| "grad_norm": 5.520730018615723, | |
| "learning_rate": 5.802966061093762e-05, | |
| "loss": 1.4169, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.7735334242837655, | |
| "grad_norm": 6.108306407928467, | |
| "learning_rate": 5.7722933017036515e-05, | |
| "loss": 1.4631, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.7871759890859482, | |
| "grad_norm": 3.4584460258483887, | |
| "learning_rate": 5.741490320204644e-05, | |
| "loss": 1.5939, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.800818553888131, | |
| "grad_norm": 7.870245456695557, | |
| "learning_rate": 5.7105593798954895e-05, | |
| "loss": 1.4775, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.8144611186903137, | |
| "grad_norm": 6.256468772888184, | |
| "learning_rate": 5.679502753476913e-05, | |
| "loss": 1.6513, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.8281036834924966, | |
| "grad_norm": 8.314029693603516, | |
| "learning_rate": 5.648322722884635e-05, | |
| "loss": 1.6092, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.8417462482946794, | |
| "grad_norm": 7.979041576385498, | |
| "learning_rate": 5.6170215791216926e-05, | |
| "loss": 1.5116, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.8553888130968623, | |
| "grad_norm": 7.900625228881836, | |
| "learning_rate": 5.58560162209011e-05, | |
| "loss": 1.3667, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.869031377899045, | |
| "grad_norm": 7.304281234741211, | |
| "learning_rate": 5.554065160421907e-05, | |
| "loss": 1.4584, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.8826739427012278, | |
| "grad_norm": 5.6456618309021, | |
| "learning_rate": 5.522414511309472e-05, | |
| "loss": 1.7889, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.8963165075034105, | |
| "grad_norm": 5.90806770324707, | |
| "learning_rate": 5.490652000335297e-05, | |
| "loss": 1.5623, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.9099590723055935, | |
| "grad_norm": 6.941085338592529, | |
| "learning_rate": 5.461972027286809e-05, | |
| "loss": 1.5508, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.9236016371077762, | |
| "grad_norm": 7.18194580078125, | |
| "learning_rate": 5.430003415054097e-05, | |
| "loss": 1.4951, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.9372442019099592, | |
| "grad_norm": 9.118647575378418, | |
| "learning_rate": 5.397929731013993e-05, | |
| "loss": 1.601, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.950886766712142, | |
| "grad_norm": 7.178833961486816, | |
| "learning_rate": 5.365753331832165e-05, | |
| "loss": 1.5427, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.9645293315143246, | |
| "grad_norm": 5.578968524932861, | |
| "learning_rate": 5.3334765817214407e-05, | |
| "loss": 1.4017, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.9781718963165074, | |
| "grad_norm": 10.137267112731934, | |
| "learning_rate": 5.301101852268093e-05, | |
| "loss": 1.6255, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.9918144611186903, | |
| "grad_norm": 7.634779930114746, | |
| "learning_rate": 5.268631522257586e-05, | |
| "loss": 1.5628, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.0054570259208733, | |
| "grad_norm": 4.639617919921875, | |
| "learning_rate": 5.23606797749979e-05, | |
| "loss": 1.2903, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.019099590723056, | |
| "grad_norm": 9.048364639282227, | |
| "learning_rate": 5.2034136106536784e-05, | |
| "loss": 0.8694, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.0327421555252387, | |
| "grad_norm": 12.236319541931152, | |
| "learning_rate": 5.1706708210515225e-05, | |
| "loss": 1.0926, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.0463847203274215, | |
| "grad_norm": 8.067400932312012, | |
| "learning_rate": 5.1378420145226e-05, | |
| "loss": 0.9019, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.060027285129604, | |
| "grad_norm": 5.9596028327941895, | |
| "learning_rate": 5.104929603216422e-05, | |
| "loss": 0.9077, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.0736698499317874, | |
| "grad_norm": 16.68425941467285, | |
| "learning_rate": 5.0719360054254925e-05, | |
| "loss": 0.8833, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.08731241473397, | |
| "grad_norm": 10.05809497833252, | |
| "learning_rate": 5.0388636454076256e-05, | |
| "loss": 0.8182, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.100954979536153, | |
| "grad_norm": 7.299218654632568, | |
| "learning_rate": 5.0057149532078165e-05, | |
| "loss": 0.7097, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.1145975443383356, | |
| "grad_norm": 7.193046569824219, | |
| "learning_rate": 4.9724923644796904e-05, | |
| "loss": 0.8243, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.1282401091405183, | |
| "grad_norm": 6.852187633514404, | |
| "learning_rate": 4.939198320306537e-05, | |
| "loss": 0.8468, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.141882673942701, | |
| "grad_norm": 8.87316608428955, | |
| "learning_rate": 4.9058352670219576e-05, | |
| "loss": 0.8348, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.155525238744884, | |
| "grad_norm": 7.688050270080566, | |
| "learning_rate": 4.872405656030099e-05, | |
| "loss": 0.9292, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.169167803547067, | |
| "grad_norm": 9.83198356628418, | |
| "learning_rate": 4.83891194362555e-05, | |
| "loss": 0.7706, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.1828103683492497, | |
| "grad_norm": 10.323474884033203, | |
| "learning_rate": 4.805356590812852e-05, | |
| "loss": 0.8487, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.1964529331514324, | |
| "grad_norm": 8.7017183303833, | |
| "learning_rate": 4.771742063125674e-05, | |
| "loss": 0.8509, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.210095497953615, | |
| "grad_norm": 6.59673547744751, | |
| "learning_rate": 4.7380708304456554e-05, | |
| "loss": 0.8833, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.223738062755798, | |
| "grad_norm": 15.534720420837402, | |
| "learning_rate": 4.704345366820927e-05, | |
| "loss": 0.8968, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.237380627557981, | |
| "grad_norm": 11.261545181274414, | |
| "learning_rate": 4.670568150284323e-05, | |
| "loss": 0.9044, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.251023192360164, | |
| "grad_norm": 8.76096248626709, | |
| "learning_rate": 4.636741662671308e-05, | |
| "loss": 0.974, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.2646657571623465, | |
| "grad_norm": 5.140685081481934, | |
| "learning_rate": 4.602868389437622e-05, | |
| "loss": 0.8121, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.2783083219645293, | |
| "grad_norm": 9.906394004821777, | |
| "learning_rate": 4.568950819476648e-05, | |
| "loss": 1.1324, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.291950886766712, | |
| "grad_norm": 13.181589126586914, | |
| "learning_rate": 4.5349914449365435e-05, | |
| "loss": 0.8856, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.305593451568895, | |
| "grad_norm": 12.309771537780762, | |
| "learning_rate": 4.500992761037129e-05, | |
| "loss": 0.8276, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.319236016371078, | |
| "grad_norm": 10.110664367675781, | |
| "learning_rate": 4.4669572658865405e-05, | |
| "loss": 0.7492, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.3328785811732606, | |
| "grad_norm": 8.248074531555176, | |
| "learning_rate": 4.4328874602976786e-05, | |
| "loss": 0.9167, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.3465211459754434, | |
| "grad_norm": 8.356489181518555, | |
| "learning_rate": 4.3987858476044617e-05, | |
| "loss": 0.9453, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.360163710777626, | |
| "grad_norm": 5.316605091094971, | |
| "learning_rate": 4.364654933477886e-05, | |
| "loss": 0.8714, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.373806275579809, | |
| "grad_norm": 9.405285835266113, | |
| "learning_rate": 4.330497225741917e-05, | |
| "loss": 0.9251, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.3874488403819916, | |
| "grad_norm": 9.250997543334961, | |
| "learning_rate": 4.296315234189223e-05, | |
| "loss": 0.9479, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.4010914051841747, | |
| "grad_norm": 7.570913791656494, | |
| "learning_rate": 4.262111470396766e-05, | |
| "loss": 0.8346, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.4147339699863575, | |
| "grad_norm": 7.284623146057129, | |
| "learning_rate": 4.2278884475412585e-05, | |
| "loss": 0.754, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.42837653478854, | |
| "grad_norm": 8.737554550170898, | |
| "learning_rate": 4.193648680214505e-05, | |
| "loss": 0.7743, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.442019099590723, | |
| "grad_norm": 9.590819358825684, | |
| "learning_rate": 4.159394684238635e-05, | |
| "loss": 0.9273, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.4556616643929057, | |
| "grad_norm": 9.176666259765625, | |
| "learning_rate": 4.1251289764812495e-05, | |
| "loss": 1.0037, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.469304229195089, | |
| "grad_norm": 9.587228775024414, | |
| "learning_rate": 4.090854074670495e-05, | |
| "loss": 0.7728, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.4829467939972716, | |
| "grad_norm": 3.603977918624878, | |
| "learning_rate": 4.056572497210066e-05, | |
| "loss": 0.8098, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.4965893587994543, | |
| "grad_norm": 10.71971321105957, | |
| "learning_rate": 4.0222867629941554e-05, | |
| "loss": 0.8404, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 2.510231923601637, | |
| "grad_norm": 8.777016639709473, | |
| "learning_rate": 3.987999391222389e-05, | |
| "loss": 0.7021, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.52387448840382, | |
| "grad_norm": 9.952046394348145, | |
| "learning_rate": 3.953712901214707e-05, | |
| "loss": 0.9656, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.5375170532060025, | |
| "grad_norm": 10.772456169128418, | |
| "learning_rate": 3.9194298122262666e-05, | |
| "loss": 0.7487, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 2.5511596180081857, | |
| "grad_norm": 7.986364364624023, | |
| "learning_rate": 3.8851526432623254e-05, | |
| "loss": 0.8693, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 2.5648021828103684, | |
| "grad_norm": 8.993766784667969, | |
| "learning_rate": 3.850883912893158e-05, | |
| "loss": 0.8933, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 2.578444747612551, | |
| "grad_norm": 8.045419692993164, | |
| "learning_rate": 3.816626139069004e-05, | |
| "loss": 0.7098, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 2.592087312414734, | |
| "grad_norm": 7.299800872802734, | |
| "learning_rate": 3.782381838935047e-05, | |
| "loss": 0.8742, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.6057298772169166, | |
| "grad_norm": 8.76268482208252, | |
| "learning_rate": 3.748153528646472e-05, | |
| "loss": 0.7846, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 2.6193724420191, | |
| "grad_norm": 13.77757453918457, | |
| "learning_rate": 3.713943723183587e-05, | |
| "loss": 0.9032, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 2.6330150068212825, | |
| "grad_norm": 9.216800689697266, | |
| "learning_rate": 3.6797549361670257e-05, | |
| "loss": 0.8044, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 2.6466575716234653, | |
| "grad_norm": 12.811419486999512, | |
| "learning_rate": 3.6455896796730554e-05, | |
| "loss": 0.8528, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 2.660300136425648, | |
| "grad_norm": 8.257354736328125, | |
| "learning_rate": 3.611450464049005e-05, | |
| "loss": 0.9294, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.6739427012278307, | |
| "grad_norm": 6.599920272827148, | |
| "learning_rate": 3.577339797728805e-05, | |
| "loss": 0.9441, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 2.6875852660300135, | |
| "grad_norm": 10.485870361328125, | |
| "learning_rate": 3.5432601870486795e-05, | |
| "loss": 0.8572, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 2.701227830832196, | |
| "grad_norm": 5.167399883270264, | |
| "learning_rate": 3.509214136062993e-05, | |
| "loss": 0.9912, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 2.7148703956343794, | |
| "grad_norm": 9.000060081481934, | |
| "learning_rate": 3.475204146360254e-05, | |
| "loss": 0.9876, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 2.728512960436562, | |
| "grad_norm": 7.356120586395264, | |
| "learning_rate": 3.44123271687931e-05, | |
| "loss": 0.8796, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.742155525238745, | |
| "grad_norm": 10.706610679626465, | |
| "learning_rate": 3.407302343725737e-05, | |
| "loss": 0.8394, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 2.7557980900409276, | |
| "grad_norm": 6.853879928588867, | |
| "learning_rate": 3.3734155199884275e-05, | |
| "loss": 0.7603, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 2.7694406548431107, | |
| "grad_norm": 8.325788497924805, | |
| "learning_rate": 3.339574735556412e-05, | |
| "loss": 0.8944, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 2.7830832196452935, | |
| "grad_norm": 12.427726745605469, | |
| "learning_rate": 3.3057824769359104e-05, | |
| "loss": 0.9737, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 2.796725784447476, | |
| "grad_norm": 11.044093132019043, | |
| "learning_rate": 3.2720412270676275e-05, | |
| "loss": 0.9483, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 2.810368349249659, | |
| "grad_norm": 5.907009601593018, | |
| "learning_rate": 3.2383534651443206e-05, | |
| "loss": 0.9254, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 2.8240109140518417, | |
| "grad_norm": 5.986649036407471, | |
| "learning_rate": 3.204721666428631e-05, | |
| "loss": 0.7472, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 2.8376534788540244, | |
| "grad_norm": 9.0693359375, | |
| "learning_rate": 3.171148302071215e-05, | |
| "loss": 0.8689, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 2.851296043656207, | |
| "grad_norm": 7.178049087524414, | |
| "learning_rate": 3.137635838929169e-05, | |
| "loss": 0.7493, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 2.8649386084583903, | |
| "grad_norm": 8.808353424072266, | |
| "learning_rate": 3.1041867393847764e-05, | |
| "loss": 1.0867, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.878581173260573, | |
| "grad_norm": 11.312020301818848, | |
| "learning_rate": 3.070803461164575e-05, | |
| "loss": 0.9846, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 2.892223738062756, | |
| "grad_norm": 2.6739227771759033, | |
| "learning_rate": 3.0374884571587776e-05, | |
| "loss": 0.7862, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 2.9058663028649385, | |
| "grad_norm": 9.54980182647705, | |
| "learning_rate": 3.004244175241038e-05, | |
| "loss": 1.0586, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 2.9195088676671213, | |
| "grad_norm": 10.564549446105957, | |
| "learning_rate": 2.971073058088587e-05, | |
| "loss": 0.8757, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 2.9331514324693044, | |
| "grad_norm": 9.608818054199219, | |
| "learning_rate": 2.937977543002764e-05, | |
| "loss": 0.9026, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.946793997271487, | |
| "grad_norm": 4.548944473266602, | |
| "learning_rate": 2.9049600617299188e-05, | |
| "loss": 0.8864, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 2.96043656207367, | |
| "grad_norm": 7.556874752044678, | |
| "learning_rate": 2.872023040282739e-05, | |
| "loss": 0.8224, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 2.9740791268758526, | |
| "grad_norm": 9.024819374084473, | |
| "learning_rate": 2.8391688987620045e-05, | |
| "loss": 0.794, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 2.9877216916780354, | |
| "grad_norm": 5.675693035125732, | |
| "learning_rate": 2.8064000511787523e-05, | |
| "loss": 0.9792, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 3.001364256480218, | |
| "grad_norm": 6.426331520080566, | |
| "learning_rate": 2.77371890527691e-05, | |
| "loss": 0.7812, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.0150068212824013, | |
| "grad_norm": 7.24752140045166, | |
| "learning_rate": 2.741127862356389e-05, | |
| "loss": 0.496, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 3.028649386084584, | |
| "grad_norm": 8.4566650390625, | |
| "learning_rate": 2.7086293170966312e-05, | |
| "loss": 0.3416, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.0422919508867667, | |
| "grad_norm": 5.965997219085693, | |
| "learning_rate": 2.6762256573806664e-05, | |
| "loss": 0.3462, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 3.0559345156889495, | |
| "grad_norm": 10.749195098876953, | |
| "learning_rate": 2.6439192641196583e-05, | |
| "loss": 0.4756, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.069577080491132, | |
| "grad_norm": 7.666036605834961, | |
| "learning_rate": 2.611712511077959e-05, | |
| "loss": 0.357, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.083219645293315, | |
| "grad_norm": 7.038111209869385, | |
| "learning_rate": 2.5796077646986922e-05, | |
| "loss": 0.3361, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 3.096862210095498, | |
| "grad_norm": 10.389034271240234, | |
| "learning_rate": 2.5476073839298857e-05, | |
| "loss": 0.419, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 3.110504774897681, | |
| "grad_norm": 6.829967021942139, | |
| "learning_rate": 2.5157137200511253e-05, | |
| "loss": 0.4211, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 3.1241473396998636, | |
| "grad_norm": 4.85243558883667, | |
| "learning_rate": 2.4839291165008073e-05, | |
| "loss": 0.5248, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 3.1377899045020463, | |
| "grad_norm": 7.922482490539551, | |
| "learning_rate": 2.452255908703945e-05, | |
| "loss": 0.3983, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.151432469304229, | |
| "grad_norm": 4.674210071563721, | |
| "learning_rate": 2.420696423900567e-05, | |
| "loss": 0.3728, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 3.1650750341064118, | |
| "grad_norm": 16.18085289001465, | |
| "learning_rate": 2.3892529809747195e-05, | |
| "loss": 0.4099, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 3.178717598908595, | |
| "grad_norm": 10.127204895019531, | |
| "learning_rate": 2.35792789028409e-05, | |
| "loss": 0.5466, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 3.1923601637107777, | |
| "grad_norm": 6.371789932250977, | |
| "learning_rate": 2.32672345349024e-05, | |
| "loss": 0.3956, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 3.2060027285129604, | |
| "grad_norm": 6.170881748199463, | |
| "learning_rate": 2.2956419633894922e-05, | |
| "loss": 0.2686, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.219645293315143, | |
| "grad_norm": 8.651374816894531, | |
| "learning_rate": 2.264685703744466e-05, | |
| "loss": 0.3545, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 3.233287858117326, | |
| "grad_norm": 9.146707534790039, | |
| "learning_rate": 2.2338569491162688e-05, | |
| "loss": 0.4361, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 3.246930422919509, | |
| "grad_norm": 10.473475456237793, | |
| "learning_rate": 2.2031579646973662e-05, | |
| "loss": 0.4616, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 3.260572987721692, | |
| "grad_norm": 7.665262222290039, | |
| "learning_rate": 2.1725910061451582e-05, | |
| "loss": 0.4095, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 3.2742155525238745, | |
| "grad_norm": 14.176889419555664, | |
| "learning_rate": 2.1421583194162237e-05, | |
| "loss": 0.4428, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.2878581173260573, | |
| "grad_norm": 10.604450225830078, | |
| "learning_rate": 2.1118621406013045e-05, | |
| "loss": 0.4946, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 3.30150068212824, | |
| "grad_norm": 9.188895225524902, | |
| "learning_rate": 2.0817046957610073e-05, | |
| "loss": 0.3676, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 3.3151432469304227, | |
| "grad_norm": 5.148584842681885, | |
| "learning_rate": 2.0516882007622318e-05, | |
| "loss": 0.4319, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 3.328785811732606, | |
| "grad_norm": 10.153374671936035, | |
| "learning_rate": 2.0218148611153614e-05, | |
| "loss": 0.3973, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 3.3424283765347886, | |
| "grad_norm": 2.798651695251465, | |
| "learning_rate": 1.99208687181221e-05, | |
| "loss": 0.3728, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.3560709413369714, | |
| "grad_norm": 7.391672611236572, | |
| "learning_rate": 1.9625064171647403e-05, | |
| "loss": 0.4029, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 3.369713506139154, | |
| "grad_norm": 8.293444633483887, | |
| "learning_rate": 1.933075670644566e-05, | |
| "loss": 0.5182, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 3.383356070941337, | |
| "grad_norm": 12.008163452148438, | |
| "learning_rate": 1.903796794723261e-05, | |
| "loss": 0.3211, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 3.39699863574352, | |
| "grad_norm": 4.839990139007568, | |
| "learning_rate": 1.8746719407134558e-05, | |
| "loss": 0.3218, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 3.4106412005457027, | |
| "grad_norm": 5.815826892852783, | |
| "learning_rate": 1.8457032486107733e-05, | |
| "loss": 0.3561, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.4242837653478855, | |
| "grad_norm": 5.609132766723633, | |
| "learning_rate": 1.816892846936592e-05, | |
| "loss": 0.4223, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 3.437926330150068, | |
| "grad_norm": 6.10213041305542, | |
| "learning_rate": 1.7882428525816434e-05, | |
| "loss": 0.2911, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 3.451568894952251, | |
| "grad_norm": 6.413115501403809, | |
| "learning_rate": 1.759755370650472e-05, | |
| "loss": 0.3991, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 3.4652114597544337, | |
| "grad_norm": 10.340503692626953, | |
| "learning_rate": 1.7314324943067598e-05, | |
| "loss": 0.3823, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 3.4788540245566164, | |
| "grad_norm": 10.008712768554688, | |
| "learning_rate": 1.70327630461953e-05, | |
| "loss": 0.3564, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 3.4924965893587996, | |
| "grad_norm": 7.237100601196289, | |
| "learning_rate": 1.6752888704102304e-05, | |
| "loss": 0.4012, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 3.5061391541609823, | |
| "grad_norm": 7.369434833526611, | |
| "learning_rate": 1.6474722481007344e-05, | |
| "loss": 0.3622, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 3.519781718963165, | |
| "grad_norm": 10.581721305847168, | |
| "learning_rate": 1.619828481562229e-05, | |
| "loss": 0.3502, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 3.533424283765348, | |
| "grad_norm": 7.968098163604736, | |
| "learning_rate": 1.5923596019650517e-05, | |
| "loss": 0.4129, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 3.547066848567531, | |
| "grad_norm": 9.581936836242676, | |
| "learning_rate": 1.565067627629432e-05, | |
| "loss": 0.455, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 3.5607094133697137, | |
| "grad_norm": 7.407210350036621, | |
| "learning_rate": 1.5379545638772032e-05, | |
| "loss": 0.3905, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 3.5743519781718964, | |
| "grad_norm": 9.555901527404785, | |
| "learning_rate": 1.511022402884459e-05, | |
| "loss": 0.4603, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 3.587994542974079, | |
| "grad_norm": 5.677056312561035, | |
| "learning_rate": 1.4842731235351653e-05, | |
| "loss": 0.3654, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 3.601637107776262, | |
| "grad_norm": 8.808677673339844, | |
| "learning_rate": 1.4577086912757659e-05, | |
| "loss": 0.4481, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 3.6152796725784446, | |
| "grad_norm": 11.250635147094727, | |
| "learning_rate": 1.4313310579707697e-05, | |
| "loss": 0.4829, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 3.6289222373806274, | |
| "grad_norm": 8.87308406829834, | |
| "learning_rate": 1.405142161759327e-05, | |
| "loss": 0.3937, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 3.64256480218281, | |
| "grad_norm": 7.986633777618408, | |
| "learning_rate": 1.3791439269128274e-05, | |
| "loss": 0.3902, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 3.6562073669849933, | |
| "grad_norm": 6.5852155685424805, | |
| "learning_rate": 1.3533382636935092e-05, | |
| "loss": 0.4495, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 3.669849931787176, | |
| "grad_norm": 22.476903915405273, | |
| "learning_rate": 1.3277270682140996e-05, | |
| "loss": 0.4707, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 3.6834924965893587, | |
| "grad_norm": 6.103644371032715, | |
| "learning_rate": 1.3023122222984941e-05, | |
| "loss": 0.4189, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 3.6971350613915415, | |
| "grad_norm": 8.584049224853516, | |
| "learning_rate": 1.2770955933434906e-05, | |
| "loss": 0.3741, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 3.7107776261937246, | |
| "grad_norm": 8.539772033691406, | |
| "learning_rate": 1.2520790341815726e-05, | |
| "loss": 0.3724, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 3.7244201909959074, | |
| "grad_norm": 8.408806800842285, | |
| "learning_rate": 1.2272643829447723e-05, | |
| "loss": 0.4002, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 3.73806275579809, | |
| "grad_norm": 13.712347030639648, | |
| "learning_rate": 1.2026534629296168e-05, | |
| "loss": 0.4129, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 3.751705320600273, | |
| "grad_norm": 7.925304412841797, | |
| "learning_rate": 1.1782480824631478e-05, | |
| "loss": 0.3354, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 3.7653478854024556, | |
| "grad_norm": 6.753428936004639, | |
| "learning_rate": 1.154050034770057e-05, | |
| "loss": 0.3, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 3.7789904502046383, | |
| "grad_norm": 11.436506271362305, | |
| "learning_rate": 1.1300610978409301e-05, | |
| "loss": 0.4613, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 3.792633015006821, | |
| "grad_norm": 7.002927780151367, | |
| "learning_rate": 1.1062830343015998e-05, | |
| "loss": 0.3782, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 3.806275579809004, | |
| "grad_norm": 8.752483367919922, | |
| "learning_rate": 1.0827175912836352e-05, | |
| "loss": 0.3154, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 3.819918144611187, | |
| "grad_norm": 8.298131942749023, | |
| "learning_rate": 1.059366500295973e-05, | |
| "loss": 0.3941, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 3.8335607094133697, | |
| "grad_norm": 9.577731132507324, | |
| "learning_rate": 1.0362314770976858e-05, | |
| "loss": 0.4402, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 3.8472032742155524, | |
| "grad_norm": 8.360587120056152, | |
| "learning_rate": 1.0133142215719176e-05, | |
| "loss": 0.3588, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 3.8608458390177356, | |
| "grad_norm": 7.106631755828857, | |
| "learning_rate": 9.906164176009825e-06, | |
| "loss": 0.4171, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 3.8744884038199183, | |
| "grad_norm": 7.905004501342773, | |
| "learning_rate": 9.681397329426363e-06, | |
| "loss": 0.4008, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 3.888130968622101, | |
| "grad_norm": 7.430079460144043, | |
| "learning_rate": 9.458858191075358e-06, | |
| "loss": 0.2851, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 3.901773533424284, | |
| "grad_norm": 6.089357852935791, | |
| "learning_rate": 9.238563112378967e-06, | |
| "loss": 0.486, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 3.9154160982264665, | |
| "grad_norm": 10.433391571044922, | |
| "learning_rate": 9.02052827987339e-06, | |
| "loss": 0.3321, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 3.9290586630286493, | |
| "grad_norm": 11.072164535522461, | |
| "learning_rate": 8.804769714019619e-06, | |
| "loss": 0.2981, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 3.942701227830832, | |
| "grad_norm": 6.538973331451416, | |
| "learning_rate": 8.591303268026293e-06, | |
| "loss": 0.3412, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 3.956343792633015, | |
| "grad_norm": 6.748257637023926, | |
| "learning_rate": 8.380144626684829e-06, | |
| "loss": 0.3648, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 3.969986357435198, | |
| "grad_norm": 8.447965621948242, | |
| "learning_rate": 8.171309305216973e-06, | |
| "loss": 0.4028, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 3.9836289222373806, | |
| "grad_norm": 11.228553771972656, | |
| "learning_rate": 7.96481264813481e-06, | |
| "loss": 0.5365, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 3.9972714870395634, | |
| "grad_norm": 16.4576416015625, | |
| "learning_rate": 7.760669828113276e-06, | |
| "loss": 0.3312, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 4.0109140518417465, | |
| "grad_norm": 9.181668281555176, | |
| "learning_rate": 7.558895844875325e-06, | |
| "loss": 0.2586, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 4.024556616643929, | |
| "grad_norm": 6.005313873291016, | |
| "learning_rate": 7.359505524089843e-06, | |
| "loss": 0.174, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.038199181446112, | |
| "grad_norm": 9.592679977416992, | |
| "learning_rate": 7.162513516282236e-06, | |
| "loss": 0.1319, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 4.051841746248295, | |
| "grad_norm": 4.006359100341797, | |
| "learning_rate": 6.967934295758003e-06, | |
| "loss": 0.1098, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 4.0654843110504775, | |
| "grad_norm": 8.887368202209473, | |
| "learning_rate": 6.775782159539237e-06, | |
| "loss": 0.1351, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 4.07912687585266, | |
| "grad_norm": 4.626531600952148, | |
| "learning_rate": 6.586071226314046e-06, | |
| "loss": 0.2408, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 4.092769440654843, | |
| "grad_norm": 3.412660598754883, | |
| "learning_rate": 6.3988154353992285e-06, | |
| "loss": 0.1759, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.106412005457026, | |
| "grad_norm": 3.9307303428649902, | |
| "learning_rate": 6.214028545716071e-06, | |
| "loss": 0.2398, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 4.120054570259208, | |
| "grad_norm": 4.589839935302734, | |
| "learning_rate": 6.031724134779331e-06, | |
| "loss": 0.1491, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 4.133697135061391, | |
| "grad_norm": 5.797046184539795, | |
| "learning_rate": 5.851915597699638e-06, | |
| "loss": 0.1711, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 4.147339699863575, | |
| "grad_norm": 6.0500030517578125, | |
| "learning_rate": 5.674616146199277e-06, | |
| "loss": 0.1815, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 4.1609822646657575, | |
| "grad_norm": 5.424590587615967, | |
| "learning_rate": 5.499838807641413e-06, | |
| "loss": 0.2796, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.17462482946794, | |
| "grad_norm": 10.55384349822998, | |
| "learning_rate": 5.327596424072896e-06, | |
| "loss": 0.2307, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 4.188267394270123, | |
| "grad_norm": 11.20399284362793, | |
| "learning_rate": 5.157901651280672e-06, | |
| "loss": 0.1955, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 4.201909959072306, | |
| "grad_norm": 5.694499492645264, | |
| "learning_rate": 4.990766957861875e-06, | |
| "loss": 0.2332, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 4.215552523874488, | |
| "grad_norm": 3.623840093612671, | |
| "learning_rate": 4.826204624307665e-06, | |
| "loss": 0.1317, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 4.229195088676671, | |
| "grad_norm": 6.306783199310303, | |
| "learning_rate": 4.664226742100946e-06, | |
| "loss": 0.1772, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.242837653478854, | |
| "grad_norm": 5.530797004699707, | |
| "learning_rate": 4.504845212827848e-06, | |
| "loss": 0.1893, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 4.256480218281037, | |
| "grad_norm": 6.482370853424072, | |
| "learning_rate": 4.348071747303322e-06, | |
| "loss": 0.2459, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 4.270122783083219, | |
| "grad_norm": 6.513676643371582, | |
| "learning_rate": 4.193917864710599e-06, | |
| "loss": 0.1615, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 4.283765347885402, | |
| "grad_norm": 7.2312703132629395, | |
| "learning_rate": 4.042394891754846e-06, | |
| "loss": 0.1488, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 4.297407912687586, | |
| "grad_norm": 8.509649276733398, | |
| "learning_rate": 3.893513961830886e-06, | |
| "loss": 0.1507, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 4.311050477489768, | |
| "grad_norm": 8.615159034729004, | |
| "learning_rate": 3.74728601420518e-06, | |
| "loss": 0.1768, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 4.324693042291951, | |
| "grad_norm": 9.077027320861816, | |
| "learning_rate": 3.6037217932120272e-06, | |
| "loss": 0.3008, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 4.338335607094134, | |
| "grad_norm": 4.60724401473999, | |
| "learning_rate": 3.4628318474641344e-06, | |
| "loss": 0.1839, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 4.351978171896317, | |
| "grad_norm": 3.5988879203796387, | |
| "learning_rate": 3.3246265290775013e-06, | |
| "loss": 0.1681, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 4.365620736698499, | |
| "grad_norm": 3.6327638626098633, | |
| "learning_rate": 3.1891159929108074e-06, | |
| "loss": 0.2091, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.379263301500682, | |
| "grad_norm": 7.489198207855225, | |
| "learning_rate": 3.0563101958192677e-06, | |
| "loss": 0.175, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 4.392905866302865, | |
| "grad_norm": 7.076563835144043, | |
| "learning_rate": 2.9262188959230297e-06, | |
| "loss": 0.2715, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 4.406548431105048, | |
| "grad_norm": 4.808744430541992, | |
| "learning_rate": 2.7988516518901643e-06, | |
| "loss": 0.205, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 4.42019099590723, | |
| "grad_norm": 12.732096672058105, | |
| "learning_rate": 2.674217822234382e-06, | |
| "loss": 0.1804, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 4.433833560709413, | |
| "grad_norm": 4.898436546325684, | |
| "learning_rate": 2.5523265646273252e-06, | |
| "loss": 0.2301, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 4.447476125511596, | |
| "grad_norm": 3.7656116485595703, | |
| "learning_rate": 2.433186835225745e-06, | |
| "loss": 0.2665, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 4.461118690313779, | |
| "grad_norm": 4.772073745727539, | |
| "learning_rate": 2.316807388013431e-06, | |
| "loss": 0.1443, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 4.474761255115962, | |
| "grad_norm": 7.5682830810546875, | |
| "learning_rate": 2.203196774157972e-06, | |
| "loss": 0.2292, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 4.488403819918145, | |
| "grad_norm": 4.153563976287842, | |
| "learning_rate": 2.0923633413824663e-06, | |
| "loss": 0.1765, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 4.502046384720328, | |
| "grad_norm": 3.0278851985931396, | |
| "learning_rate": 1.98431523335215e-06, | |
| "loss": 0.1966, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 4.51568894952251, | |
| "grad_norm": 3.25349497795105, | |
| "learning_rate": 1.8790603890760328e-06, | |
| "loss": 0.2525, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 4.529331514324693, | |
| "grad_norm": 6.384403228759766, | |
| "learning_rate": 1.7766065423235624e-06, | |
| "loss": 0.1502, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 4.542974079126876, | |
| "grad_norm": 8.483981132507324, | |
| "learning_rate": 1.6769612210563834e-06, | |
| "loss": 0.2251, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 4.5566166439290585, | |
| "grad_norm": 5.477599620819092, | |
| "learning_rate": 1.5801317468751954e-06, | |
| "loss": 0.1724, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 4.570259208731241, | |
| "grad_norm": 6.450118541717529, | |
| "learning_rate": 1.4861252344817812e-06, | |
| "loss": 0.1416, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 4.583901773533424, | |
| "grad_norm": 8.463088035583496, | |
| "learning_rate": 1.3949485911562799e-06, | |
| "loss": 0.1507, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 4.597544338335607, | |
| "grad_norm": 5.667326927185059, | |
| "learning_rate": 1.3066085162496057e-06, | |
| "loss": 0.1486, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 4.61118690313779, | |
| "grad_norm": 8.173833847045898, | |
| "learning_rate": 1.2211115006912499e-06, | |
| "loss": 0.1823, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 4.624829467939973, | |
| "grad_norm": 7.300376892089844, | |
| "learning_rate": 1.1384638265123305e-06, | |
| "loss": 0.1146, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 4.638472032742156, | |
| "grad_norm": 5.938165664672852, | |
| "learning_rate": 1.0586715663840175e-06, | |
| "loss": 0.2206, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 4.6521145975443385, | |
| "grad_norm": 5.717035293579102, | |
| "learning_rate": 9.817405831713135e-07, | |
| "loss": 0.137, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 4.665757162346521, | |
| "grad_norm": 9.58558464050293, | |
| "learning_rate": 9.076765295022949e-07, | |
| "loss": 0.1969, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 4.679399727148704, | |
| "grad_norm": 5.858943462371826, | |
| "learning_rate": 8.364848473527698e-07, | |
| "loss": 0.1712, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 4.693042291950887, | |
| "grad_norm": 5.517107009887695, | |
| "learning_rate": 7.748725387928791e-07, | |
| "loss": 0.1606, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 4.7066848567530695, | |
| "grad_norm": 2.8044917583465576, | |
| "learning_rate": 7.09152598980536e-07, | |
| "loss": 0.1722, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 4.720327421555252, | |
| "grad_norm": 5.587821960449219, | |
| "learning_rate": 6.463196175536768e-07, | |
| "loss": 0.2592, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 4.733969986357435, | |
| "grad_norm": 8.284170150756836, | |
| "learning_rate": 5.863782112669647e-07, | |
| "loss": 0.2729, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 4.747612551159618, | |
| "grad_norm": 7.003023147583008, | |
| "learning_rate": 5.293327844118956e-07, | |
| "loss": 0.1888, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 4.7612551159618, | |
| "grad_norm": 1.9856537580490112, | |
| "learning_rate": 4.751875284932217e-07, | |
| "loss": 0.1367, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 4.774897680763983, | |
| "grad_norm": 6.205258846282959, | |
| "learning_rate": 4.2394642192095327e-07, | |
| "loss": 0.2003, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 4.788540245566167, | |
| "grad_norm": 5.503448009490967, | |
| "learning_rate": 3.7561322971803706e-07, | |
| "loss": 0.1503, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 4.8021828103683495, | |
| "grad_norm": 5.814334392547607, | |
| "learning_rate": 3.301915032437375e-07, | |
| "loss": 0.1425, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 4.815825375170532, | |
| "grad_norm": 7.47700309753418, | |
| "learning_rate": 2.8768457993266775e-07, | |
| "loss": 0.1265, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 4.829467939972715, | |
| "grad_norm": 6.920642375946045, | |
| "learning_rate": 2.480955830495679e-07, | |
| "loss": 0.2103, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 4.843110504774898, | |
| "grad_norm": 8.557290077209473, | |
| "learning_rate": 2.1142742145984442e-07, | |
| "loss": 0.2395, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 4.85675306957708, | |
| "grad_norm": 7.730421543121338, | |
| "learning_rate": 1.7768278941581617e-07, | |
| "loss": 0.1317, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 4.870395634379263, | |
| "grad_norm": 4.612334728240967, | |
| "learning_rate": 1.4686416635874445e-07, | |
| "loss": 0.1802, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 4.884038199181446, | |
| "grad_norm": 6.5652899742126465, | |
| "learning_rate": 1.1897381673666719e-07, | |
| "loss": 0.2484, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 4.897680763983629, | |
| "grad_norm": 8.53544807434082, | |
| "learning_rate": 9.40137898380078e-08, | |
| "loss": 0.1346, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 4.911323328785811, | |
| "grad_norm": 6.831415176391602, | |
| "learning_rate": 7.198591964099777e-08, | |
| "loss": 0.2097, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 4.924965893587995, | |
| "grad_norm": 11.197553634643555, | |
| "learning_rate": 5.289182467893561e-08, | |
| "loss": 0.1865, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 4.938608458390178, | |
| "grad_norm": 9.906291007995605, | |
| "learning_rate": 3.6732907921241956e-08, | |
| "loss": 0.2137, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 4.95225102319236, | |
| "grad_norm": 6.347635269165039, | |
| "learning_rate": 2.351035667038648e-08, | |
| "loss": 0.1699, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 4.965893587994543, | |
| "grad_norm": 2.5308990478515625, | |
| "learning_rate": 1.3225142474651009e-08, | |
| "loss": 0.247, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 4.979536152796726, | |
| "grad_norm": 8.235413551330566, | |
| "learning_rate": 5.878021056742178e-09, | |
| "loss": 0.1884, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 4.993178717598909, | |
| "grad_norm": 8.303121566772461, | |
| "learning_rate": 1.4695322582491956e-09, | |
| "loss": 0.186, | |
| "step": 3660 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3665, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500.0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.2102292765343744e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |