| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.991755976916735, | |
| "eval_steps": 500, | |
| "global_step": 6060, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0016488046166529267, | |
| "grad_norm": 380.0, | |
| "learning_rate": 3.300330033003301e-07, | |
| "loss": 35.4867, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.008244023083264633, | |
| "grad_norm": 308.0, | |
| "learning_rate": 1.65016501650165e-06, | |
| "loss": 34.8873, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.016488046166529265, | |
| "grad_norm": 306.0, | |
| "learning_rate": 3.3003300330033e-06, | |
| "loss": 34.9252, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0247320692497939, | |
| "grad_norm": 163.0, | |
| "learning_rate": 4.950495049504951e-06, | |
| "loss": 31.7188, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03297609233305853, | |
| "grad_norm": 110.5, | |
| "learning_rate": 6.6006600660066e-06, | |
| "loss": 28.5443, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.041220115416323165, | |
| "grad_norm": 72.5, | |
| "learning_rate": 8.250825082508252e-06, | |
| "loss": 24.1835, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0494641384995878, | |
| "grad_norm": 39.5, | |
| "learning_rate": 9.900990099009901e-06, | |
| "loss": 21.6514, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.057708161582852434, | |
| "grad_norm": 19.0, | |
| "learning_rate": 1.155115511551155e-05, | |
| "loss": 19.5766, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06595218466611706, | |
| "grad_norm": 16.25, | |
| "learning_rate": 1.32013201320132e-05, | |
| "loss": 18.5587, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0741962077493817, | |
| "grad_norm": 13.125, | |
| "learning_rate": 1.4851485148514851e-05, | |
| "loss": 17.2984, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.08244023083264633, | |
| "grad_norm": 9.375, | |
| "learning_rate": 1.6501650165016504e-05, | |
| "loss": 16.2291, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09068425391591096, | |
| "grad_norm": 7.5, | |
| "learning_rate": 1.8151815181518153e-05, | |
| "loss": 15.5459, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0989282769991756, | |
| "grad_norm": 5.1875, | |
| "learning_rate": 1.9801980198019803e-05, | |
| "loss": 15.0494, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.10717230008244023, | |
| "grad_norm": 3.546875, | |
| "learning_rate": 2.1452145214521452e-05, | |
| "loss": 14.5263, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.11541632316570487, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 2.31023102310231e-05, | |
| "loss": 14.0492, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1236603462489695, | |
| "grad_norm": 3.34375, | |
| "learning_rate": 2.4752475247524754e-05, | |
| "loss": 14.0437, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.13190436933223412, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 2.64026402640264e-05, | |
| "loss": 13.4269, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14014839241549876, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 2.8052805280528056e-05, | |
| "loss": 13.1438, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1483924154987634, | |
| "grad_norm": 5.4375, | |
| "learning_rate": 2.9702970297029702e-05, | |
| "loss": 12.8682, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.15663643858202803, | |
| "grad_norm": 7.5625, | |
| "learning_rate": 3.135313531353136e-05, | |
| "loss": 12.1682, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.16488046166529266, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 3.300330033003301e-05, | |
| "loss": 11.6708, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1731244847485573, | |
| "grad_norm": 13.875, | |
| "learning_rate": 3.465346534653465e-05, | |
| "loss": 10.6173, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.18136850783182193, | |
| "grad_norm": 20.375, | |
| "learning_rate": 3.6303630363036307e-05, | |
| "loss": 9.3535, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.18961253091508656, | |
| "grad_norm": 23.0, | |
| "learning_rate": 3.7953795379537956e-05, | |
| "loss": 7.0394, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.1978565539983512, | |
| "grad_norm": 19.0, | |
| "learning_rate": 3.9603960396039605e-05, | |
| "loss": 4.5084, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.20610057708161583, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 4.1254125412541255e-05, | |
| "loss": 2.7198, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.21434460016488047, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 4.2904290429042904e-05, | |
| "loss": 1.9952, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2225886232481451, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 4.455445544554456e-05, | |
| "loss": 1.7411, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.23083264633140974, | |
| "grad_norm": 0.95703125, | |
| "learning_rate": 4.62046204620462e-05, | |
| "loss": 1.5998, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.23907666941467437, | |
| "grad_norm": 1.0859375, | |
| "learning_rate": 4.785478547854786e-05, | |
| "loss": 1.5183, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.247320692497939, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 4.950495049504951e-05, | |
| "loss": 1.4489, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.25556471558120364, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 5.115511551155116e-05, | |
| "loss": 1.3924, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.26380873866446825, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 5.28052805280528e-05, | |
| "loss": 1.3648, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2720527617477329, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 5.445544554455446e-05, | |
| "loss": 1.3461, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.2802967848309975, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 5.610561056105611e-05, | |
| "loss": 1.3065, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2885408079142622, | |
| "grad_norm": 0.98828125, | |
| "learning_rate": 5.7755775577557755e-05, | |
| "loss": 1.2809, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.2967848309975268, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 5.9405940594059404e-05, | |
| "loss": 1.2647, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.30502885408079145, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 6.105610561056106e-05, | |
| "loss": 1.2387, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.31327287716405605, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 6.270627062706272e-05, | |
| "loss": 1.24, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3215169002473207, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 6.435643564356436e-05, | |
| "loss": 1.2108, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.3297609233305853, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 6.600660066006602e-05, | |
| "loss": 1.2026, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.33800494641385, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 6.765676567656766e-05, | |
| "loss": 1.1894, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.3462489694971146, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 6.93069306930693e-05, | |
| "loss": 1.2093, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.35449299258037925, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 7.095709570957097e-05, | |
| "loss": 1.1768, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.36273701566364386, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 7.260726072607261e-05, | |
| "loss": 1.1946, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.37098103874690846, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 7.425742574257426e-05, | |
| "loss": 1.1678, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.3792250618301731, | |
| "grad_norm": 1.5, | |
| "learning_rate": 7.590759075907591e-05, | |
| "loss": 1.1618, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.38746908491343773, | |
| "grad_norm": 0.84375, | |
| "learning_rate": 7.755775577557755e-05, | |
| "loss": 1.1585, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.3957131079967024, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 7.920792079207921e-05, | |
| "loss": 1.1519, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.403957131079967, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 8.085808580858087e-05, | |
| "loss": 1.1408, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.41220115416323166, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 8.250825082508251e-05, | |
| "loss": 1.138, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.42044517724649627, | |
| "grad_norm": 0.80859375, | |
| "learning_rate": 8.415841584158417e-05, | |
| "loss": 1.1375, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.42868920032976093, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 8.580858085808581e-05, | |
| "loss": 1.1193, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.43693322341302554, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 8.745874587458746e-05, | |
| "loss": 1.1178, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.4451772464962902, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 8.910891089108912e-05, | |
| "loss": 1.1038, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4534212695795548, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 9.075907590759076e-05, | |
| "loss": 1.1148, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.46166529266281947, | |
| "grad_norm": 0.84765625, | |
| "learning_rate": 9.24092409240924e-05, | |
| "loss": 1.112, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.4699093157460841, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 9.405940594059406e-05, | |
| "loss": 1.0882, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.47815333882934874, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 9.570957095709572e-05, | |
| "loss": 1.0873, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.48639736191261335, | |
| "grad_norm": 0.80078125, | |
| "learning_rate": 9.735973597359736e-05, | |
| "loss": 1.0982, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.494641384995878, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 9.900990099009902e-05, | |
| "loss": 1.074, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5028854080791426, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 0.00010066006600660067, | |
| "loss": 1.0719, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.5111294311624073, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 0.00010231023102310232, | |
| "loss": 1.0816, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5193734542456719, | |
| "grad_norm": 0.95703125, | |
| "learning_rate": 0.00010396039603960397, | |
| "loss": 1.0681, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.5276174773289365, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 0.0001056105610561056, | |
| "loss": 1.0689, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5358615004122012, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 0.00010726072607260727, | |
| "loss": 1.0712, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.5441055234954658, | |
| "grad_norm": 0.9375, | |
| "learning_rate": 0.00010891089108910893, | |
| "loss": 1.063, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5523495465787304, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 0.00011056105610561056, | |
| "loss": 1.0622, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.560593569661995, | |
| "grad_norm": 5.0, | |
| "learning_rate": 0.00011221122112211223, | |
| "loss": 1.0614, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5688375927452597, | |
| "grad_norm": 2.0, | |
| "learning_rate": 0.00011386138613861385, | |
| "loss": 1.0611, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.5770816158285244, | |
| "grad_norm": 1.75, | |
| "learning_rate": 0.00011551155115511551, | |
| "loss": 1.0451, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5853256389117889, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 0.00011716171617161718, | |
| "loss": 1.0506, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.5935696619950536, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 0.00011881188118811881, | |
| "loss": 1.0414, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.6018136850783182, | |
| "grad_norm": 2.703125, | |
| "learning_rate": 0.00012046204620462047, | |
| "loss": 1.0334, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.6100577081615829, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 0.00012211221122112212, | |
| "loss": 1.0388, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.6183017312448474, | |
| "grad_norm": 0.86328125, | |
| "learning_rate": 0.00012376237623762376, | |
| "loss": 1.0251, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.6265457543281121, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 0.00012541254125412543, | |
| "loss": 1.0315, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6347897774113768, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 0.00012706270627062708, | |
| "loss": 1.0342, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.6430338004946414, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 0.00012871287128712872, | |
| "loss": 1.0258, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.651277823577906, | |
| "grad_norm": 1.53125, | |
| "learning_rate": 0.00013036303630363036, | |
| "loss": 1.02, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.6595218466611706, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 0.00013201320132013203, | |
| "loss": 1.0053, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6677658697444353, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 0.00013366336633663367, | |
| "loss": 1.0217, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.6760098928277, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 0.00013531353135313532, | |
| "loss": 1.0066, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6842539159109645, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 0.00013696369636963699, | |
| "loss": 1.0141, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.6924979389942292, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 0.0001386138613861386, | |
| "loss": 1.0028, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.7007419620774938, | |
| "grad_norm": 3.875, | |
| "learning_rate": 0.00014026402640264027, | |
| "loss": 1.0207, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.7089859851607585, | |
| "grad_norm": 1.0, | |
| "learning_rate": 0.00014191419141914194, | |
| "loss": 1.0122, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.717230008244023, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 0.00014356435643564356, | |
| "loss": 1.0145, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.7254740313272877, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 0.00014521452145214523, | |
| "loss": 1.0031, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.7337180544105524, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 0.00014686468646864687, | |
| "loss": 0.9987, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.7419620774938169, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 0.0001485148514851485, | |
| "loss": 1.0024, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7502061005770816, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 0.00015016501650165018, | |
| "loss": 1.0048, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.7584501236603463, | |
| "grad_norm": 3.5, | |
| "learning_rate": 0.00015181518151815182, | |
| "loss": 1.0039, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7666941467436109, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 0.00015346534653465347, | |
| "loss": 1.0092, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.7749381698268755, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 0.0001551155115511551, | |
| "loss": 1.0078, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7831821929101401, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 0.00015676567656765678, | |
| "loss": 1.0005, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.7914262159934048, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 0.00015841584158415842, | |
| "loss": 0.9895, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7996702390766695, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 0.00016006600660066006, | |
| "loss": 0.9923, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.807914262159934, | |
| "grad_norm": 2.703125, | |
| "learning_rate": 0.00016171617161716173, | |
| "loss": 0.9996, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.8161582852431987, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 0.00016336633663366338, | |
| "loss": 0.9955, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.8244023083264633, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 0.00016501650165016502, | |
| "loss": 0.9931, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.832646331409728, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 1.0061, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.8408903544929925, | |
| "grad_norm": 2.640625, | |
| "learning_rate": 0.00016831683168316833, | |
| "loss": 1.0086, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.8491343775762572, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 0.00016996699669966997, | |
| "loss": 0.9886, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.8573784006595219, | |
| "grad_norm": 14.4375, | |
| "learning_rate": 0.00017161716171617162, | |
| "loss": 0.9933, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8656224237427865, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 0.00017326732673267329, | |
| "loss": 1.0033, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.8738664468260511, | |
| "grad_norm": 0.68359375, | |
| "learning_rate": 0.00017491749174917493, | |
| "loss": 0.9905, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8821104699093157, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 0.00017656765676567657, | |
| "loss": 0.9717, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.8903544929925804, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 0.00017821782178217824, | |
| "loss": 0.9656, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8985985160758451, | |
| "grad_norm": 3.625, | |
| "learning_rate": 0.00017986798679867986, | |
| "loss": 0.9827, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.9068425391591096, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 0.00018151815181518153, | |
| "loss": 0.9865, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.9150865622423743, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 0.0001831683168316832, | |
| "loss": 0.9815, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.9233305853256389, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 0.0001848184818481848, | |
| "loss": 0.9799, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.9315746084089035, | |
| "grad_norm": 3.4375, | |
| "learning_rate": 0.00018646864686468648, | |
| "loss": 0.9611, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.9398186314921682, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 0.00018811881188118812, | |
| "loss": 0.9652, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.9480626545754328, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 0.00018976897689768977, | |
| "loss": 0.9676, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.9563066776586975, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 0.00019141914191419144, | |
| "loss": 0.9692, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.964550700741962, | |
| "grad_norm": 6.4375, | |
| "learning_rate": 0.00019306930693069308, | |
| "loss": 0.9694, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.9727947238252267, | |
| "grad_norm": 1.25, | |
| "learning_rate": 0.00019471947194719472, | |
| "loss": 0.9823, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.9810387469084914, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 0.00019636963696369636, | |
| "loss": 0.97, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.989282769991756, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 0.00019801980198019803, | |
| "loss": 0.9746, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9975267930750206, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 0.00019966996699669968, | |
| "loss": 0.964, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.9991755976916735, | |
| "eval_loss": 2.485042095184326, | |
| "eval_runtime": 0.2808, | |
| "eval_samples_per_second": 35.608, | |
| "eval_steps_per_second": 3.561, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.0057708161582852, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 0.00019999973456433681, | |
| "loss": 0.9535, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.0140148392415498, | |
| "grad_norm": 2.6875, | |
| "learning_rate": 0.00019999865623437013, | |
| "loss": 0.9553, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.0222588623248146, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 0.00019999674842930876, | |
| "loss": 0.9556, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.030502885408079, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 0.00019999401116497763, | |
| "loss": 0.9746, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.0387469084913439, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 0.000199990444464082, | |
| "loss": 0.9639, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.0469909315746084, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 0.00019998604835620717, | |
| "loss": 0.9585, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.055234954657873, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 0.00019998082287781826, | |
| "loss": 0.9563, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.0634789777411378, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 0.00019997476807225985, | |
| "loss": 0.9489, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.0717230008244023, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 0.00019996788398975578, | |
| "loss": 0.9474, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0799670239076669, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 0.0001999601706874085, | |
| "loss": 0.9407, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.0882110469909316, | |
| "grad_norm": 2.875, | |
| "learning_rate": 0.00019995162822919883, | |
| "loss": 0.9514, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.0964550700741962, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 0.00019994225668598526, | |
| "loss": 0.9502, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.104699093157461, | |
| "grad_norm": 2.796875, | |
| "learning_rate": 0.0001999320561355035, | |
| "loss": 0.9502, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.1129431162407255, | |
| "grad_norm": 2.0, | |
| "learning_rate": 0.00019992102666236566, | |
| "loss": 0.9455, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.12118713932399, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 0.00019990916835805974, | |
| "loss": 0.9429, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.1294311624072548, | |
| "grad_norm": 0.74609375, | |
| "learning_rate": 0.00019989648132094873, | |
| "loss": 0.9348, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.1376751854905194, | |
| "grad_norm": 0.76171875, | |
| "learning_rate": 0.00019988296565626987, | |
| "loss": 0.939, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.145919208573784, | |
| "grad_norm": 0.93359375, | |
| "learning_rate": 0.0001998686214761337, | |
| "loss": 0.9374, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.1541632316570487, | |
| "grad_norm": 1.375, | |
| "learning_rate": 0.00019985344889952327, | |
| "loss": 0.9326, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.1624072547403133, | |
| "grad_norm": 0.91015625, | |
| "learning_rate": 0.00019983744805229296, | |
| "loss": 0.9308, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.1706512778235778, | |
| "grad_norm": 1.75, | |
| "learning_rate": 0.00019982061906716764, | |
| "loss": 0.9436, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.1788953009068426, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 0.00019980296208374143, | |
| "loss": 0.9369, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.1871393239901071, | |
| "grad_norm": 20.0, | |
| "learning_rate": 0.00019978447724847652, | |
| "loss": 0.9334, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.195383347073372, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 0.00019976516471470216, | |
| "loss": 0.9416, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.2036273701566365, | |
| "grad_norm": 8.75, | |
| "learning_rate": 0.0001997450246426131, | |
| "loss": 0.9382, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.211871393239901, | |
| "grad_norm": 0.86328125, | |
| "learning_rate": 0.0001997240571992685, | |
| "loss": 0.9315, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.2201154163231658, | |
| "grad_norm": 0.98046875, | |
| "learning_rate": 0.00019970226255859038, | |
| "loss": 0.9266, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.2283594394064303, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 0.0001996796409013623, | |
| "loss": 0.9299, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.2366034624896949, | |
| "grad_norm": 1.0, | |
| "learning_rate": 0.0001996561924152278, | |
| "loss": 0.9202, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.2448474855729597, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 0.00019963191729468888, | |
| "loss": 0.9149, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.2530915086562242, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 0.00019960681574110426, | |
| "loss": 0.9165, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.2613355317394888, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 0.00019958088796268793, | |
| "loss": 0.9188, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.2695795548227535, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 0.0001995541341745072, | |
| "loss": 0.9274, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.277823577906018, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 0.0001995265545984811, | |
| "loss": 0.9136, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.2860676009892829, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 0.00019949814946337838, | |
| "loss": 0.9251, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.2943116240725474, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 0.00019946891900481578, | |
| "loss": 0.9176, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.302555647155812, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 0.0001994388634652559, | |
| "loss": 0.9283, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.3107996702390767, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 0.00019940798309400526, | |
| "loss": 0.9221, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.3190436933223413, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 0.00019937627814721237, | |
| "loss": 0.9199, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.327287716405606, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 0.00019934374888786537, | |
| "loss": 0.9163, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.3355317394888706, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 0.00019931039558578997, | |
| "loss": 0.9181, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.3437757625721352, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 0.00019927621851764725, | |
| "loss": 0.9276, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.3520197856553997, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 0.00019924121796693127, | |
| "loss": 0.9199, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.3602638087386645, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 0.0001992053942239668, | |
| "loss": 0.9213, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.368507831821929, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 0.00019916874758590684, | |
| "loss": 0.9228, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.3767518549051938, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 0.00019913127835673023, | |
| "loss": 0.9149, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.3849958779884584, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 0.00019909298684723904, | |
| "loss": 0.9086, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.393239901071723, | |
| "grad_norm": 1.125, | |
| "learning_rate": 0.00019905387337505612, | |
| "loss": 0.9092, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.4014839241549877, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 0.0001990139382646223, | |
| "loss": 0.9041, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.4097279472382522, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 0.00019897318184719385, | |
| "loss": 0.9093, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.417971970321517, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 0.00019893160446083963, | |
| "loss": 0.909, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.4262159934047816, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 0.00019888920645043831, | |
| "loss": 0.9014, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.434460016488046, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 0.00019884598816767563, | |
| "loss": 0.9036, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.4427040395713109, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 0.00019880194997104123, | |
| "loss": 0.8999, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.4509480626545754, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 0.00019875709222582594, | |
| "loss": 0.9, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.45919208573784, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 0.00019871141530411853, | |
| "loss": 0.8955, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.4674361088211048, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 0.00019866491958480284, | |
| "loss": 0.9042, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.4756801319043693, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 0.00019861760545355442, | |
| "loss": 0.9177, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.4839241549876339, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 0.00019856947330283752, | |
| "loss": 0.8974, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.4921681780708986, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 0.00019852052353190166, | |
| "loss": 0.9064, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.5004122011541632, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 0.0001984707565467785, | |
| "loss": 0.9086, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.508656224237428, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 0.00019842017276027832, | |
| "loss": 0.9069, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.5169002473206925, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 0.00019836877259198662, | |
| "loss": 0.898, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.525144270403957, | |
| "grad_norm": 2.484375, | |
| "learning_rate": 0.0001983165564682608, | |
| "loss": 0.8999, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.5333882934872216, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 0.00019826352482222638, | |
| "loss": 0.8987, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.5416323165704864, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 0.00019820967809377357, | |
| "loss": 0.8791, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 1.5498763396537512, | |
| "grad_norm": 0.80859375, | |
| "learning_rate": 0.00019815501672955358, | |
| "loss": 0.8887, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.5581203627370157, | |
| "grad_norm": 6.0, | |
| "learning_rate": 0.0001980995411829749, | |
| "loss": 0.8955, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.5663643858202803, | |
| "grad_norm": 0.8984375, | |
| "learning_rate": 0.00019804325191419956, | |
| "loss": 0.8991, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.5746084089035448, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 0.00019798614939013932, | |
| "loss": 0.8916, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.5828524319868096, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 0.00019792823408445174, | |
| "loss": 0.9048, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.5910964550700744, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 0.0001978695064775363, | |
| "loss": 0.8828, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 1.599340478153339, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 0.00019780996705653044, | |
| "loss": 0.8864, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.6075845012366035, | |
| "grad_norm": 0.99609375, | |
| "learning_rate": 0.00019774961631530545, | |
| "loss": 0.8908, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.615828524319868, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 0.0001976884547544624, | |
| "loss": 0.8853, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.6240725474031328, | |
| "grad_norm": 2.84375, | |
| "learning_rate": 0.0001976264828813281, | |
| "loss": 0.8835, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 1.6323165704863973, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 0.00019756370120995066, | |
| "loss": 0.8817, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.640560593569662, | |
| "grad_norm": 27.25, | |
| "learning_rate": 0.0001975001102610954, | |
| "loss": 0.8972, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 1.6488046166529267, | |
| "grad_norm": 9.75, | |
| "learning_rate": 0.0001974357105622405, | |
| "loss": 0.9076, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.6570486397361912, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 0.0001973705026475726, | |
| "loss": 0.9001, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 1.6652926628194558, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 0.00019730448705798239, | |
| "loss": 0.9172, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.6735366859027205, | |
| "grad_norm": 1.375, | |
| "learning_rate": 0.0001972376643410601, | |
| "loss": 0.8945, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 1.6817807089859853, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 0.00019717003505109095, | |
| "loss": 0.8857, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.6900247320692499, | |
| "grad_norm": 1.4375, | |
| "learning_rate": 0.00019710159974905064, | |
| "loss": 0.8852, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.6982687551525144, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 0.00019703235900260055, | |
| "loss": 0.8795, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.706512778235779, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 0.00019696231338608316, | |
| "loss": 0.8926, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 1.7147568013190437, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 0.00019689146348051719, | |
| "loss": 0.8927, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.7230008244023083, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 0.0001968198098735929, | |
| "loss": 0.8762, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 1.731244847485573, | |
| "grad_norm": 6.75, | |
| "learning_rate": 0.0001967473531596671, | |
| "loss": 0.8886, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.7394888705688376, | |
| "grad_norm": 12.125, | |
| "learning_rate": 0.00019667409393975822, | |
| "loss": 0.8865, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 1.7477328936521022, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 0.00019660003282154147, | |
| "loss": 0.887, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.7559769167353667, | |
| "grad_norm": 0.84765625, | |
| "learning_rate": 0.00019652517041934356, | |
| "loss": 0.8669, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 1.7642209398186315, | |
| "grad_norm": 0.7890625, | |
| "learning_rate": 0.00019644950735413788, | |
| "loss": 0.8774, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.7724649629018963, | |
| "grad_norm": 0.98828125, | |
| "learning_rate": 0.00019637304425353916, | |
| "loss": 0.8717, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.7807089859851608, | |
| "grad_norm": 0.7578125, | |
| "learning_rate": 0.0001962957817517982, | |
| "loss": 0.8769, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.7889530090684254, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 0.0001962177204897969, | |
| "loss": 0.872, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 1.79719703215169, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 0.0001961388611150427, | |
| "loss": 0.8727, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.8054410552349547, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 0.00019605920428166323, | |
| "loss": 0.8671, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 1.8136850783182195, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 0.00019597875065040094, | |
| "loss": 0.8927, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.821929101401484, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 0.00019589750088860766, | |
| "loss": 0.881, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 1.8301731244847486, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 0.000195815455670239, | |
| "loss": 0.8793, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.838417147568013, | |
| "grad_norm": 3.890625, | |
| "learning_rate": 0.00019573261567584874, | |
| "loss": 0.8795, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 1.8466611706512777, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 0.00019564898159258324, | |
| "loss": 0.8933, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.8549051937345424, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 0.00019556455411417573, | |
| "loss": 0.8626, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.8631492168178072, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 0.0001954793339409405, | |
| "loss": 0.8616, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.8713932399010718, | |
| "grad_norm": 2.625, | |
| "learning_rate": 0.00019539332177976714, | |
| "loss": 0.8693, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 1.8796372629843363, | |
| "grad_norm": 0.875, | |
| "learning_rate": 0.00019530651834411474, | |
| "loss": 0.8659, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.8878812860676009, | |
| "grad_norm": 6.0, | |
| "learning_rate": 0.00019521892435400587, | |
| "loss": 0.8666, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 1.8961253091508656, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 0.00019513054053602055, | |
| "loss": 0.8601, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.9043693322341304, | |
| "grad_norm": 2.125, | |
| "learning_rate": 0.00019504136762329047, | |
| "loss": 0.8631, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 1.912613355317395, | |
| "grad_norm": 3.296875, | |
| "learning_rate": 0.00019495140635549261, | |
| "loss": 0.8833, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.9208573784006595, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 0.00019486065747884333, | |
| "loss": 0.8555, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 1.929101401483924, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 0.0001947691217460921, | |
| "loss": 0.8602, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.9373454245671888, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 0.0001946767999165152, | |
| "loss": 0.8553, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.9455894476504534, | |
| "grad_norm": 0.94921875, | |
| "learning_rate": 0.00019458369275590954, | |
| "loss": 0.8588, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.9538334707337182, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 0.00019448980103658613, | |
| "loss": 0.8529, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 1.9620774938169827, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 0.00019439512553736394, | |
| "loss": 0.8441, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.9703215169002473, | |
| "grad_norm": 0.83984375, | |
| "learning_rate": 0.0001942996670435632, | |
| "loss": 0.8526, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 1.9785655399835118, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 0.0001942034263469989, | |
| "loss": 0.8547, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.9868095630667766, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 0.0001941064042459745, | |
| "loss": 0.8686, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 1.9950535861500414, | |
| "grad_norm": 0.7734375, | |
| "learning_rate": 0.00019400860154527493, | |
| "loss": 0.8499, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.4393434524536133, | |
| "eval_runtime": 0.2359, | |
| "eval_samples_per_second": 42.391, | |
| "eval_steps_per_second": 4.239, | |
| "step": 1213 | |
| }, | |
| { | |
| "epoch": 2.003297609233306, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 0.0001939100190561601, | |
| "loss": 0.8486, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 2.0115416323165705, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 0.00019381065759635822, | |
| "loss": 0.8375, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.019785655399835, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 0.0001937105179900589, | |
| "loss": 0.8531, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.0280296784830996, | |
| "grad_norm": 1.75, | |
| "learning_rate": 0.00019360960106790643, | |
| "loss": 0.8369, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.0362737015663646, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 0.00019350790766699282, | |
| "loss": 0.8276, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 2.044517724649629, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 0.0001934054386308508, | |
| "loss": 0.8289, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.0527617477328937, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.00019330219480944694, | |
| "loss": 0.8292, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 2.061005770816158, | |
| "grad_norm": 0.828125, | |
| "learning_rate": 0.0001931981770591745, | |
| "loss": 0.8305, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.0692497938994228, | |
| "grad_norm": 0.77734375, | |
| "learning_rate": 0.00019309338624284644, | |
| "loss": 0.8243, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 2.0774938169826878, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 0.00019298782322968815, | |
| "loss": 0.8225, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.0857378400659523, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 0.0001928814888953303, | |
| "loss": 0.8212, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 2.093981863149217, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 0.0001927743841218016, | |
| "loss": 0.8188, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.1022258862324814, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 0.00019266650979752136, | |
| "loss": 0.8209, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 2.110469909315746, | |
| "grad_norm": 8.25, | |
| "learning_rate": 0.00019255786681729225, | |
| "loss": 0.8242, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.1187139323990105, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 0.00019244845608229293, | |
| "loss": 0.828, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 2.1269579554822755, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 0.00019233827850007027, | |
| "loss": 0.8159, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.13520197856554, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 0.00019222733498453222, | |
| "loss": 0.8196, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 2.1434460016488046, | |
| "grad_norm": 0.84765625, | |
| "learning_rate": 0.00019211562645594002, | |
| "loss": 0.8231, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.151690024732069, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.00019200315384090044, | |
| "loss": 0.8073, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 2.1599340478153337, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 0.00019188991807235844, | |
| "loss": 0.8255, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.1681780708985987, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 0.0001917759200895891, | |
| "loss": 0.8185, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 2.1764220939818633, | |
| "grad_norm": 8.0, | |
| "learning_rate": 0.00019166116083819002, | |
| "loss": 0.8174, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.184666117065128, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 0.00019154564127007336, | |
| "loss": 0.8263, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 2.1929101401483924, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 0.0001914293623434581, | |
| "loss": 0.8333, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.201154163231657, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 0.00019131232502286188, | |
| "loss": 0.8227, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 2.209398186314922, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 0.00019119453027909323, | |
| "loss": 0.8123, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.2176422093981865, | |
| "grad_norm": 0.96484375, | |
| "learning_rate": 0.0001910759790892433, | |
| "loss": 0.8129, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 2.225886232481451, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 0.0001909566724366779, | |
| "loss": 0.8101, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.2341302555647156, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 0.00019083661131102933, | |
| "loss": 0.8205, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 2.24237427864798, | |
| "grad_norm": 0.9921875, | |
| "learning_rate": 0.00019071579670818808, | |
| "loss": 0.8228, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.2506183017312447, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00019059422963029464, | |
| "loss": 0.8123, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 2.2588623248145097, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 0.00019047191108573125, | |
| "loss": 0.8227, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.267106347897774, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 0.00019034884208911335, | |
| "loss": 0.814, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 2.2753503709810388, | |
| "grad_norm": 0.78515625, | |
| "learning_rate": 0.00019022502366128135, | |
| "loss": 0.819, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.2835943940643033, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 0.00019010045682929213, | |
| "loss": 0.8074, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 2.291838417147568, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 0.00018997514262641035, | |
| "loss": 0.8224, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.300082440230833, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 0.0001898490820921001, | |
| "loss": 0.8096, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 2.3083264633140974, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00018972227627201617, | |
| "loss": 0.8102, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.316570486397362, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.0001895947262179954, | |
| "loss": 0.8113, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 2.3248145094806265, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 0.00018946643298804793, | |
| "loss": 0.8109, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.333058532563891, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.00018933739764634847, | |
| "loss": 0.809, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 2.3413025556471556, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 0.0001892076212632274, | |
| "loss": 0.8153, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.3495465787304206, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 0.00018907710491516199, | |
| "loss": 0.8161, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 2.357790601813685, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 0.00018894584968476733, | |
| "loss": 0.8141, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.3660346248969497, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 0.00018881385666078755, | |
| "loss": 0.8102, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 2.3742786479802143, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.00018868112693808665, | |
| "loss": 0.8124, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.382522671063479, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 0.00018854766161763932, | |
| "loss": 0.8033, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 2.390766694146744, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00018841346180652213, | |
| "loss": 0.812, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.3990107172300084, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 0.00018827852861790398, | |
| "loss": 0.8059, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 2.407254740313273, | |
| "grad_norm": 0.70703125, | |
| "learning_rate": 0.00018814286317103714, | |
| "loss": 0.8021, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.4154987633965375, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 0.00018800646659124782, | |
| "loss": 0.8036, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 2.423742786479802, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 0.00018786934000992688, | |
| "loss": 0.8045, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.4319868095630666, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.00018773148456452046, | |
| "loss": 0.8108, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 2.4402308326463316, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 0.00018759290139852048, | |
| "loss": 0.8097, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.448474855729596, | |
| "grad_norm": 0.79296875, | |
| "learning_rate": 0.00018745359166145523, | |
| "loss": 0.8052, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 2.4567188788128607, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 0.00018731355650887985, | |
| "loss": 0.8016, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.464962901896125, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 0.00018717279710236666, | |
| "loss": 0.8077, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 2.4732069249793898, | |
| "grad_norm": 0.9921875, | |
| "learning_rate": 0.00018703131460949554, | |
| "loss": 0.8031, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.4814509480626548, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 0.00018688911020384432, | |
| "loss": 0.8062, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 2.4896949711459193, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 0.000186746185064979, | |
| "loss": 0.8156, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.497938994229184, | |
| "grad_norm": 0.77734375, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 0.8083, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 2.5061830173124484, | |
| "grad_norm": 0.70703125, | |
| "learning_rate": 0.00018645817733575193, | |
| "loss": 0.812, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.514427040395713, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 0.00018631309713437467, | |
| "loss": 0.796, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 2.5226710634789775, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 0.0001861673009777325, | |
| "loss": 0.7988, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.5309150865622425, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 0.00018602079007518438, | |
| "loss": 0.7988, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 2.539159109645507, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.00018587356564201817, | |
| "loss": 0.8045, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.5474031327287716, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.0001857256288994402, | |
| "loss": 0.8112, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 2.555647155812036, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 0.00018557698107456549, | |
| "loss": 0.808, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.563891178895301, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 0.00018542762340040722, | |
| "loss": 0.7958, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 2.5721352019785657, | |
| "grad_norm": 0.859375, | |
| "learning_rate": 0.00018527755711586678, | |
| "loss": 0.8008, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.5803792250618303, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 0.00018512678346572337, | |
| "loss": 0.7995, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 2.588623248145095, | |
| "grad_norm": 0.734375, | |
| "learning_rate": 0.00018497530370062363, | |
| "loss": 0.7974, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.5968672712283594, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.0001848231190770714, | |
| "loss": 0.7929, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 2.605111294311624, | |
| "grad_norm": 0.78125, | |
| "learning_rate": 0.00018467023085741717, | |
| "loss": 0.8014, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.6133553173948885, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 0.00018451664030984773, | |
| "loss": 0.7944, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 2.6215993404781535, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 0.00018436234870837547, | |
| "loss": 0.7937, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.629843363561418, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 0.00018420735733282807, | |
| "loss": 0.7983, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 2.6380873866446826, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 0.00018405166746883762, | |
| "loss": 0.7924, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.646331409727947, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.00018389528040783012, | |
| "loss": 0.7953, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 2.654575432811212, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 0.00018373819744701476, | |
| "loss": 0.7893, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.6628194558944767, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.00018358041988937305, | |
| "loss": 0.7945, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 2.671063478977741, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 0.00018342194904364813, | |
| "loss": 0.7894, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.6793075020610058, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 0.00018326278622433386, | |
| "loss": 0.7925, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 2.6875515251442703, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 0.00018310293275166392, | |
| "loss": 0.7978, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.695795548227535, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 0.00018294238995160094, | |
| "loss": 0.792, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 2.7040395713107994, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 0.00018278115915582526, | |
| "loss": 0.8069, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.7122835943940644, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 0.0001826192417017242, | |
| "loss": 0.8048, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 2.720527617477329, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 0.00018245663893238075, | |
| "loss": 0.8009, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.7287716405605935, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 0.0001822933521965625, | |
| "loss": 0.7903, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 2.737015663643858, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.00018212938284871047, | |
| "loss": 0.7917, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.745259686727123, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 0.00018196473224892784, | |
| "loss": 0.7886, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 2.7535037098103876, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 0.0001817994017629687, | |
| "loss": 0.7933, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.761747732893652, | |
| "grad_norm": 0.78515625, | |
| "learning_rate": 0.00018163339276222666, | |
| "loss": 0.792, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 2.7699917559769167, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 0.00018146670662372354, | |
| "loss": 0.7825, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.7782357790601813, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 0.0001812993447300979, | |
| "loss": 0.7929, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 2.786479802143446, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 0.00018113130846959368, | |
| "loss": 0.7925, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.7947238252267104, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 0.0001809625992360485, | |
| "loss": 0.7888, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 2.8029678483099754, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.00018079321842888227, | |
| "loss": 0.7995, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.81121187139324, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 0.00018062316745308542, | |
| "loss": 0.7939, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 2.8194558944765045, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 0.0001804524477192075, | |
| "loss": 0.79, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.827699917559769, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 0.0001802810606433451, | |
| "loss": 0.7927, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 2.835943940643034, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 0.00018010900764713048, | |
| "loss": 0.796, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.8441879637262986, | |
| "grad_norm": 0.75, | |
| "learning_rate": 0.0001799362901577196, | |
| "loss": 0.7921, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 2.852431986809563, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.00017976290960778024, | |
| "loss": 0.79, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.8606760098928277, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 0.0001795888674354802, | |
| "loss": 0.7927, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 2.868920032976092, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 0.00017941416508447536, | |
| "loss": 0.7917, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.8771640560593568, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 0.0001792388040038977, | |
| "loss": 0.7905, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 2.8854080791426218, | |
| "grad_norm": 0.7578125, | |
| "learning_rate": 0.00017906278564834324, | |
| "loss": 0.7934, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.8936521022258863, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 0.00017888611147786002, | |
| "loss": 0.7957, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 2.901896125309151, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00017870878295793598, | |
| "loss": 0.7793, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.9101401483924154, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 0.0001785308015594868, | |
| "loss": 0.7912, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 2.91838417147568, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 0.00017835216875884368, | |
| "loss": 0.7842, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.926628194558945, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 0.00017817288603774116, | |
| "loss": 0.784, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 2.9348722176422095, | |
| "grad_norm": 0.828125, | |
| "learning_rate": 0.00017799295488330467, | |
| "loss": 0.7934, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.943116240725474, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00017781237678803847, | |
| "loss": 0.7867, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 2.9513602638087386, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 0.00017763115324981294, | |
| "loss": 0.7911, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.959604286892003, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 0.00017744928577185243, | |
| "loss": 0.7914, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 2.9678483099752677, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 0.00017726677586272263, | |
| "loss": 0.7917, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.9760923330585327, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 0.00017708362503631814, | |
| "loss": 0.7819, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 2.9843363561417973, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 0.00017689983481184989, | |
| "loss": 0.7842, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.992580379225062, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.00017671540671383243, | |
| "loss": 0.7939, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 2.9991755976916736, | |
| "eval_loss": 2.4241690635681152, | |
| "eval_runtime": 0.2578, | |
| "eval_samples_per_second": 38.793, | |
| "eval_steps_per_second": 3.879, | |
| "step": 1819 | |
| }, | |
| { | |
| "epoch": 3.0008244023083264, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00017653034227207152, | |
| "loss": 0.7885, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 3.009068425391591, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 0.00017634464302165124, | |
| "loss": 0.772, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 3.017312448474856, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 0.0001761583105029213, | |
| "loss": 0.7668, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 3.0255564715581205, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 0.00017597134626148427, | |
| "loss": 0.77, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 3.033800494641385, | |
| "grad_norm": 0.88671875, | |
| "learning_rate": 0.0001757837518481829, | |
| "loss": 0.7713, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 3.0420445177246496, | |
| "grad_norm": 0.74609375, | |
| "learning_rate": 0.00017559552881908695, | |
| "loss": 0.7748, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 3.050288540807914, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.00017540667873548063, | |
| "loss": 0.7653, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 3.058532563891179, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00017521720316384935, | |
| "loss": 0.7706, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 3.0667765869744437, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 0.00017502710367586687, | |
| "loss": 0.7633, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 3.075020610057708, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.00017483638184838239, | |
| "loss": 0.7568, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 3.0832646331409728, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 0.0001746450392634071, | |
| "loss": 0.757, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 3.0915086562242373, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.0001744530775081015, | |
| "loss": 0.7701, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 3.099752679307502, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.00017426049817476197, | |
| "loss": 0.7717, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 3.107996702390767, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 0.00017406730286080753, | |
| "loss": 0.7647, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 3.1162407254740314, | |
| "grad_norm": 0.5, | |
| "learning_rate": 0.00017387349316876666, | |
| "loss": 0.7618, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 3.124484748557296, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.00017367907070626424, | |
| "loss": 0.7712, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 3.1327287716405605, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00017348403708600772, | |
| "loss": 0.7635, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.140972794723825, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 0.0001732883939257742, | |
| "loss": 0.7591, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 3.14921681780709, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.00017309214284839678, | |
| "loss": 0.7664, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 3.1574608408903546, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00017289528548175114, | |
| "loss": 0.7633, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 3.165704863973619, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.00017269782345874203, | |
| "loss": 0.7676, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 3.1739488870568837, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 0.0001724997584172898, | |
| "loss": 0.7712, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 3.1821929101401483, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.00017230109200031668, | |
| "loss": 0.7631, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 3.190436933223413, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.00017210182585573327, | |
| "loss": 0.7664, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 3.198680956306678, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 0.00017190196163642483, | |
| "loss": 0.7653, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.2069249793899424, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 0.0001717015010002376, | |
| "loss": 0.7677, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 3.215169002473207, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00017150044560996488, | |
| "loss": 0.7628, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.2234130255564715, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 0.00017129879713333356, | |
| "loss": 0.7604, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 3.231657048639736, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 0.00017109655724298995, | |
| "loss": 0.7664, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.239901071723001, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 0.00017089372761648616, | |
| "loss": 0.7679, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 3.2481450948062656, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00017069030993626603, | |
| "loss": 0.7621, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 3.25638911788953, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 0.00017048630588965117, | |
| "loss": 0.7747, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 3.2646331409727947, | |
| "grad_norm": 0.625, | |
| "learning_rate": 0.00017028171716882714, | |
| "loss": 0.7655, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.272877164056059, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 0.00017007654547082922, | |
| "loss": 0.768, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 3.281121187139324, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 0.00016987079249752843, | |
| "loss": 0.7631, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 3.2893652102225888, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 0.00016966445995561727, | |
| "loss": 0.7686, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 3.2976092333058533, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 0.00016945754955659595, | |
| "loss": 0.7695, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.305853256389118, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00016925006301675763, | |
| "loss": 0.7548, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 3.3140972794723824, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.0001690420020571747, | |
| "loss": 0.7642, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 3.322341302555647, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 0.00016883336840368412, | |
| "loss": 0.7706, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 3.330585325638912, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.0001686241637868734, | |
| "loss": 0.7693, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.3388293487221765, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00016841438994206595, | |
| "loss": 0.7616, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 3.347073371805441, | |
| "grad_norm": 0.99609375, | |
| "learning_rate": 0.0001682040486093071, | |
| "loss": 0.7661, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 3.3553173948887056, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 0.00016799314153334916, | |
| "loss": 0.7543, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 3.36356141797197, | |
| "grad_norm": 0.92578125, | |
| "learning_rate": 0.00016778167046363734, | |
| "loss": 0.757, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.371805441055235, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 0.00016756963715429502, | |
| "loss": 0.7647, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 3.3800494641384997, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00016735704336410943, | |
| "loss": 0.7562, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.3882934872217643, | |
| "grad_norm": 0.70703125, | |
| "learning_rate": 0.0001671438908565167, | |
| "loss": 0.7573, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 3.396537510305029, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00016693018139958763, | |
| "loss": 0.7585, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.4047815333882934, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 0.00016671591676601272, | |
| "loss": 0.7538, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 3.413025556471558, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 0.00016650109873308765, | |
| "loss": 0.7635, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 3.421269579554823, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.00016628572908269841, | |
| "loss": 0.7605, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 3.4295136026380875, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.00016606980960130665, | |
| "loss": 0.7511, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.437757625721352, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 0.00016585334207993476, | |
| "loss": 0.757, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 3.4460016488046166, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 0.00016563632831415102, | |
| "loss": 0.7616, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 3.454245671887881, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.00016541877010405477, | |
| "loss": 0.7605, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 3.462489694971146, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 0.00016520066925426144, | |
| "loss": 0.7564, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.4707337180544107, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 0.00016498202757388758, | |
| "loss": 0.7627, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 3.478977741137675, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 0.0001647628468765358, | |
| "loss": 0.7514, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 3.4872217642209398, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 0.0001645431289802799, | |
| "loss": 0.7616, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 3.4954657873042043, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00016432287570764952, | |
| "loss": 0.7639, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.503709810387469, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 0.0001641020888856153, | |
| "loss": 0.7642, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 3.511953833470734, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 0.00016388077034557355, | |
| "loss": 0.7511, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 3.5201978565539984, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 0.0001636589219233311, | |
| "loss": 0.7513, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 3.528441879637263, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 0.00016343654545909007, | |
| "loss": 0.7568, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.5366859027205275, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 0.00016321364279743266, | |
| "loss": 0.7562, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 3.5449299258037925, | |
| "grad_norm": 0.70703125, | |
| "learning_rate": 0.00016299021578730579, | |
| "loss": 0.7591, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.553173948887057, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 0.00016276626628200568, | |
| "loss": 0.7665, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 3.5614179719703216, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00016254179613916278, | |
| "loss": 0.7604, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 3.569661995053586, | |
| "grad_norm": 1.0, | |
| "learning_rate": 0.000162316807220726, | |
| "loss": 0.7504, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 3.5779060181368507, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 0.00016209130139294744, | |
| "loss": 0.7646, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 3.5861500412201153, | |
| "grad_norm": 0.69921875, | |
| "learning_rate": 0.00016186528052636692, | |
| "loss": 0.7562, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 3.59439406430338, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 0.00016163874649579647, | |
| "loss": 0.7501, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 3.602638087386645, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.00016141170118030463, | |
| "loss": 0.7548, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 3.6108821104699094, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 0.0001611841464632011, | |
| "loss": 0.7582, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 3.619126133553174, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00016095608423202098, | |
| "loss": 0.7517, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 3.6273701566364385, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.00016072751637850904, | |
| "loss": 0.7563, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.6356141797197035, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 0.00016049844479860422, | |
| "loss": 0.7566, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 3.643858202802968, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.00016026887139242372, | |
| "loss": 0.7515, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 3.6521022258862326, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 0.0001600387980642474, | |
| "loss": 0.754, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 3.660346248969497, | |
| "grad_norm": 0.74609375, | |
| "learning_rate": 0.0001598082267225018, | |
| "loss": 0.7608, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.6685902720527617, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 0.0001595771592797445, | |
| "loss": 0.7574, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 3.676834295136026, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.0001593455976526482, | |
| "loss": 0.7526, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 3.6850783182192908, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.0001591135437619847, | |
| "loss": 0.7546, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 3.6933223413025558, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 0.00015888099953260905, | |
| "loss": 0.7574, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.7015663643858203, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 0.0001586479668934437, | |
| "loss": 0.7548, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 3.709810387469085, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 0.0001584144477774623, | |
| "loss": 0.7519, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.7180544105523494, | |
| "grad_norm": 0.75, | |
| "learning_rate": 0.0001581804441216738, | |
| "loss": 0.761, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 3.7262984336356144, | |
| "grad_norm": 0.86328125, | |
| "learning_rate": 0.00015794595786710632, | |
| "loss": 0.7552, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 3.734542456718879, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.00015771099095879108, | |
| "loss": 0.7573, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 3.7427864798021435, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 0.00015747554534574626, | |
| "loss": 0.753, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 3.751030502885408, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 0.0001572396229809608, | |
| "loss": 0.7587, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 3.7592745259686726, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 0.00015700322582137827, | |
| "loss": 0.7505, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 3.767518549051937, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.0001567663558278806, | |
| "loss": 0.747, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 3.7757625721352017, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 0.0001565290149652718, | |
| "loss": 0.763, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 3.7840065952184667, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 0.00015629120520226165, | |
| "loss": 0.7547, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 3.7922506183017313, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 0.00015605292851144942, | |
| "loss": 0.7537, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.800494641384996, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.00015581418686930743, | |
| "loss": 0.754, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 3.8087386644682604, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 0.00015557498225616487, | |
| "loss": 0.7407, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 3.8169826875515254, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00015533531665619098, | |
| "loss": 0.7556, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 3.82522671063479, | |
| "grad_norm": 0.97265625, | |
| "learning_rate": 0.00015509519205737896, | |
| "loss": 0.7516, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 3.8334707337180545, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 0.0001548546104515294, | |
| "loss": 0.7506, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 3.841714756801319, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.0001546135738342335, | |
| "loss": 0.7524, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 3.8499587798845836, | |
| "grad_norm": 0.81640625, | |
| "learning_rate": 0.0001543720842048569, | |
| "loss": 0.748, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 3.858202802967848, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.00015413014356652286, | |
| "loss": 0.7503, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 3.8664468260511127, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00015388775392609564, | |
| "loss": 0.754, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 3.8746908491343777, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 0.000153644917294164, | |
| "loss": 0.7511, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.882934872217642, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.0001534016356850244, | |
| "loss": 0.7492, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 3.8911788953009068, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.00015315791111666425, | |
| "loss": 0.7529, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 3.8994229183841713, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.00015291374561074536, | |
| "loss": 0.7481, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 3.9076669414674363, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 0.000152669141192587, | |
| "loss": 0.752, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 3.915910964550701, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 0.00015242409989114916, | |
| "loss": 0.7389, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 3.9241549876339654, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 0.00015217862373901575, | |
| "loss": 0.7521, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 3.93239901071723, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 0.0001519327147723776, | |
| "loss": 0.742, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 3.9406430338004945, | |
| "grad_norm": 0.73046875, | |
| "learning_rate": 0.00015168637503101584, | |
| "loss": 0.7499, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 3.948887056883759, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00015143960655828468, | |
| "loss": 0.7516, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 3.957131079967024, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.00015119241140109467, | |
| "loss": 0.7493, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.9653751030502886, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 0.0001509447916098956, | |
| "loss": 0.7445, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 3.973619126133553, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.0001506967492386596, | |
| "loss": 0.7535, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 3.9818631492168177, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 0.000150448286344864, | |
| "loss": 0.7411, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 3.9901071723000827, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 0.00015019940498947428, | |
| "loss": 0.7484, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 3.9983511953833473, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 0.00014995010723692714, | |
| "loss": 0.7465, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.436275005340576, | |
| "eval_runtime": 0.2365, | |
| "eval_samples_per_second": 42.283, | |
| "eval_steps_per_second": 4.228, | |
| "step": 2426 | |
| }, | |
| { | |
| "epoch": 4.006595218466612, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 0.00014970039515511304, | |
| "loss": 0.7483, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 4.014839241549876, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 0.00014945027081535937, | |
| "loss": 0.7256, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 4.023083264633141, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 0.00014919973629241314, | |
| "loss": 0.7386, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 4.0313272877164055, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 0.0001489487936644237, | |
| "loss": 0.7329, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 4.03957131079967, | |
| "grad_norm": 0.84765625, | |
| "learning_rate": 0.00014869744501292561, | |
| "loss": 0.7317, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 4.047815333882935, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.00014844569242282148, | |
| "loss": 0.7278, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 4.056059356966199, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00014819353798236427, | |
| "loss": 0.73, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 4.064303380049465, | |
| "grad_norm": 0.91796875, | |
| "learning_rate": 0.0001479409837831404, | |
| "loss": 0.7357, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 4.072547403132729, | |
| "grad_norm": 0.78125, | |
| "learning_rate": 0.00014768803192005223, | |
| "loss": 0.7341, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 4.080791426215994, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00014743468449130063, | |
| "loss": 0.7367, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 4.089035449299258, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00014718094359836772, | |
| "loss": 0.7322, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 4.097279472382523, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 0.00014692681134599925, | |
| "loss": 0.73, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 4.105523495465787, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.0001466722898421873, | |
| "loss": 0.7364, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 4.113767518549052, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.00014641738119815266, | |
| "loss": 0.7267, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 4.122011541632316, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.00014616208752832758, | |
| "loss": 0.7282, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.130255564715581, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 0.00014590641095033787, | |
| "loss": 0.7251, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 4.1384995877988455, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.0001456503535849855, | |
| "loss": 0.7391, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 4.14674361088211, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.0001453939175562312, | |
| "loss": 0.7346, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 4.1549876339653755, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 0.00014513710499117647, | |
| "loss": 0.7362, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 4.16323165704864, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 0.00014487991802004623, | |
| "loss": 0.731, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 4.171475680131905, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.00014462235877617098, | |
| "loss": 0.7285, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 4.179719703215169, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 0.0001443644293959693, | |
| "loss": 0.7386, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 4.187963726298434, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 0.00014410613201892985, | |
| "loss": 0.7376, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 4.196207749381698, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 0.0001438474687875938, | |
| "loss": 0.731, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 4.204451772464963, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.00014358844184753712, | |
| "loss": 0.7238, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 4.212695795548227, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 0.00014332905334735261, | |
| "loss": 0.7246, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 4.220939818631492, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 0.00014306930543863219, | |
| "loss": 0.7394, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 4.2291838417147565, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 0.00014280920027594907, | |
| "loss": 0.7306, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 4.237427864798021, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.00014254874001683976, | |
| "loss": 0.7418, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 4.2456718878812865, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 0.00014228792682178623, | |
| "loss": 0.7291, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 4.253915910964551, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.00014202676285419812, | |
| "loss": 0.7273, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 4.262159934047816, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00014176525028039452, | |
| "loss": 0.7311, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 4.27040395713108, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.00014150339126958633, | |
| "loss": 0.7214, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 4.278647980214345, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.00014124118799385796, | |
| "loss": 0.7324, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 4.286892003297609, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 0.00014097864262814955, | |
| "loss": 0.7397, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.295136026380874, | |
| "grad_norm": 0.625, | |
| "learning_rate": 0.00014071575735023875, | |
| "loss": 0.7382, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 4.303380049464138, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 0.0001404525343407228, | |
| "loss": 0.7324, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 4.311624072547403, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 0.00014018897578300035, | |
| "loss": 0.7327, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 4.319868095630667, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.0001399250838632533, | |
| "loss": 0.7419, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 4.328112118713932, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.0001396608607704289, | |
| "loss": 0.738, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 4.336356141797197, | |
| "grad_norm": 0.70703125, | |
| "learning_rate": 0.00013939630869622133, | |
| "loss": 0.7412, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 4.344600164880462, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 0.00013913142983505364, | |
| "loss": 0.7336, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 4.3528441879637265, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00013886622638405952, | |
| "loss": 0.7282, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 4.361088211046991, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.00013860070054306516, | |
| "loss": 0.7306, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 4.369332234130256, | |
| "grad_norm": 0.84765625, | |
| "learning_rate": 0.0001383348545145708, | |
| "loss": 0.7279, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.37757625721352, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 0.0001380686905037327, | |
| "loss": 0.7355, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 4.385820280296785, | |
| "grad_norm": 1.0546875, | |
| "learning_rate": 0.00013780221071834476, | |
| "loss": 0.7336, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 4.394064303380049, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.0001375354173688201, | |
| "loss": 0.7314, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 4.402308326463314, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.00013726831266817278, | |
| "loss": 0.7344, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 4.410552349546578, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 0.00013700089883199966, | |
| "loss": 0.7361, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 4.418796372629844, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.0001367331780784616, | |
| "loss": 0.7322, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 4.427040395713108, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00013646515262826552, | |
| "loss": 0.7332, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 4.435284418796373, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 0.00013619682470464558, | |
| "loss": 0.7321, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 4.4435284418796375, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 0.00013592819653334505, | |
| "loss": 0.7262, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 4.451772464962902, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.0001356592703425976, | |
| "loss": 0.7273, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.460016488046167, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 0.00013539004836310894, | |
| "loss": 0.7378, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 4.468260511129431, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.0001351205328280385, | |
| "loss": 0.7254, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 4.476504534212696, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00013485072597298038, | |
| "loss": 0.729, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 4.48474855729596, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.00013458063003594543, | |
| "loss": 0.7375, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 4.492992580379225, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.0001343102472573423, | |
| "loss": 0.7278, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 4.501236603462489, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 0.00013403957987995882, | |
| "loss": 0.7363, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 4.509480626545754, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00013376863014894375, | |
| "loss": 0.7341, | |
| "step": 2735 | |
| }, | |
| { | |
| "epoch": 4.517724649629019, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.00013349740031178784, | |
| "loss": 0.7325, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 4.525968672712284, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 0.00013322589261830517, | |
| "loss": 0.7376, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 4.534212695795548, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.00013295410932061478, | |
| "loss": 0.727, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.542456718878813, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 0.00013268205267312174, | |
| "loss": 0.729, | |
| "step": 2755 | |
| }, | |
| { | |
| "epoch": 4.5507007419620775, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.00013240972493249847, | |
| "loss": 0.7355, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 4.558944765045342, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.00013213712835766607, | |
| "loss": 0.7362, | |
| "step": 2765 | |
| }, | |
| { | |
| "epoch": 4.567188788128607, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 0.0001318642652097757, | |
| "loss": 0.7319, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 4.575432811211871, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.00013159113775218964, | |
| "loss": 0.7265, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 4.583676834295136, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00013131774825046245, | |
| "loss": 0.7343, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 4.5919208573784, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 0.00013104409897232258, | |
| "loss": 0.7231, | |
| "step": 2785 | |
| }, | |
| { | |
| "epoch": 4.600164880461666, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 0.00013077019218765305, | |
| "loss": 0.7305, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 4.60840890354493, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.00013049603016847296, | |
| "loss": 0.7311, | |
| "step": 2795 | |
| }, | |
| { | |
| "epoch": 4.616652926628195, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.00013022161518891855, | |
| "loss": 0.7347, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.624896949711459, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.00012994694952522435, | |
| "loss": 0.7395, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 4.633140972794724, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.00012967203545570418, | |
| "loss": 0.7332, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 4.6413849958779885, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 0.0001293968752607325, | |
| "loss": 0.7326, | |
| "step": 2815 | |
| }, | |
| { | |
| "epoch": 4.649629018961253, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00012912147122272523, | |
| "loss": 0.7317, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 4.657873042044518, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 0.00012884582562612095, | |
| "loss": 0.7336, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 4.666117065127782, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.00012856994075736197, | |
| "loss": 0.7283, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 4.674361088211047, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 0.00012829381890487536, | |
| "loss": 0.7366, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 4.682605111294311, | |
| "grad_norm": 0.8515625, | |
| "learning_rate": 0.00012801746235905384, | |
| "loss": 0.7377, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 4.690849134377576, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.00012774087341223695, | |
| "loss": 0.7357, | |
| "step": 2845 | |
| }, | |
| { | |
| "epoch": 4.699093157460841, | |
| "grad_norm": 0.490234375, | |
| "learning_rate": 0.00012746405435869198, | |
| "loss": 0.7307, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.707337180544106, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.00012718700749459486, | |
| "loss": 0.7307, | |
| "step": 2855 | |
| }, | |
| { | |
| "epoch": 4.71558120362737, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 0.0001269097351180112, | |
| "loss": 0.7244, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 4.723825226710635, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.00012663223952887723, | |
| "loss": 0.7321, | |
| "step": 2865 | |
| }, | |
| { | |
| "epoch": 4.732069249793899, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.0001263545230289807, | |
| "loss": 0.7243, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 4.740313272877164, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.00012607658792194174, | |
| "loss": 0.7282, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 4.7485572959604285, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.0001257984365131938, | |
| "loss": 0.7239, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 4.756801319043693, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 0.00012552007110996463, | |
| "loss": 0.7273, | |
| "step": 2885 | |
| }, | |
| { | |
| "epoch": 4.765045342126958, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 0.00012524149402125685, | |
| "loss": 0.7251, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 4.773289365210223, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00012496270755782914, | |
| "loss": 0.739, | |
| "step": 2895 | |
| }, | |
| { | |
| "epoch": 4.781533388293488, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 0.00012468371403217684, | |
| "loss": 0.7344, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.789777411376752, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.00012440451575851285, | |
| "loss": 0.7314, | |
| "step": 2905 | |
| }, | |
| { | |
| "epoch": 4.798021434460017, | |
| "grad_norm": 0.5, | |
| "learning_rate": 0.00012412511505274844, | |
| "loss": 0.7269, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 4.806265457543281, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 0.00012384551423247407, | |
| "loss": 0.7292, | |
| "step": 2915 | |
| }, | |
| { | |
| "epoch": 4.814509480626546, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 0.00012356571561693996, | |
| "loss": 0.7227, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 4.82275350370981, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.00012328572152703725, | |
| "loss": 0.7311, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 4.830997526793075, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 0.00012300553428527832, | |
| "loss": 0.7315, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 4.8392415498763395, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 0.00012272515621577782, | |
| "loss": 0.7376, | |
| "step": 2935 | |
| }, | |
| { | |
| "epoch": 4.847485572959604, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 0.00012244458964423327, | |
| "loss": 0.7305, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 4.855729596042869, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00012216383689790574, | |
| "loss": 0.7279, | |
| "step": 2945 | |
| }, | |
| { | |
| "epoch": 4.863973619126133, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.00012188290030560063, | |
| "loss": 0.7299, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.872217642209399, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 0.00012160178219764837, | |
| "loss": 0.7253, | |
| "step": 2955 | |
| }, | |
| { | |
| "epoch": 4.880461665292663, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 0.00012132048490588492, | |
| "loss": 0.7291, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 4.888705688375928, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 0.00012103901076363269, | |
| "loss": 0.7244, | |
| "step": 2965 | |
| }, | |
| { | |
| "epoch": 4.896949711459192, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.0001207573621056809, | |
| "loss": 0.7279, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 4.905193734542457, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00012047554126826643, | |
| "loss": 0.7297, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 4.913437757625721, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00012019355058905435, | |
| "loss": 0.7285, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 4.921681780708986, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 0.00011991139240711857, | |
| "loss": 0.7312, | |
| "step": 2985 | |
| }, | |
| { | |
| "epoch": 4.92992580379225, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 0.00011962906906292238, | |
| "loss": 0.7284, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 4.938169826875515, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.00011934658289829902, | |
| "loss": 0.7336, | |
| "step": 2995 | |
| }, | |
| { | |
| "epoch": 4.9464138499587795, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00011906393625643244, | |
| "loss": 0.7281, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.954657873042045, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 0.00011878113148183758, | |
| "loss": 0.7271, | |
| "step": 3005 | |
| }, | |
| { | |
| "epoch": 4.9629018961253095, | |
| "grad_norm": 0.5, | |
| "learning_rate": 0.00011849817092034118, | |
| "loss": 0.7229, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 4.971145919208574, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.00011821505691906216, | |
| "loss": 0.7318, | |
| "step": 3015 | |
| }, | |
| { | |
| "epoch": 4.979389942291839, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00011793179182639218, | |
| "loss": 0.7366, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 4.987633965375103, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.00011764837799197622, | |
| "loss": 0.7337, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 4.995877988458368, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 0.7312, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 4.999175597691673, | |
| "eval_loss": 2.439051389694214, | |
| "eval_runtime": 0.2596, | |
| "eval_samples_per_second": 38.523, | |
| "eval_steps_per_second": 3.852, | |
| "step": 3032 | |
| }, | |
| { | |
| "epoch": 5.004122011541632, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 0.0001170811135026357, | |
| "loss": 0.7263, | |
| "step": 3035 | |
| }, | |
| { | |
| "epoch": 5.012366034624897, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 0.00011679726755309205, | |
| "loss": 0.7183, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 5.020610057708161, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00011651328227252517, | |
| "loss": 0.723, | |
| "step": 3045 | |
| }, | |
| { | |
| "epoch": 5.028854080791426, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00011622916001655388, | |
| "loss": 0.7185, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 5.0370981038746905, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00011594490314193323, | |
| "loss": 0.7132, | |
| "step": 3055 | |
| }, | |
| { | |
| "epoch": 5.045342126957956, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00011566051400653486, | |
| "loss": 0.7054, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 5.0535861500412205, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.00011537599496932752, | |
| "loss": 0.7197, | |
| "step": 3065 | |
| }, | |
| { | |
| "epoch": 5.061830173124485, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.00011509134839035748, | |
| "loss": 0.7157, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 5.07007419620775, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 0.00011480657663072896, | |
| "loss": 0.7093, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 5.078318219291014, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 0.0001145216820525845, | |
| "loss": 0.7286, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 5.086562242374279, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.00011423666701908547, | |
| "loss": 0.7105, | |
| "step": 3085 | |
| }, | |
| { | |
| "epoch": 5.094806265457543, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.00011395153389439233, | |
| "loss": 0.7072, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 5.103050288540808, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.00011366628504364509, | |
| "loss": 0.7156, | |
| "step": 3095 | |
| }, | |
| { | |
| "epoch": 5.111294311624072, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 0.00011338092283294377, | |
| "loss": 0.7052, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 5.119538334707337, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00011309544962932862, | |
| "loss": 0.7197, | |
| "step": 3105 | |
| }, | |
| { | |
| "epoch": 5.127782357790601, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.00011280986780076057, | |
| "loss": 0.7195, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 5.136026380873867, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.00011252417971610163, | |
| "loss": 0.7062, | |
| "step": 3115 | |
| }, | |
| { | |
| "epoch": 5.144270403957131, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00011223838774509514, | |
| "loss": 0.7225, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 5.152514427040396, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.00011195249425834615, | |
| "loss": 0.7106, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 5.1607584501236605, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 0.00011166650162730188, | |
| "loss": 0.7174, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 5.169002473206925, | |
| "grad_norm": 0.76953125, | |
| "learning_rate": 0.00011138041222423177, | |
| "loss": 0.7208, | |
| "step": 3135 | |
| }, | |
| { | |
| "epoch": 5.17724649629019, | |
| "grad_norm": 0.7265625, | |
| "learning_rate": 0.00011109422842220805, | |
| "loss": 0.716, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 5.185490519373454, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 0.00011080795259508608, | |
| "loss": 0.717, | |
| "step": 3145 | |
| }, | |
| { | |
| "epoch": 5.193734542456719, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.00011052158711748434, | |
| "loss": 0.7093, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 5.201978565539983, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 0.00011023513436476511, | |
| "loss": 0.7129, | |
| "step": 3155 | |
| }, | |
| { | |
| "epoch": 5.210222588623248, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.00010994859671301462, | |
| "loss": 0.7168, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 5.218466611706512, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 0.0001096619765390232, | |
| "loss": 0.7158, | |
| "step": 3165 | |
| }, | |
| { | |
| "epoch": 5.226710634789778, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 0.00010937527622026575, | |
| "loss": 0.7229, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 5.234954657873042, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 0.00010908849813488203, | |
| "loss": 0.7151, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 5.243198680956307, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.00010880164466165674, | |
| "loss": 0.7185, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 5.2514427040395715, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 0.00010851471817999997, | |
| "loss": 0.7113, | |
| "step": 3185 | |
| }, | |
| { | |
| "epoch": 5.259686727122836, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.00010822772106992747, | |
| "loss": 0.7178, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 5.267930750206101, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00010794065571204072, | |
| "loss": 0.7106, | |
| "step": 3195 | |
| }, | |
| { | |
| "epoch": 5.276174773289365, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.0001076535244875074, | |
| "loss": 0.7136, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.28441879637263, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00010736632977804149, | |
| "loss": 0.7138, | |
| "step": 3205 | |
| }, | |
| { | |
| "epoch": 5.292662819455894, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 0.00010707907396588361, | |
| "loss": 0.7192, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 5.300906842539159, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 0.00010679175943378119, | |
| "loss": 0.7068, | |
| "step": 3215 | |
| }, | |
| { | |
| "epoch": 5.309150865622423, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 0.00010650438856496872, | |
| "loss": 0.7095, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 5.317394888705689, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 0.00010621696374314807, | |
| "loss": 0.7118, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 5.325638911788953, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.00010592948735246854, | |
| "loss": 0.711, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 5.333882934872218, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00010564196177750725, | |
| "loss": 0.7172, | |
| "step": 3235 | |
| }, | |
| { | |
| "epoch": 5.342126957955482, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 0.0001053543894032493, | |
| "loss": 0.7171, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 5.350370981038747, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 0.00010506677261506797, | |
| "loss": 0.7153, | |
| "step": 3245 | |
| }, | |
| { | |
| "epoch": 5.3586150041220115, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 0.00010477911379870488, | |
| "loss": 0.7162, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 5.366859027205276, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.00010449141534025045, | |
| "loss": 0.7067, | |
| "step": 3255 | |
| }, | |
| { | |
| "epoch": 5.375103050288541, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 0.00010420367962612372, | |
| "loss": 0.7117, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 5.383347073371805, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00010391590904305284, | |
| "loss": 0.7175, | |
| "step": 3265 | |
| }, | |
| { | |
| "epoch": 5.39159109645507, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.00010362810597805526, | |
| "loss": 0.7109, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 5.399835119538334, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 0.00010334027281841781, | |
| "loss": 0.7136, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 5.4080791426216, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.00010305241195167687, | |
| "loss": 0.7123, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 5.416323165704864, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 0.00010276452576559879, | |
| "loss": 0.7132, | |
| "step": 3285 | |
| }, | |
| { | |
| "epoch": 5.424567188788129, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 0.00010247661664815986, | |
| "loss": 0.7161, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 5.432811211871393, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00010218868698752658, | |
| "loss": 0.7122, | |
| "step": 3295 | |
| }, | |
| { | |
| "epoch": 5.441055234954658, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.00010190073917203589, | |
| "loss": 0.7167, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 5.4492992580379225, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00010161277559017528, | |
| "loss": 0.7143, | |
| "step": 3305 | |
| }, | |
| { | |
| "epoch": 5.457543281121187, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 0.00010132479863056303, | |
| "loss": 0.7163, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 5.465787304204452, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 0.00010103681068192845, | |
| "loss": 0.7173, | |
| "step": 3315 | |
| }, | |
| { | |
| "epoch": 5.474031327287716, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.00010074881413309193, | |
| "loss": 0.714, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 5.482275350370981, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00010046081137294516, | |
| "loss": 0.7128, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 5.490519373454246, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00010017280479043147, | |
| "loss": 0.7242, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 5.498763396537511, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 9.988479677452584e-05, | |
| "loss": 0.7196, | |
| "step": 3335 | |
| }, | |
| { | |
| "epoch": 5.507007419620775, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 9.959678971421508e-05, | |
| "loss": 0.714, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 5.51525144270404, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 9.930878599847821e-05, | |
| "loss": 0.7173, | |
| "step": 3345 | |
| }, | |
| { | |
| "epoch": 5.523495465787304, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 9.902078801626636e-05, | |
| "loss": 0.7137, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 5.531739488870569, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 9.873279815648318e-05, | |
| "loss": 0.7125, | |
| "step": 3355 | |
| }, | |
| { | |
| "epoch": 5.539983511953833, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 9.844481880796491e-05, | |
| "loss": 0.7173, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 5.548227535037098, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 9.815685235946068e-05, | |
| "loss": 0.7134, | |
| "step": 3365 | |
| }, | |
| { | |
| "epoch": 5.5564715581203625, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 9.786890119961253e-05, | |
| "loss": 0.7199, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 5.564715581203627, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 9.758096771693573e-05, | |
| "loss": 0.7116, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 5.572959604286892, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 9.729305429979887e-05, | |
| "loss": 0.7131, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 5.581203627370156, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 9.700516333640415e-05, | |
| "loss": 0.7172, | |
| "step": 3385 | |
| }, | |
| { | |
| "epoch": 5.589447650453422, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 9.671729721476746e-05, | |
| "loss": 0.7121, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 5.597691673536686, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 9.642945832269874e-05, | |
| "loss": 0.7187, | |
| "step": 3395 | |
| }, | |
| { | |
| "epoch": 5.605935696619951, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 9.614164904778196e-05, | |
| "loss": 0.7108, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 5.614179719703215, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 9.585387177735547e-05, | |
| "loss": 0.7099, | |
| "step": 3405 | |
| }, | |
| { | |
| "epoch": 5.62242374278648, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 9.556612889849214e-05, | |
| "loss": 0.7169, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 5.630667765869744, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 9.527842279797953e-05, | |
| "loss": 0.7118, | |
| "step": 3415 | |
| }, | |
| { | |
| "epoch": 5.638911788953009, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 9.499075586230013e-05, | |
| "loss": 0.7148, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 5.6471558120362735, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 9.470313047761167e-05, | |
| "loss": 0.7166, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 5.655399835119538, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 9.44155490297271e-05, | |
| "loss": 0.7156, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 5.663643858202803, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 9.412801390409497e-05, | |
| "loss": 0.707, | |
| "step": 3435 | |
| }, | |
| { | |
| "epoch": 5.671887881286068, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 9.38405274857796e-05, | |
| "loss": 0.7125, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 5.680131904369333, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 9.355309215944124e-05, | |
| "loss": 0.7153, | |
| "step": 3445 | |
| }, | |
| { | |
| "epoch": 5.688375927452597, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 9.326571030931637e-05, | |
| "loss": 0.7143, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 5.696619950535862, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 9.297838431919794e-05, | |
| "loss": 0.7192, | |
| "step": 3455 | |
| }, | |
| { | |
| "epoch": 5.704863973619126, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 9.269111657241548e-05, | |
| "loss": 0.7151, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 5.713107996702391, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 9.240390945181543e-05, | |
| "loss": 0.7171, | |
| "step": 3465 | |
| }, | |
| { | |
| "epoch": 5.721352019785655, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 9.211676533974131e-05, | |
| "loss": 0.7111, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 5.72959604286892, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 9.182968661801412e-05, | |
| "loss": 0.7111, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 5.737840065952184, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 9.154267566791223e-05, | |
| "loss": 0.7211, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 5.746084089035449, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 9.125573487015203e-05, | |
| "loss": 0.7165, | |
| "step": 3485 | |
| }, | |
| { | |
| "epoch": 5.7543281121187135, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 9.096886660486797e-05, | |
| "loss": 0.7082, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 5.762572135201978, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 9.068207325159284e-05, | |
| "loss": 0.7136, | |
| "step": 3495 | |
| }, | |
| { | |
| "epoch": 5.7708161582852435, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 9.039535718923804e-05, | |
| "loss": 0.714, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.779060181368508, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 9.01087207960739e-05, | |
| "loss": 0.7174, | |
| "step": 3505 | |
| }, | |
| { | |
| "epoch": 5.787304204451773, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 8.982216644970979e-05, | |
| "loss": 0.7071, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 5.795548227535037, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 8.953569652707459e-05, | |
| "loss": 0.7081, | |
| "step": 3515 | |
| }, | |
| { | |
| "epoch": 5.803792250618302, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 8.924931340439694e-05, | |
| "loss": 0.7124, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 5.812036273701566, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 8.896301945718541e-05, | |
| "loss": 0.7115, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 5.820280296784831, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 8.867681706020894e-05, | |
| "loss": 0.7134, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 5.828524319868095, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 8.839070858747697e-05, | |
| "loss": 0.7169, | |
| "step": 3535 | |
| }, | |
| { | |
| "epoch": 5.83676834295136, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 8.810469641222001e-05, | |
| "loss": 0.7154, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 5.845012366034625, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 8.781878290686959e-05, | |
| "loss": 0.7182, | |
| "step": 3545 | |
| }, | |
| { | |
| "epoch": 5.85325638911789, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 8.753297044303896e-05, | |
| "loss": 0.7128, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 5.8615004122011545, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 8.724726139150318e-05, | |
| "loss": 0.7083, | |
| "step": 3555 | |
| }, | |
| { | |
| "epoch": 5.869744435284419, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 8.696165812217953e-05, | |
| "loss": 0.7175, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 5.877988458367684, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 8.667616300410778e-05, | |
| "loss": 0.7174, | |
| "step": 3565 | |
| }, | |
| { | |
| "epoch": 5.886232481450948, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 8.639077840543077e-05, | |
| "loss": 0.7173, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 5.894476504534213, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 8.610550669337433e-05, | |
| "loss": 0.7147, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 5.902720527617477, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 8.582035023422815e-05, | |
| "loss": 0.7169, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 5.910964550700742, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 8.553531139332582e-05, | |
| "loss": 0.7237, | |
| "step": 3585 | |
| }, | |
| { | |
| "epoch": 5.919208573784006, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 8.525039253502529e-05, | |
| "loss": 0.7134, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 5.927452596867271, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 8.496559602268928e-05, | |
| "loss": 0.7189, | |
| "step": 3595 | |
| }, | |
| { | |
| "epoch": 5.935696619950535, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 8.468092421866573e-05, | |
| "loss": 0.717, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 5.943940643033801, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 8.439637948426801e-05, | |
| "loss": 0.7094, | |
| "step": 3605 | |
| }, | |
| { | |
| "epoch": 5.952184666117065, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 8.411196417975558e-05, | |
| "loss": 0.7019, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 5.96042868920033, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 8.382768066431425e-05, | |
| "loss": 0.7127, | |
| "step": 3615 | |
| }, | |
| { | |
| "epoch": 5.9686727122835945, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 8.354353129603668e-05, | |
| "loss": 0.7133, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 5.976916735366859, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 8.325951843190274e-05, | |
| "loss": 0.7182, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 5.985160758450124, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 8.297564442776014e-05, | |
| "loss": 0.7053, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 5.993404781533388, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 8.269191163830467e-05, | |
| "loss": 0.7253, | |
| "step": 3635 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 2.459299325942993, | |
| "eval_runtime": 0.2463, | |
| "eval_samples_per_second": 40.595, | |
| "eval_steps_per_second": 4.059, | |
| "step": 3639 | |
| }, | |
| { | |
| "epoch": 6.001648804616653, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 8.240832241706068e-05, | |
| "loss": 0.7144, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 6.009892827699917, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 8.212487911636184e-05, | |
| "loss": 0.7102, | |
| "step": 3645 | |
| }, | |
| { | |
| "epoch": 6.018136850783182, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 8.184158408733131e-05, | |
| "loss": 0.7073, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 6.026380873866446, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 8.155843967986236e-05, | |
| "loss": 0.6914, | |
| "step": 3655 | |
| }, | |
| { | |
| "epoch": 6.034624896949712, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 8.127544824259889e-05, | |
| "loss": 0.7095, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 6.042868920032976, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 8.099261212291601e-05, | |
| "loss": 0.7006, | |
| "step": 3665 | |
| }, | |
| { | |
| "epoch": 6.051112943116241, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 8.070993366690029e-05, | |
| "loss": 0.6983, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 6.0593569661995055, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 8.042741521933071e-05, | |
| "loss": 0.7086, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 6.06760098928277, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 8.014505912365893e-05, | |
| "loss": 0.7039, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 6.075845012366035, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 7.986286772198986e-05, | |
| "loss": 0.7056, | |
| "step": 3685 | |
| }, | |
| { | |
| "epoch": 6.084089035449299, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 7.958084335506239e-05, | |
| "loss": 0.6957, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 6.092333058532564, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 7.929898836222983e-05, | |
| "loss": 0.7052, | |
| "step": 3695 | |
| }, | |
| { | |
| "epoch": 6.100577081615828, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 7.90173050814406e-05, | |
| "loss": 0.6982, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 6.108821104699093, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 7.873579584921869e-05, | |
| "loss": 0.7029, | |
| "step": 3705 | |
| }, | |
| { | |
| "epoch": 6.117065127782358, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 7.84544630006445e-05, | |
| "loss": 0.7015, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 6.125309150865623, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 7.817330886933527e-05, | |
| "loss": 0.7073, | |
| "step": 3715 | |
| }, | |
| { | |
| "epoch": 6.133553173948887, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 7.789233578742582e-05, | |
| "loss": 0.7092, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 6.141797197032152, | |
| "grad_norm": 0.490234375, | |
| "learning_rate": 7.761154608554927e-05, | |
| "loss": 0.7025, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 6.150041220115416, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 7.733094209281756e-05, | |
| "loss": 0.7048, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 6.158285243198681, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 7.705052613680211e-05, | |
| "loss": 0.7029, | |
| "step": 3735 | |
| }, | |
| { | |
| "epoch": 6.1665292662819455, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 7.677030054351477e-05, | |
| "loss": 0.701, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 6.17477328936521, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 7.649026763738827e-05, | |
| "loss": 0.7067, | |
| "step": 3745 | |
| }, | |
| { | |
| "epoch": 6.183017312448475, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 7.6210429741257e-05, | |
| "loss": 0.7055, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 6.191261335531739, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 7.593078917633787e-05, | |
| "loss": 0.7104, | |
| "step": 3755 | |
| }, | |
| { | |
| "epoch": 6.199505358615004, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 7.565134826221083e-05, | |
| "loss": 0.703, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 6.207749381698269, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 7.537210931679987e-05, | |
| "loss": 0.6998, | |
| "step": 3765 | |
| }, | |
| { | |
| "epoch": 6.215993404781534, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 7.509307465635358e-05, | |
| "loss": 0.6976, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 6.224237427864798, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 7.481424659542609e-05, | |
| "loss": 0.7025, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 6.232481450948063, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 7.453562744685778e-05, | |
| "loss": 0.6971, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 6.240725474031327, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 7.425721952175618e-05, | |
| "loss": 0.6984, | |
| "step": 3785 | |
| }, | |
| { | |
| "epoch": 6.248969497114592, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 7.39790251294767e-05, | |
| "loss": 0.7012, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 6.2572135201978565, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 7.370104657760361e-05, | |
| "loss": 0.7012, | |
| "step": 3795 | |
| }, | |
| { | |
| "epoch": 6.265457543281121, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 7.342328617193067e-05, | |
| "loss": 0.7069, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 6.273701566364386, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 7.314574621644225e-05, | |
| "loss": 0.6998, | |
| "step": 3805 | |
| }, | |
| { | |
| "epoch": 6.28194558944765, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 7.286842901329412e-05, | |
| "loss": 0.695, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 6.290189612530915, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 7.259133686279429e-05, | |
| "loss": 0.7045, | |
| "step": 3815 | |
| }, | |
| { | |
| "epoch": 6.29843363561418, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 7.231447206338407e-05, | |
| "loss": 0.7062, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 6.306677658697445, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 7.203783691161883e-05, | |
| "loss": 0.6975, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 6.314921681780709, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 7.176143370214914e-05, | |
| "loss": 0.7035, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 6.323165704863974, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 7.148526472770154e-05, | |
| "loss": 0.7071, | |
| "step": 3835 | |
| }, | |
| { | |
| "epoch": 6.331409727947238, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 7.12093322790597e-05, | |
| "loss": 0.7022, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 6.339653751030503, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 7.09336386450453e-05, | |
| "loss": 0.7104, | |
| "step": 3845 | |
| }, | |
| { | |
| "epoch": 6.347897774113767, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 7.065818611249915e-05, | |
| "loss": 0.7028, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 6.356141797197032, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 7.038297696626206e-05, | |
| "loss": 0.7049, | |
| "step": 3855 | |
| }, | |
| { | |
| "epoch": 6.3643858202802965, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 7.010801348915608e-05, | |
| "loss": 0.7074, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 6.372629843363561, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 6.983329796196534e-05, | |
| "loss": 0.7001, | |
| "step": 3865 | |
| }, | |
| { | |
| "epoch": 6.380873866446826, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 6.955883266341741e-05, | |
| "loss": 0.7006, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 6.389117889530091, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 6.928461987016413e-05, | |
| "loss": 0.7113, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 6.397361912613356, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 6.901066185676295e-05, | |
| "loss": 0.6964, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 6.40560593569662, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 6.873696089565786e-05, | |
| "loss": 0.7086, | |
| "step": 3885 | |
| }, | |
| { | |
| "epoch": 6.413849958779885, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 6.846351925716068e-05, | |
| "loss": 0.698, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 6.422093981863149, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 6.819033920943219e-05, | |
| "loss": 0.6997, | |
| "step": 3895 | |
| }, | |
| { | |
| "epoch": 6.430338004946414, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 6.791742301846326e-05, | |
| "loss": 0.7031, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 6.438582028029678, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 6.764477294805615e-05, | |
| "loss": 0.7026, | |
| "step": 3905 | |
| }, | |
| { | |
| "epoch": 6.446826051112943, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 6.737239125980573e-05, | |
| "loss": 0.7006, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 6.4550700741962075, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 6.710028021308061e-05, | |
| "loss": 0.6971, | |
| "step": 3915 | |
| }, | |
| { | |
| "epoch": 6.463314097279472, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 6.682844206500445e-05, | |
| "loss": 0.7028, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 6.471558120362737, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 6.655687907043734e-05, | |
| "loss": 0.7053, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 6.479802143446002, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 6.62855934819569e-05, | |
| "loss": 0.6995, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 6.488046166529267, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 6.601458754983978e-05, | |
| "loss": 0.6971, | |
| "step": 3935 | |
| }, | |
| { | |
| "epoch": 6.496290189612531, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 6.574386352204289e-05, | |
| "loss": 0.7029, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 6.504534212695796, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 6.547342364418481e-05, | |
| "loss": 0.7011, | |
| "step": 3945 | |
| }, | |
| { | |
| "epoch": 6.51277823577906, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 6.520327015952713e-05, | |
| "loss": 0.699, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 6.521022258862325, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 6.493340530895583e-05, | |
| "loss": 0.6987, | |
| "step": 3955 | |
| }, | |
| { | |
| "epoch": 6.529266281945589, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 6.466383133096267e-05, | |
| "loss": 0.7095, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 6.537510305028854, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 6.439455046162677e-05, | |
| "loss": 0.704, | |
| "step": 3965 | |
| }, | |
| { | |
| "epoch": 6.545754328112118, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 6.412556493459581e-05, | |
| "loss": 0.7127, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 6.553998351195383, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 6.385687698106781e-05, | |
| "loss": 0.7019, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 6.562242374278648, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 6.358848882977233e-05, | |
| "loss": 0.702, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 6.570486397361913, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 6.332040270695219e-05, | |
| "loss": 0.7086, | |
| "step": 3985 | |
| }, | |
| { | |
| "epoch": 6.5787304204451775, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 6.305262083634488e-05, | |
| "loss": 0.7086, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 6.586974443528442, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 6.278514543916415e-05, | |
| "loss": 0.7087, | |
| "step": 3995 | |
| }, | |
| { | |
| "epoch": 6.595218466611707, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 6.251797873408161e-05, | |
| "loss": 0.6976, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.603462489694971, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 6.225112293720836e-05, | |
| "loss": 0.6968, | |
| "step": 4005 | |
| }, | |
| { | |
| "epoch": 6.611706512778236, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 6.198458026207652e-05, | |
| "loss": 0.7039, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 6.6199505358615, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 6.171835291962088e-05, | |
| "loss": 0.702, | |
| "step": 4015 | |
| }, | |
| { | |
| "epoch": 6.628194558944765, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 6.145244311816063e-05, | |
| "loss": 0.7004, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 6.636438582028029, | |
| "grad_norm": 0.5, | |
| "learning_rate": 6.1186853063381e-05, | |
| "loss": 0.6988, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 6.644682605111294, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 6.092158495831486e-05, | |
| "loss": 0.7019, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 6.6529266281945585, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 6.065664100332478e-05, | |
| "loss": 0.7082, | |
| "step": 4035 | |
| }, | |
| { | |
| "epoch": 6.661170651277824, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 6.039202339608432e-05, | |
| "loss": 0.7008, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 6.6694146743610885, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 6.012773433156017e-05, | |
| "loss": 0.7022, | |
| "step": 4045 | |
| }, | |
| { | |
| "epoch": 6.677658697444353, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 5.986377600199371e-05, | |
| "loss": 0.6986, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 6.685902720527618, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 5.9600150596883066e-05, | |
| "loss": 0.6989, | |
| "step": 4055 | |
| }, | |
| { | |
| "epoch": 6.694146743610882, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 5.933686030296459e-05, | |
| "loss": 0.6993, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 6.702390766694147, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 5.907390730419507e-05, | |
| "loss": 0.6977, | |
| "step": 4065 | |
| }, | |
| { | |
| "epoch": 6.710634789777411, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 5.881129378173347e-05, | |
| "loss": 0.7019, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 6.718878812860676, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 5.854902191392284e-05, | |
| "loss": 0.6936, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 6.72712283594394, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 5.828709387627218e-05, | |
| "loss": 0.7002, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 6.735366859027205, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 5.802551184143865e-05, | |
| "loss": 0.7026, | |
| "step": 4085 | |
| }, | |
| { | |
| "epoch": 6.74361088211047, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 5.7764277979209094e-05, | |
| "loss": 0.7151, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 6.751854905193735, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 5.750339445648252e-05, | |
| "loss": 0.7055, | |
| "step": 4095 | |
| }, | |
| { | |
| "epoch": 6.760098928276999, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 5.724286343725185e-05, | |
| "loss": 0.7032, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 6.768342951360264, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 5.6982687082585994e-05, | |
| "loss": 0.7008, | |
| "step": 4105 | |
| }, | |
| { | |
| "epoch": 6.7765869744435285, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 5.6722867550612116e-05, | |
| "loss": 0.6998, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 6.784830997526793, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 5.6463406996497456e-05, | |
| "loss": 0.6961, | |
| "step": 4115 | |
| }, | |
| { | |
| "epoch": 6.793075020610058, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 5.620430757243156e-05, | |
| "loss": 0.6963, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 6.801319043693322, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 5.5945571427608526e-05, | |
| "loss": 0.7083, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 6.809563066776587, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 5.5687200708209076e-05, | |
| "loss": 0.704, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 6.817807089859851, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 5.542919755738275e-05, | |
| "loss": 0.7061, | |
| "step": 4135 | |
| }, | |
| { | |
| "epoch": 6.826051112943116, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 5.5171564115230254e-05, | |
| "loss": 0.7037, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 6.83429513602638, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 5.491430251878551e-05, | |
| "loss": 0.715, | |
| "step": 4145 | |
| }, | |
| { | |
| "epoch": 6.842539159109646, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 5.4657414901998095e-05, | |
| "loss": 0.7023, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 6.85078318219291, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 5.4400903395715366e-05, | |
| "loss": 0.6967, | |
| "step": 4155 | |
| }, | |
| { | |
| "epoch": 6.859027205276175, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 5.4144770127665024e-05, | |
| "loss": 0.7073, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 6.8672712283594395, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 5.388901722243724e-05, | |
| "loss": 0.6954, | |
| "step": 4165 | |
| }, | |
| { | |
| "epoch": 6.875515251442704, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 5.363364680146725e-05, | |
| "loss": 0.7044, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 6.883759274525969, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 5.3378660983017536e-05, | |
| "loss": 0.7045, | |
| "step": 4175 | |
| }, | |
| { | |
| "epoch": 6.892003297609233, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 5.31240618821604e-05, | |
| "loss": 0.7029, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 6.900247320692498, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 5.286985161076029e-05, | |
| "loss": 0.7018, | |
| "step": 4185 | |
| }, | |
| { | |
| "epoch": 6.908491343775762, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 5.2616032277456463e-05, | |
| "loss": 0.7102, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 6.916735366859028, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 5.236260598764535e-05, | |
| "loss": 0.7078, | |
| "step": 4195 | |
| }, | |
| { | |
| "epoch": 6.924979389942292, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 5.210957484346314e-05, | |
| "loss": 0.7055, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 6.933223413025557, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 5.185694094376843e-05, | |
| "loss": 0.7068, | |
| "step": 4205 | |
| }, | |
| { | |
| "epoch": 6.941467436108821, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 5.160470638412461e-05, | |
| "loss": 0.6911, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 6.949711459192086, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 5.135287325678271e-05, | |
| "loss": 0.7047, | |
| "step": 4215 | |
| }, | |
| { | |
| "epoch": 6.95795548227535, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 5.1101443650663764e-05, | |
| "loss": 0.6989, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 6.966199505358615, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 5.085041965134183e-05, | |
| "loss": 0.6975, | |
| "step": 4225 | |
| }, | |
| { | |
| "epoch": 6.9744435284418795, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 5.059980334102637e-05, | |
| "loss": 0.7055, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 6.982687551525144, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 5.034959679854532e-05, | |
| "loss": 0.6983, | |
| "step": 4235 | |
| }, | |
| { | |
| "epoch": 6.990931574608409, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 5.009980209932743e-05, | |
| "loss": 0.7046, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 6.999175597691673, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 4.985042131538545e-05, | |
| "loss": 0.7042, | |
| "step": 4245 | |
| }, | |
| { | |
| "epoch": 6.999175597691673, | |
| "eval_loss": 2.4711008071899414, | |
| "eval_runtime": 0.2631, | |
| "eval_samples_per_second": 38.011, | |
| "eval_steps_per_second": 3.801, | |
| "step": 4245 | |
| }, | |
| { | |
| "epoch": 7.007419620774938, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 4.960145651529856e-05, | |
| "loss": 0.6792, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 7.015663643858203, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 4.9352909764195576e-05, | |
| "loss": 0.6999, | |
| "step": 4255 | |
| }, | |
| { | |
| "epoch": 7.023907666941468, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 4.9104783123737566e-05, | |
| "loss": 0.6999, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 7.032151690024732, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 4.885707865210093e-05, | |
| "loss": 0.7018, | |
| "step": 4265 | |
| }, | |
| { | |
| "epoch": 7.040395713107997, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 4.860979840396016e-05, | |
| "loss": 0.6912, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 7.048639736191261, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 4.836294443047088e-05, | |
| "loss": 0.6945, | |
| "step": 4275 | |
| }, | |
| { | |
| "epoch": 7.056883759274526, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 4.8116518779252885e-05, | |
| "loss": 0.6905, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 7.0651277823577905, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 4.787052349437295e-05, | |
| "loss": 0.691, | |
| "step": 4285 | |
| }, | |
| { | |
| "epoch": 7.073371805441055, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 4.762496061632814e-05, | |
| "loss": 0.6843, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 7.08161582852432, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 4.7379832182028814e-05, | |
| "loss": 0.6951, | |
| "step": 4295 | |
| }, | |
| { | |
| "epoch": 7.089859851607584, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 4.713514022478155e-05, | |
| "loss": 0.6893, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 7.09810387469085, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 4.689088677427249e-05, | |
| "loss": 0.6952, | |
| "step": 4305 | |
| }, | |
| { | |
| "epoch": 7.106347897774114, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 4.6647073856550415e-05, | |
| "loss": 0.6958, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 7.114591920857379, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 4.6403703494009875e-05, | |
| "loss": 0.6946, | |
| "step": 4315 | |
| }, | |
| { | |
| "epoch": 7.122835943940643, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 4.6160777705374524e-05, | |
| "loss": 0.6996, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 7.131079967023908, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 4.591829850568046e-05, | |
| "loss": 0.6969, | |
| "step": 4325 | |
| }, | |
| { | |
| "epoch": 7.139323990107172, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 4.567626790625921e-05, | |
| "loss": 0.6868, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 7.147568013190437, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 4.543468791472131e-05, | |
| "loss": 0.69, | |
| "step": 4335 | |
| }, | |
| { | |
| "epoch": 7.155812036273701, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 4.519356053493958e-05, | |
| "loss": 0.6979, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 7.164056059356966, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 4.495288776703241e-05, | |
| "loss": 0.7022, | |
| "step": 4345 | |
| }, | |
| { | |
| "epoch": 7.1723000824402305, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 4.471267160734731e-05, | |
| "loss": 0.6874, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 7.180544105523495, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 4.447291404844424e-05, | |
| "loss": 0.6982, | |
| "step": 4355 | |
| }, | |
| { | |
| "epoch": 7.18878812860676, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 4.4233617079079236e-05, | |
| "loss": 0.7015, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 7.197032151690025, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 4.399478268418771e-05, | |
| "loss": 0.6919, | |
| "step": 4365 | |
| }, | |
| { | |
| "epoch": 7.20527617477329, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 4.375641284486808e-05, | |
| "loss": 0.6867, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 7.213520197856554, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 4.3518509538365425e-05, | |
| "loss": 0.6929, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 7.221764220939819, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 4.328107473805487e-05, | |
| "loss": 0.7013, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 7.230008244023083, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 4.3044110413425395e-05, | |
| "loss": 0.6879, | |
| "step": 4385 | |
| }, | |
| { | |
| "epoch": 7.238252267106348, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 4.2807618530063565e-05, | |
| "loss": 0.6918, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 7.246496290189612, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 4.257160104963696e-05, | |
| "loss": 0.6965, | |
| "step": 4395 | |
| }, | |
| { | |
| "epoch": 7.254740313272877, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 4.23360599298781e-05, | |
| "loss": 0.6963, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 7.2629843363561415, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 4.210099712456822e-05, | |
| "loss": 0.69, | |
| "step": 4405 | |
| }, | |
| { | |
| "epoch": 7.271228359439406, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 4.1866414583520877e-05, | |
| "loss": 0.6955, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 7.2794723825226715, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 4.163231425256595e-05, | |
| "loss": 0.6888, | |
| "step": 4415 | |
| }, | |
| { | |
| "epoch": 7.287716405605936, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 4.139869807353357e-05, | |
| "loss": 0.6998, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 7.295960428689201, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 4.1165567984237764e-05, | |
| "loss": 0.6963, | |
| "step": 4425 | |
| }, | |
| { | |
| "epoch": 7.304204451772465, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 4.0932925918460516e-05, | |
| "loss": 0.6922, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 7.31244847485573, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 4.070077380593579e-05, | |
| "loss": 0.6969, | |
| "step": 4435 | |
| }, | |
| { | |
| "epoch": 7.320692497938994, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 4.046911357233343e-05, | |
| "loss": 0.6893, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 7.328936521022259, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 4.02379471392431e-05, | |
| "loss": 0.6902, | |
| "step": 4445 | |
| }, | |
| { | |
| "epoch": 7.337180544105523, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 4.000727642415867e-05, | |
| "loss": 0.7053, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 7.345424567188788, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 3.977710334046193e-05, | |
| "loss": 0.6942, | |
| "step": 4455 | |
| }, | |
| { | |
| "epoch": 7.353668590272052, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 3.954742979740695e-05, | |
| "loss": 0.7078, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 7.361912613355317, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 3.9318257700104174e-05, | |
| "loss": 0.6932, | |
| "step": 4465 | |
| }, | |
| { | |
| "epoch": 7.370156636438582, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 3.9089588949504655e-05, | |
| "loss": 0.6955, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 7.378400659521847, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 3.8861425442384135e-05, | |
| "loss": 0.6969, | |
| "step": 4475 | |
| }, | |
| { | |
| "epoch": 7.3866446826051115, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 3.863376907132752e-05, | |
| "loss": 0.6949, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 7.394888705688376, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 3.840662172471315e-05, | |
| "loss": 0.7005, | |
| "step": 4485 | |
| }, | |
| { | |
| "epoch": 7.403132728771641, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 3.8179985286696986e-05, | |
| "loss": 0.6935, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 7.411376751854905, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 3.7953861637197085e-05, | |
| "loss": 0.6923, | |
| "step": 4495 | |
| }, | |
| { | |
| "epoch": 7.41962077493817, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 3.772825265187802e-05, | |
| "loss": 0.6923, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 7.427864798021434, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 3.75031602021353e-05, | |
| "loss": 0.6979, | |
| "step": 4505 | |
| }, | |
| { | |
| "epoch": 7.436108821104699, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 3.727858615507974e-05, | |
| "loss": 0.6977, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 7.444352844187963, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 3.705453237352227e-05, | |
| "loss": 0.7043, | |
| "step": 4515 | |
| }, | |
| { | |
| "epoch": 7.452596867271229, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 3.683100071595813e-05, | |
| "loss": 0.6956, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 7.460840890354493, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 3.660799303655166e-05, | |
| "loss": 0.6974, | |
| "step": 4525 | |
| }, | |
| { | |
| "epoch": 7.469084913437758, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 3.638551118512089e-05, | |
| "loss": 0.7013, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 7.4773289365210225, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 3.616355700712221e-05, | |
| "loss": 0.6966, | |
| "step": 4535 | |
| }, | |
| { | |
| "epoch": 7.485572959604287, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 3.594213234363486e-05, | |
| "loss": 0.6964, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 7.493816982687552, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 0.6922, | |
| "step": 4545 | |
| }, | |
| { | |
| "epoch": 7.502061005770816, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 3.550087890253544e-05, | |
| "loss": 0.6948, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 7.510305028854081, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 3.5281053785059925e-05, | |
| "loss": 0.695, | |
| "step": 4555 | |
| }, | |
| { | |
| "epoch": 7.518549051937345, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 3.506176550233863e-05, | |
| "loss": 0.6949, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 7.52679307502061, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 3.484301587333772e-05, | |
| "loss": 0.6903, | |
| "step": 4565 | |
| }, | |
| { | |
| "epoch": 7.535037098103874, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 3.462480671255515e-05, | |
| "loss": 0.6983, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 7.543281121187139, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 3.440713983000601e-05, | |
| "loss": 0.6964, | |
| "step": 4575 | |
| }, | |
| { | |
| "epoch": 7.551525144270404, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 3.419001703120709e-05, | |
| "loss": 0.6934, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 7.559769167353669, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 3.397344011716216e-05, | |
| "loss": 0.7035, | |
| "step": 4585 | |
| }, | |
| { | |
| "epoch": 7.568013190436933, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 3.3757410884346894e-05, | |
| "loss": 0.6827, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 7.576257213520198, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 3.354193112469407e-05, | |
| "loss": 0.6979, | |
| "step": 4595 | |
| }, | |
| { | |
| "epoch": 7.5845012366034625, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 3.332700262557864e-05, | |
| "loss": 0.7002, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 7.592745259686727, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 3.3112627169802946e-05, | |
| "loss": 0.6996, | |
| "step": 4605 | |
| }, | |
| { | |
| "epoch": 7.600989282769992, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 3.289880653558188e-05, | |
| "loss": 0.6942, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 7.609233305853256, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 3.2685542496528185e-05, | |
| "loss": 0.7002, | |
| "step": 4615 | |
| }, | |
| { | |
| "epoch": 7.617477328936521, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 3.2472836821637744e-05, | |
| "loss": 0.6953, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 7.625721352019785, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 3.2260691275274835e-05, | |
| "loss": 0.7001, | |
| "step": 4625 | |
| }, | |
| { | |
| "epoch": 7.633965375103051, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 3.204910761715763e-05, | |
| "loss": 0.6935, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 7.642209398186315, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 3.1838087602343344e-05, | |
| "loss": 0.6973, | |
| "step": 4635 | |
| }, | |
| { | |
| "epoch": 7.65045342126958, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 3.162763298121408e-05, | |
| "loss": 0.6962, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 7.658697444352844, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 3.1417745499461934e-05, | |
| "loss": 0.6986, | |
| "step": 4645 | |
| }, | |
| { | |
| "epoch": 7.666941467436109, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 3.120842689807468e-05, | |
| "loss": 0.7008, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 7.6751854905193735, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 3.099967891332132e-05, | |
| "loss": 0.698, | |
| "step": 4655 | |
| }, | |
| { | |
| "epoch": 7.683429513602638, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 3.079150327673766e-05, | |
| "loss": 0.6996, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 7.691673536685903, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 3.058390171511196e-05, | |
| "loss": 0.6973, | |
| "step": 4665 | |
| }, | |
| { | |
| "epoch": 7.699917559769167, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 3.0376875950470617e-05, | |
| "loss": 0.6972, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 7.708161582852432, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 3.0170427700063873e-05, | |
| "loss": 0.6962, | |
| "step": 4675 | |
| }, | |
| { | |
| "epoch": 7.716405605935696, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 2.996455867635155e-05, | |
| "loss": 0.7006, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 7.724649629018961, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 2.9759270586988865e-05, | |
| "loss": 0.7017, | |
| "step": 4685 | |
| }, | |
| { | |
| "epoch": 7.732893652102226, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 2.9554565134812294e-05, | |
| "loss": 0.7051, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 7.741137675185491, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 2.9350444017825385e-05, | |
| "loss": 0.6909, | |
| "step": 4695 | |
| }, | |
| { | |
| "epoch": 7.749381698268755, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 2.9146908929184713e-05, | |
| "loss": 0.6939, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 7.75762572135202, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 2.894396155718585e-05, | |
| "loss": 0.6956, | |
| "step": 4705 | |
| }, | |
| { | |
| "epoch": 7.765869744435284, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 2.874160358524931e-05, | |
| "loss": 0.6962, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 7.774113767518549, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 2.853983669190664e-05, | |
| "loss": 0.6911, | |
| "step": 4715 | |
| }, | |
| { | |
| "epoch": 7.7823577906018135, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 2.8338662550786443e-05, | |
| "loss": 0.6954, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 7.790601813685078, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 2.8138082830600554e-05, | |
| "loss": 0.694, | |
| "step": 4725 | |
| }, | |
| { | |
| "epoch": 7.798845836768343, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 2.7938099195130153e-05, | |
| "loss": 0.6935, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 7.807089859851608, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 2.7738713303211982e-05, | |
| "loss": 0.6885, | |
| "step": 4735 | |
| }, | |
| { | |
| "epoch": 7.815333882934873, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 2.753992680872457e-05, | |
| "loss": 0.7002, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 7.823577906018137, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 2.7341741360574548e-05, | |
| "loss": 0.6928, | |
| "step": 4745 | |
| }, | |
| { | |
| "epoch": 7.831821929101402, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 2.7144158602682924e-05, | |
| "loss": 0.6959, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 7.840065952184666, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 2.6947180173971508e-05, | |
| "loss": 0.6907, | |
| "step": 4755 | |
| }, | |
| { | |
| "epoch": 7.848309975267931, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 2.6750807708349267e-05, | |
| "loss": 0.6982, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 7.856553998351195, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 2.6555042834698773e-05, | |
| "loss": 0.6945, | |
| "step": 4765 | |
| }, | |
| { | |
| "epoch": 7.86479802143446, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 2.6359887176862718e-05, | |
| "loss": 0.695, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 7.8730420445177245, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 2.6165342353630428e-05, | |
| "loss": 0.694, | |
| "step": 4775 | |
| }, | |
| { | |
| "epoch": 7.881286067600989, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 2.5971409978724458e-05, | |
| "loss": 0.6986, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 7.889530090684254, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 2.577809166078716e-05, | |
| "loss": 0.6935, | |
| "step": 4785 | |
| }, | |
| { | |
| "epoch": 7.897774113767518, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 2.558538900336741e-05, | |
| "loss": 0.6991, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 7.906018136850783, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 2.5393303604907205e-05, | |
| "loss": 0.6974, | |
| "step": 4795 | |
| }, | |
| { | |
| "epoch": 7.914262159934048, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 2.5201837058728505e-05, | |
| "loss": 0.6956, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 7.922506183017313, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 2.5010990953019975e-05, | |
| "loss": 0.6927, | |
| "step": 4805 | |
| }, | |
| { | |
| "epoch": 7.930750206100577, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 2.4820766870823807e-05, | |
| "loss": 0.688, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 7.938994229183842, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 2.4631166390022574e-05, | |
| "loss": 0.695, | |
| "step": 4815 | |
| }, | |
| { | |
| "epoch": 7.947238252267106, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 2.4442191083326195e-05, | |
| "loss": 0.7014, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 7.955482275350371, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 2.425384251825882e-05, | |
| "loss": 0.6955, | |
| "step": 4825 | |
| }, | |
| { | |
| "epoch": 7.963726298433635, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 2.4066122257145894e-05, | |
| "loss": 0.6934, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 7.9719703215169, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 2.387903185710115e-05, | |
| "loss": 0.6909, | |
| "step": 4835 | |
| }, | |
| { | |
| "epoch": 7.9802143446001645, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 2.3692572870013718e-05, | |
| "loss": 0.691, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 7.98845836768343, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 2.3506746842535242e-05, | |
| "loss": 0.6929, | |
| "step": 4845 | |
| }, | |
| { | |
| "epoch": 7.9967023907666945, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 2.3321555316067045e-05, | |
| "loss": 0.6928, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.471337080001831, | |
| "eval_runtime": 0.2361, | |
| "eval_samples_per_second": 42.357, | |
| "eval_steps_per_second": 4.236, | |
| "step": 4852 | |
| }, | |
| { | |
| "epoch": 8.004946413849959, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 2.313699982674736e-05, | |
| "loss": 0.6913, | |
| "step": 4855 | |
| }, | |
| { | |
| "epoch": 8.013190436933224, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 2.295308190543859e-05, | |
| "loss": 0.6943, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 8.021434460016488, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 2.276980307771458e-05, | |
| "loss": 0.6958, | |
| "step": 4865 | |
| }, | |
| { | |
| "epoch": 8.029678483099753, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 2.2587164863847975e-05, | |
| "loss": 0.6957, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 8.037922506183017, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 2.2405168778797646e-05, | |
| "loss": 0.6914, | |
| "step": 4875 | |
| }, | |
| { | |
| "epoch": 8.046166529266282, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 2.222381633219608e-05, | |
| "loss": 0.6904, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 8.054410552349546, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 2.204310902833685e-05, | |
| "loss": 0.6921, | |
| "step": 4885 | |
| }, | |
| { | |
| "epoch": 8.062654575432811, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 2.1863048366162208e-05, | |
| "loss": 0.6926, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 8.070898598516075, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 2.1683635839250537e-05, | |
| "loss": 0.6938, | |
| "step": 4895 | |
| }, | |
| { | |
| "epoch": 8.07914262159934, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 2.15048729358041e-05, | |
| "loss": 0.6936, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 8.087386644682605, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 2.1326761138636553e-05, | |
| "loss": 0.6959, | |
| "step": 4905 | |
| }, | |
| { | |
| "epoch": 8.09563066776587, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 2.114930192516076e-05, | |
| "loss": 0.6883, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 8.103874690849134, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 2.097249676737648e-05, | |
| "loss": 0.6989, | |
| "step": 4915 | |
| }, | |
| { | |
| "epoch": 8.112118713932398, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 2.0796347131858186e-05, | |
| "loss": 0.6915, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 8.120362737015663, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 2.0620854479742834e-05, | |
| "loss": 0.6893, | |
| "step": 4925 | |
| }, | |
| { | |
| "epoch": 8.12860676009893, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 2.044602026671786e-05, | |
| "loss": 0.699, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 8.136850783182194, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 2.027184594300898e-05, | |
| "loss": 0.6962, | |
| "step": 4935 | |
| }, | |
| { | |
| "epoch": 8.145094806265458, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 2.0098332953368272e-05, | |
| "loss": 0.6869, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 8.153338829348723, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 1.9925482737062085e-05, | |
| "loss": 0.6957, | |
| "step": 4945 | |
| }, | |
| { | |
| "epoch": 8.161582852431987, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 1.9753296727859195e-05, | |
| "loss": 0.692, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 8.169826875515252, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 1.9581776354018854e-05, | |
| "loss": 0.6985, | |
| "step": 4955 | |
| }, | |
| { | |
| "epoch": 8.178070898598516, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 1.941092303827896e-05, | |
| "loss": 0.6876, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 8.186314921681781, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 1.9240738197844278e-05, | |
| "loss": 0.6863, | |
| "step": 4965 | |
| }, | |
| { | |
| "epoch": 8.194558944765046, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 1.9071223244374614e-05, | |
| "loss": 0.694, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 8.20280296784831, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 1.8902379583973208e-05, | |
| "loss": 0.6936, | |
| "step": 4975 | |
| }, | |
| { | |
| "epoch": 8.211046990931575, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 1.8734208617174988e-05, | |
| "loss": 0.6926, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 8.21929101401484, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 1.856671173893497e-05, | |
| "loss": 0.6921, | |
| "step": 4985 | |
| }, | |
| { | |
| "epoch": 8.227535037098104, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 1.839989033861673e-05, | |
| "loss": 0.6893, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 8.235779060181368, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 1.8233745799980817e-05, | |
| "loss": 0.6931, | |
| "step": 4995 | |
| }, | |
| { | |
| "epoch": 8.244023083264633, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 1.8068279501173335e-05, | |
| "loss": 0.6842, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 8.252267106347897, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 1.790349281471445e-05, | |
| "loss": 0.6998, | |
| "step": 5005 | |
| }, | |
| { | |
| "epoch": 8.260511129431162, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 1.773938710748706e-05, | |
| "loss": 0.6946, | |
| "step": 5010 | |
| }, | |
| { | |
| "epoch": 8.268755152514426, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 1.757596374072543e-05, | |
| "loss": 0.6901, | |
| "step": 5015 | |
| }, | |
| { | |
| "epoch": 8.276999175597691, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 1.741322407000391e-05, | |
| "loss": 0.6938, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 8.285243198680956, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 1.7251169445225657e-05, | |
| "loss": 0.6922, | |
| "step": 5025 | |
| }, | |
| { | |
| "epoch": 8.29348722176422, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 1.70898012106115e-05, | |
| "loss": 0.6844, | |
| "step": 5030 | |
| }, | |
| { | |
| "epoch": 8.301731244847485, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 1.692912070468874e-05, | |
| "loss": 0.6968, | |
| "step": 5035 | |
| }, | |
| { | |
| "epoch": 8.309975267930751, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 1.676912926028007e-05, | |
| "loss": 0.6977, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 8.318219291014016, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 1.660982820449247e-05, | |
| "loss": 0.6995, | |
| "step": 5045 | |
| }, | |
| { | |
| "epoch": 8.32646331409728, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 1.6451218858706374e-05, | |
| "loss": 0.6934, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 8.334707337180545, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 1.6293302538564382e-05, | |
| "loss": 0.6954, | |
| "step": 5055 | |
| }, | |
| { | |
| "epoch": 8.34295136026381, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 1.6136080553960687e-05, | |
| "loss": 0.6942, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 8.351195383347074, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 1.5979554209030024e-05, | |
| "loss": 0.6887, | |
| "step": 5065 | |
| }, | |
| { | |
| "epoch": 8.359439406430338, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 1.5823724802136865e-05, | |
| "loss": 0.6948, | |
| "step": 5070 | |
| }, | |
| { | |
| "epoch": 8.367683429513603, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 1.5668593625864715e-05, | |
| "loss": 0.695, | |
| "step": 5075 | |
| }, | |
| { | |
| "epoch": 8.375927452596867, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 1.5514161967005337e-05, | |
| "loss": 0.7057, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 8.384171475680132, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 1.536043110654809e-05, | |
| "loss": 0.6906, | |
| "step": 5085 | |
| }, | |
| { | |
| "epoch": 8.392415498763397, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 1.5207402319669306e-05, | |
| "loss": 0.6909, | |
| "step": 5090 | |
| }, | |
| { | |
| "epoch": 8.400659521846661, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 1.505507687572173e-05, | |
| "loss": 0.6841, | |
| "step": 5095 | |
| }, | |
| { | |
| "epoch": 8.408903544929926, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 1.4903456038223939e-05, | |
| "loss": 0.6889, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 8.41714756801319, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 1.4752541064849946e-05, | |
| "loss": 0.6908, | |
| "step": 5105 | |
| }, | |
| { | |
| "epoch": 8.425391591096455, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 1.4602333207418651e-05, | |
| "loss": 0.6949, | |
| "step": 5110 | |
| }, | |
| { | |
| "epoch": 8.43363561417972, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 1.4452833711883628e-05, | |
| "loss": 0.691, | |
| "step": 5115 | |
| }, | |
| { | |
| "epoch": 8.441879637262984, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 1.4304043818322565e-05, | |
| "loss": 0.6855, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 8.450123660346248, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 1.4155964760927176e-05, | |
| "loss": 0.6937, | |
| "step": 5125 | |
| }, | |
| { | |
| "epoch": 8.458367683429513, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 1.4008597767992871e-05, | |
| "loss": 0.6922, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 8.466611706512778, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 1.3861944061908583e-05, | |
| "loss": 0.6929, | |
| "step": 5135 | |
| }, | |
| { | |
| "epoch": 8.474855729596042, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 1.3716004859146592e-05, | |
| "loss": 0.6898, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 8.483099752679308, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 1.3570781370252582e-05, | |
| "loss": 0.6851, | |
| "step": 5145 | |
| }, | |
| { | |
| "epoch": 8.491343775762573, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 1.3426274799835337e-05, | |
| "loss": 0.6846, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 8.499587798845837, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 1.328248634655701e-05, | |
| "loss": 0.7024, | |
| "step": 5155 | |
| }, | |
| { | |
| "epoch": 8.507831821929102, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 1.3139417203123027e-05, | |
| "loss": 0.6881, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 8.516075845012367, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 1.2997068556272263e-05, | |
| "loss": 0.7002, | |
| "step": 5165 | |
| }, | |
| { | |
| "epoch": 8.524319868095631, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 1.2855441586767113e-05, | |
| "loss": 0.6909, | |
| "step": 5170 | |
| }, | |
| { | |
| "epoch": 8.532563891178896, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 1.2714537469383858e-05, | |
| "loss": 0.6878, | |
| "step": 5175 | |
| }, | |
| { | |
| "epoch": 8.54080791426216, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 1.2574357372902767e-05, | |
| "loss": 0.6917, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 8.549051937345425, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 1.243490246009842e-05, | |
| "loss": 0.689, | |
| "step": 5185 | |
| }, | |
| { | |
| "epoch": 8.55729596042869, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 1.2296173887730123e-05, | |
| "loss": 0.6859, | |
| "step": 5190 | |
| }, | |
| { | |
| "epoch": 8.565539983511954, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 1.215817280653232e-05, | |
| "loss": 0.6858, | |
| "step": 5195 | |
| }, | |
| { | |
| "epoch": 8.573784006595218, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 1.2020900361204968e-05, | |
| "loss": 0.6894, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 8.582028029678483, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 1.1884357690404158e-05, | |
| "loss": 0.6886, | |
| "step": 5205 | |
| }, | |
| { | |
| "epoch": 8.590272052761748, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 1.1748545926732535e-05, | |
| "loss": 0.6903, | |
| "step": 5210 | |
| }, | |
| { | |
| "epoch": 8.598516075845012, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 1.1613466196729984e-05, | |
| "loss": 0.7021, | |
| "step": 5215 | |
| }, | |
| { | |
| "epoch": 8.606760098928277, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 1.1479119620864276e-05, | |
| "loss": 0.6826, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 8.615004122011541, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 1.1345507313521786e-05, | |
| "loss": 0.6954, | |
| "step": 5225 | |
| }, | |
| { | |
| "epoch": 8.623248145094806, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 1.1212630382998213e-05, | |
| "loss": 0.6877, | |
| "step": 5230 | |
| }, | |
| { | |
| "epoch": 8.63149216817807, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 1.1080489931489391e-05, | |
| "loss": 0.696, | |
| "step": 5235 | |
| }, | |
| { | |
| "epoch": 8.639736191261335, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 1.0949087055082252e-05, | |
| "loss": 0.6977, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 8.6479802143446, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 1.0818422843745512e-05, | |
| "loss": 0.6924, | |
| "step": 5245 | |
| }, | |
| { | |
| "epoch": 8.656224237427864, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 1.0688498381320855e-05, | |
| "loss": 0.6941, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 8.664468260511129, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 1.0559314745513805e-05, | |
| "loss": 0.6878, | |
| "step": 5255 | |
| }, | |
| { | |
| "epoch": 8.672712283594395, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 1.0430873007884857e-05, | |
| "loss": 0.6975, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 8.68095630667766, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 1.0303174233840528e-05, | |
| "loss": 0.6863, | |
| "step": 5265 | |
| }, | |
| { | |
| "epoch": 8.689200329760924, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 1.0176219482624616e-05, | |
| "loss": 0.7022, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 8.697444352844188, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 1.0050009807309325e-05, | |
| "loss": 0.6892, | |
| "step": 5275 | |
| }, | |
| { | |
| "epoch": 8.705688375927453, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 9.924546254786493e-06, | |
| "loss": 0.6839, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 8.713932399010718, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 9.799829865759069e-06, | |
| "loss": 0.6821, | |
| "step": 5285 | |
| }, | |
| { | |
| "epoch": 8.722176422093982, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 9.675861674732312e-06, | |
| "loss": 0.6885, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 8.730420445177247, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 9.552642710005299e-06, | |
| "loss": 0.6965, | |
| "step": 5295 | |
| }, | |
| { | |
| "epoch": 8.738664468260511, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 9.430173993662451e-06, | |
| "loss": 0.6971, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 8.746908491343776, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 9.308456541564881e-06, | |
| "loss": 0.6847, | |
| "step": 5305 | |
| }, | |
| { | |
| "epoch": 8.75515251442704, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 9.187491363342093e-06, | |
| "loss": 0.6982, | |
| "step": 5310 | |
| }, | |
| { | |
| "epoch": 8.763396537510305, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 9.067279462383615e-06, | |
| "loss": 0.6906, | |
| "step": 5315 | |
| }, | |
| { | |
| "epoch": 8.77164056059357, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 8.947821835830616e-06, | |
| "loss": 0.6981, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 8.779884583676834, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 8.829119474567671e-06, | |
| "loss": 0.6972, | |
| "step": 5325 | |
| }, | |
| { | |
| "epoch": 8.788128606760099, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 8.711173363214553e-06, | |
| "loss": 0.6875, | |
| "step": 5330 | |
| }, | |
| { | |
| "epoch": 8.796372629843363, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 8.593984480118011e-06, | |
| "loss": 0.6904, | |
| "step": 5335 | |
| }, | |
| { | |
| "epoch": 8.804616652926628, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 8.47755379734373e-06, | |
| "loss": 0.6886, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 8.812860676009892, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 8.361882280668165e-06, | |
| "loss": 0.6919, | |
| "step": 5345 | |
| }, | |
| { | |
| "epoch": 8.821104699093157, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 8.24697088957066e-06, | |
| "loss": 0.6934, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 8.829348722176421, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 8.132820577225387e-06, | |
| "loss": 0.6882, | |
| "step": 5355 | |
| }, | |
| { | |
| "epoch": 8.837592745259688, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 8.019432290493457e-06, | |
| "loss": 0.7015, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 8.845836768342952, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 7.906806969915148e-06, | |
| "loss": 0.689, | |
| "step": 5365 | |
| }, | |
| { | |
| "epoch": 8.854080791426217, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 7.794945549701993e-06, | |
| "loss": 0.6866, | |
| "step": 5370 | |
| }, | |
| { | |
| "epoch": 8.862324814509481, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 7.683848957729056e-06, | |
| "loss": 0.696, | |
| "step": 5375 | |
| }, | |
| { | |
| "epoch": 8.870568837592746, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 7.573518115527289e-06, | |
| "loss": 0.6824, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 8.87881286067601, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 7.463953938275858e-06, | |
| "loss": 0.6941, | |
| "step": 5385 | |
| }, | |
| { | |
| "epoch": 8.887056883759275, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 7.355157334794516e-06, | |
| "loss": 0.6901, | |
| "step": 5390 | |
| }, | |
| { | |
| "epoch": 8.89530090684254, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 7.247129207536152e-06, | |
| "loss": 0.688, | |
| "step": 5395 | |
| }, | |
| { | |
| "epoch": 8.903544929925804, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 7.1398704525792e-06, | |
| "loss": 0.6906, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 8.911788953009069, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 7.0333819596203e-06, | |
| "loss": 0.6878, | |
| "step": 5405 | |
| }, | |
| { | |
| "epoch": 8.920032976092333, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 6.927664611966811e-06, | |
| "loss": 0.6965, | |
| "step": 5410 | |
| }, | |
| { | |
| "epoch": 8.928276999175598, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 6.8227192865295995e-06, | |
| "loss": 0.69, | |
| "step": 5415 | |
| }, | |
| { | |
| "epoch": 8.936521022258862, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 6.718546853815688e-06, | |
| "loss": 0.6857, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 8.944765045342127, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 6.6151481779211155e-06, | |
| "loss": 0.6922, | |
| "step": 5425 | |
| }, | |
| { | |
| "epoch": 8.953009068425391, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 6.512524116523633e-06, | |
| "loss": 0.6885, | |
| "step": 5430 | |
| }, | |
| { | |
| "epoch": 8.961253091508656, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 6.410675520875742e-06, | |
| "loss": 0.6854, | |
| "step": 5435 | |
| }, | |
| { | |
| "epoch": 8.96949711459192, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 6.30960323579749e-06, | |
| "loss": 0.6966, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 8.977741137675185, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 6.209308099669597e-06, | |
| "loss": 0.6962, | |
| "step": 5445 | |
| }, | |
| { | |
| "epoch": 8.98598516075845, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 6.109790944426397e-06, | |
| "loss": 0.707, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 8.994229183841714, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 6.011052595549038e-06, | |
| "loss": 0.6924, | |
| "step": 5455 | |
| }, | |
| { | |
| "epoch": 8.999175597691673, | |
| "eval_loss": 2.4814510345458984, | |
| "eval_runtime": 0.2587, | |
| "eval_samples_per_second": 38.654, | |
| "eval_steps_per_second": 3.865, | |
| "step": 5458 | |
| }, | |
| { | |
| "epoch": 9.002473206924979, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 5.913093872058528e-06, | |
| "loss": 0.6875, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 9.010717230008243, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 5.81591558650898e-06, | |
| "loss": 0.6871, | |
| "step": 5465 | |
| }, | |
| { | |
| "epoch": 9.01896125309151, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 5.719518544980929e-06, | |
| "loss": 0.6887, | |
| "step": 5470 | |
| }, | |
| { | |
| "epoch": 9.027205276174774, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 5.623903547074549e-06, | |
| "loss": 0.7051, | |
| "step": 5475 | |
| }, | |
| { | |
| "epoch": 9.035449299258039, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 5.529071385903084e-06, | |
| "loss": 0.694, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 9.043693322341303, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 5.43502284808628e-06, | |
| "loss": 0.6839, | |
| "step": 5485 | |
| }, | |
| { | |
| "epoch": 9.051937345424568, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 5.341758713743828e-06, | |
| "loss": 0.6906, | |
| "step": 5490 | |
| }, | |
| { | |
| "epoch": 9.060181368507832, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 5.249279756488878e-06, | |
| "loss": 0.6895, | |
| "step": 5495 | |
| }, | |
| { | |
| "epoch": 9.068425391591097, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 5.157586743421672e-06, | |
| "loss": 0.6937, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 9.076669414674361, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 5.066680435123106e-06, | |
| "loss": 0.7007, | |
| "step": 5505 | |
| }, | |
| { | |
| "epoch": 9.084913437757626, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 4.976561585648509e-06, | |
| "loss": 0.6864, | |
| "step": 5510 | |
| }, | |
| { | |
| "epoch": 9.09315746084089, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 4.887230942521337e-06, | |
| "loss": 0.6886, | |
| "step": 5515 | |
| }, | |
| { | |
| "epoch": 9.101401483924155, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 4.798689246727006e-06, | |
| "loss": 0.6965, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 9.10964550700742, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 4.710937232706691e-06, | |
| "loss": 0.6888, | |
| "step": 5525 | |
| }, | |
| { | |
| "epoch": 9.117889530090684, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 4.623975628351273e-06, | |
| "loss": 0.6937, | |
| "step": 5530 | |
| }, | |
| { | |
| "epoch": 9.126133553173949, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 4.537805154995278e-06, | |
| "loss": 0.6989, | |
| "step": 5535 | |
| }, | |
| { | |
| "epoch": 9.134377576257213, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 4.452426527410947e-06, | |
| "loss": 0.69, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 9.142621599340478, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 4.36784045380223e-06, | |
| "loss": 0.6952, | |
| "step": 5545 | |
| }, | |
| { | |
| "epoch": 9.150865622423742, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 4.2840476357989825e-06, | |
| "loss": 0.6909, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 9.159109645507007, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 4.20104876845111e-06, | |
| "loss": 0.6835, | |
| "step": 5555 | |
| }, | |
| { | |
| "epoch": 9.167353668590271, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 4.118844540222788e-06, | |
| "loss": 0.7042, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 9.175597691673536, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 4.037435632986786e-06, | |
| "loss": 0.693, | |
| "step": 5565 | |
| }, | |
| { | |
| "epoch": 9.1838417147568, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 3.95682272201876e-06, | |
| "loss": 0.6854, | |
| "step": 5570 | |
| }, | |
| { | |
| "epoch": 9.192085737840065, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 3.877006475991729e-06, | |
| "loss": 0.6937, | |
| "step": 5575 | |
| }, | |
| { | |
| "epoch": 9.200329760923331, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 3.797987556970495e-06, | |
| "loss": 0.6968, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 9.208573784006596, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 3.7197666204060955e-06, | |
| "loss": 0.6902, | |
| "step": 5585 | |
| }, | |
| { | |
| "epoch": 9.21681780708986, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 3.6423443151304526e-06, | |
| "loss": 0.6896, | |
| "step": 5590 | |
| }, | |
| { | |
| "epoch": 9.225061830173125, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 3.565721283350931e-06, | |
| "loss": 0.696, | |
| "step": 5595 | |
| }, | |
| { | |
| "epoch": 9.23330585325639, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 3.4898981606450333e-06, | |
| "loss": 0.6895, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 9.241549876339654, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 3.414875575955101e-06, | |
| "loss": 0.6845, | |
| "step": 5605 | |
| }, | |
| { | |
| "epoch": 9.249793899422919, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 3.3406541515832003e-06, | |
| "loss": 0.6908, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 9.258037922506183, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 3.267234503185823e-06, | |
| "loss": 0.6885, | |
| "step": 5615 | |
| }, | |
| { | |
| "epoch": 9.266281945589448, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 3.1946172397688267e-06, | |
| "loss": 0.6921, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 9.274525968672712, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 3.1228029636824475e-06, | |
| "loss": 0.6927, | |
| "step": 5625 | |
| }, | |
| { | |
| "epoch": 9.282769991755977, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 3.051792270616216e-06, | |
| "loss": 0.689, | |
| "step": 5630 | |
| }, | |
| { | |
| "epoch": 9.291014014839241, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 2.981585749594051e-06, | |
| "loss": 0.6962, | |
| "step": 5635 | |
| }, | |
| { | |
| "epoch": 9.299258037922506, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 2.912183982969385e-06, | |
| "loss": 0.6873, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 9.30750206100577, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 2.8435875464203343e-06, | |
| "loss": 0.6839, | |
| "step": 5645 | |
| }, | |
| { | |
| "epoch": 9.315746084089035, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 2.7757970089449024e-06, | |
| "loss": 0.6884, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 9.3239901071723, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 2.708812932856253e-06, | |
| "loss": 0.6865, | |
| "step": 5655 | |
| }, | |
| { | |
| "epoch": 9.332234130255564, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 2.6426358737781098e-06, | |
| "loss": 0.6944, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 9.340478153338829, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 2.577266380640053e-06, | |
| "loss": 0.6866, | |
| "step": 5665 | |
| }, | |
| { | |
| "epoch": 9.348722176422093, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 2.5127049956730207e-06, | |
| "loss": 0.6917, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 9.356966199505358, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 2.448952254404846e-06, | |
| "loss": 0.6984, | |
| "step": 5675 | |
| }, | |
| { | |
| "epoch": 9.365210222588622, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 2.3860086856557383e-06, | |
| "loss": 0.6881, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 9.373454245671887, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 2.3238748115339324e-06, | |
| "loss": 0.689, | |
| "step": 5685 | |
| }, | |
| { | |
| "epoch": 9.381698268755153, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 2.2625511474313685e-06, | |
| "loss": 0.6968, | |
| "step": 5690 | |
| }, | |
| { | |
| "epoch": 9.389942291838418, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 2.2020382020194074e-06, | |
| "loss": 0.6923, | |
| "step": 5695 | |
| }, | |
| { | |
| "epoch": 9.398186314921682, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 2.1423364772445887e-06, | |
| "loss": 0.6929, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 9.406430338004947, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 2.0834464683245346e-06, | |
| "loss": 0.6948, | |
| "step": 5705 | |
| }, | |
| { | |
| "epoch": 9.414674361088212, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 2.025368663743743e-06, | |
| "loss": 0.6956, | |
| "step": 5710 | |
| }, | |
| { | |
| "epoch": 9.422918384171476, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 1.968103545249611e-06, | |
| "loss": 0.6857, | |
| "step": 5715 | |
| }, | |
| { | |
| "epoch": 9.43116240725474, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 1.91165158784844e-06, | |
| "loss": 0.6871, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 9.439406430338005, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 1.8560132598014368e-06, | |
| "loss": 0.6864, | |
| "step": 5725 | |
| }, | |
| { | |
| "epoch": 9.44765045342127, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 1.8011890226208527e-06, | |
| "loss": 0.6922, | |
| "step": 5730 | |
| }, | |
| { | |
| "epoch": 9.455894476504534, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 1.7471793310662287e-06, | |
| "loss": 0.6973, | |
| "step": 5735 | |
| }, | |
| { | |
| "epoch": 9.464138499587799, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 1.6939846331405108e-06, | |
| "loss": 0.6954, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 9.472382522671063, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 1.6416053700863964e-06, | |
| "loss": 0.6983, | |
| "step": 5745 | |
| }, | |
| { | |
| "epoch": 9.480626545754328, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 1.5900419763826614e-06, | |
| "loss": 0.6904, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 9.488870568837593, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 1.5392948797405827e-06, | |
| "loss": 0.7001, | |
| "step": 5755 | |
| }, | |
| { | |
| "epoch": 9.497114591920857, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 1.489364501100332e-06, | |
| "loss": 0.6978, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 9.505358615004122, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 1.4402512546275114e-06, | |
| "loss": 0.6974, | |
| "step": 5765 | |
| }, | |
| { | |
| "epoch": 9.513602638087386, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 1.3919555477097668e-06, | |
| "loss": 0.6885, | |
| "step": 5770 | |
| }, | |
| { | |
| "epoch": 9.52184666117065, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 1.344477780953346e-06, | |
| "loss": 0.6884, | |
| "step": 5775 | |
| }, | |
| { | |
| "epoch": 9.530090684253915, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 1.2978183481797801e-06, | |
| "loss": 0.6899, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 9.53833470733718, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 1.251977636422641e-06, | |
| "loss": 0.6897, | |
| "step": 5785 | |
| }, | |
| { | |
| "epoch": 9.546578730420444, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 1.2069560259243328e-06, | |
| "loss": 0.6933, | |
| "step": 5790 | |
| }, | |
| { | |
| "epoch": 9.55482275350371, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 1.1627538901329172e-06, | |
| "loss": 0.6868, | |
| "step": 5795 | |
| }, | |
| { | |
| "epoch": 9.563066776586975, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 1.1193715956990258e-06, | |
| "loss": 0.6855, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 9.57131079967024, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 1.076809502472831e-06, | |
| "loss": 0.6977, | |
| "step": 5805 | |
| }, | |
| { | |
| "epoch": 9.579554822753504, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 1.035067963501024e-06, | |
| "loss": 0.6969, | |
| "step": 5810 | |
| }, | |
| { | |
| "epoch": 9.587798845836769, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 9.94147325023953e-07, | |
| "loss": 0.6982, | |
| "step": 5815 | |
| }, | |
| { | |
| "epoch": 9.596042868920033, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 9.540479264726676e-07, | |
| "loss": 0.6865, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 9.604286892003298, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 9.147701004661446e-07, | |
| "loss": 0.6897, | |
| "step": 5825 | |
| }, | |
| { | |
| "epoch": 9.612530915086563, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 8.763141728085789e-07, | |
| "loss": 0.6837, | |
| "step": 5830 | |
| }, | |
| { | |
| "epoch": 9.620774938169827, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 8.386804624865851e-07, | |
| "loss": 0.6865, | |
| "step": 5835 | |
| }, | |
| { | |
| "epoch": 9.629018961253092, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 8.018692816666118e-07, | |
| "loss": 0.6907, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 9.637262984336356, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 7.658809356923424e-07, | |
| "loss": 0.6902, | |
| "step": 5845 | |
| }, | |
| { | |
| "epoch": 9.64550700741962, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 7.307157230821426e-07, | |
| "loss": 0.6925, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 9.653751030502885, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 6.963739355266286e-07, | |
| "loss": 0.6911, | |
| "step": 5855 | |
| }, | |
| { | |
| "epoch": 9.66199505358615, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 6.628558578862021e-07, | |
| "loss": 0.6838, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 9.670239076669414, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 6.301617681886863e-07, | |
| "loss": 0.6883, | |
| "step": 5865 | |
| }, | |
| { | |
| "epoch": 9.678483099752679, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 5.982919376270823e-07, | |
| "loss": 0.6908, | |
| "step": 5870 | |
| }, | |
| { | |
| "epoch": 9.686727122835944, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 5.672466305572388e-07, | |
| "loss": 0.6908, | |
| "step": 5875 | |
| }, | |
| { | |
| "epoch": 9.694971145919208, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 5.370261044956971e-07, | |
| "loss": 0.6962, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 9.703215169002473, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 5.07630610117582e-07, | |
| "loss": 0.6932, | |
| "step": 5885 | |
| }, | |
| { | |
| "epoch": 9.711459192085737, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 4.790603912544489e-07, | |
| "loss": 0.6878, | |
| "step": 5890 | |
| }, | |
| { | |
| "epoch": 9.719703215169002, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 4.5131568489236166e-07, | |
| "loss": 0.6946, | |
| "step": 5895 | |
| }, | |
| { | |
| "epoch": 9.727947238252266, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 4.2439672116982855e-07, | |
| "loss": 0.6853, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 9.73619126133553, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 3.983037233759368e-07, | |
| "loss": 0.6914, | |
| "step": 5905 | |
| }, | |
| { | |
| "epoch": 9.744435284418797, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 3.73036907948543e-07, | |
| "loss": 0.6898, | |
| "step": 5910 | |
| }, | |
| { | |
| "epoch": 9.752679307502062, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 3.485964844723744e-07, | |
| "loss": 0.6888, | |
| "step": 5915 | |
| }, | |
| { | |
| "epoch": 9.760923330585326, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 3.2498265567739717e-07, | |
| "loss": 0.6824, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 9.76916735366859, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 3.0219561743707326e-07, | |
| "loss": 0.691, | |
| "step": 5925 | |
| }, | |
| { | |
| "epoch": 9.777411376751855, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 2.8023555876673937e-07, | |
| "loss": 0.6862, | |
| "step": 5930 | |
| }, | |
| { | |
| "epoch": 9.78565539983512, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 2.5910266182207486e-07, | |
| "loss": 0.6933, | |
| "step": 5935 | |
| }, | |
| { | |
| "epoch": 9.793899422918384, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 2.3879710189753656e-07, | |
| "loss": 0.6926, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 9.802143446001649, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 2.1931904742495957e-07, | |
| "loss": 0.6807, | |
| "step": 5945 | |
| }, | |
| { | |
| "epoch": 9.810387469084914, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 2.0066865997212525e-07, | |
| "loss": 0.6923, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 9.818631492168178, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 1.8284609424142895e-07, | |
| "loss": 0.6885, | |
| "step": 5955 | |
| }, | |
| { | |
| "epoch": 9.826875515251443, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 1.6585149806860324e-07, | |
| "loss": 0.6862, | |
| "step": 5960 | |
| }, | |
| { | |
| "epoch": 9.835119538334707, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 1.4968501242148547e-07, | |
| "loss": 0.6955, | |
| "step": 5965 | |
| }, | |
| { | |
| "epoch": 9.843363561417972, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 1.3434677139885222e-07, | |
| "loss": 0.6957, | |
| "step": 5970 | |
| }, | |
| { | |
| "epoch": 9.851607584501236, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 1.1983690222929778e-07, | |
| "loss": 0.6915, | |
| "step": 5975 | |
| }, | |
| { | |
| "epoch": 9.8598516075845, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 1.0615552527017958e-07, | |
| "loss": 0.701, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 9.868095630667765, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 9.330275400666332e-08, | |
| "loss": 0.6959, | |
| "step": 5985 | |
| }, | |
| { | |
| "epoch": 9.87633965375103, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 8.127869505069053e-08, | |
| "loss": 0.6885, | |
| "step": 5990 | |
| }, | |
| { | |
| "epoch": 9.884583676834295, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 7.00834481402013e-08, | |
| "loss": 0.6842, | |
| "step": 5995 | |
| }, | |
| { | |
| "epoch": 9.892827699917559, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 5.971710613821291e-08, | |
| "loss": 0.6956, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 9.901071723000824, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 5.0179755032109253e-08, | |
| "loss": 0.6898, | |
| "step": 6005 | |
| }, | |
| { | |
| "epoch": 9.90931574608409, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 4.147147393290807e-08, | |
| "loss": 0.6899, | |
| "step": 6010 | |
| }, | |
| { | |
| "epoch": 9.917559769167354, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 3.359233507459481e-08, | |
| "loss": 0.697, | |
| "step": 6015 | |
| }, | |
| { | |
| "epoch": 9.925803792250619, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 2.6542403813545334e-08, | |
| "loss": 0.6938, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 9.934047815333884, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 2.0321738627981923e-08, | |
| "loss": 0.686, | |
| "step": 6025 | |
| }, | |
| { | |
| "epoch": 9.942291838417148, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 1.4930391117451426e-08, | |
| "loss": 0.6874, | |
| "step": 6030 | |
| }, | |
| { | |
| "epoch": 9.950535861500413, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 1.0368406002436715e-08, | |
| "loss": 0.6934, | |
| "step": 6035 | |
| }, | |
| { | |
| "epoch": 9.958779884583677, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 6.635821124001406e-09, | |
| "loss": 0.6913, | |
| "step": 6040 | |
| }, | |
| { | |
| "epoch": 9.967023907666942, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 3.732667443390181e-09, | |
| "loss": 0.6895, | |
| "step": 6045 | |
| }, | |
| { | |
| "epoch": 9.975267930750206, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 1.6589690418955528e-09, | |
| "loss": 0.6968, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 9.983511953833471, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 4.147431205359098e-10, | |
| "loss": 0.6946, | |
| "step": 6055 | |
| }, | |
| { | |
| "epoch": 9.991755976916735, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.0, | |
| "loss": 0.6936, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 9.991755976916735, | |
| "eval_loss": 2.4860482215881348, | |
| "eval_runtime": 0.2343, | |
| "eval_samples_per_second": 42.675, | |
| "eval_steps_per_second": 4.267, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 9.991755976916735, | |
| "step": 6060, | |
| "total_flos": 1.8500974249565487e+19, | |
| "train_loss": 1.1020318522705104, | |
| "train_runtime": 14653.0399, | |
| "train_samples_per_second": 26.478, | |
| "train_steps_per_second": 0.414 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 6060, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "total_flos": 1.8500974249565487e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |