Invalid JSON:Unexpected token 'N', ..."ad_norm": NaN,
"... is not valid JSON
| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.07349563619660082, | |
| "eval_steps": 500, | |
| "global_step": 4200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0005249688299757202, | |
| "grad_norm": 4051.18359375, | |
| "learning_rate": 9e-07, | |
| "loss": 7.93, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0010499376599514404, | |
| "grad_norm": 111.26915740966797, | |
| "learning_rate": 3.899999999999999e-06, | |
| "loss": 7.7273, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0015749064899271605, | |
| "grad_norm": 176.40713500976562, | |
| "learning_rate": 6.899999999999999e-06, | |
| "loss": 7.5271, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.002099875319902881, | |
| "grad_norm": 668.3442993164062, | |
| "learning_rate": 9.9e-06, | |
| "loss": 7.6155, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.002624844149878601, | |
| "grad_norm": 171.85369873046875, | |
| "learning_rate": 1.2899999999999998e-05, | |
| "loss": 7.4873, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.003149812979854321, | |
| "grad_norm": 477.1031494140625, | |
| "learning_rate": 1.5899999999999997e-05, | |
| "loss": 7.4295, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0036747818098300414, | |
| "grad_norm": 256.7841491699219, | |
| "learning_rate": 1.89e-05, | |
| "loss": 7.3037, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.004199750639805762, | |
| "grad_norm": 240.86167907714844, | |
| "learning_rate": 2.1899999999999997e-05, | |
| "loss": 7.3488, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0047247194697814816, | |
| "grad_norm": 151.8752899169922, | |
| "learning_rate": 2.49e-05, | |
| "loss": 7.1664, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.005249688299757202, | |
| "grad_norm": 92.41702270507812, | |
| "learning_rate": 2.7899999999999997e-05, | |
| "loss": 7.2001, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.005774657129732922, | |
| "grad_norm": 125.6628189086914, | |
| "learning_rate": 3.09e-05, | |
| "loss": 7.1956, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.006299625959708642, | |
| "grad_norm": 105.1700210571289, | |
| "learning_rate": 3.39e-05, | |
| "loss": 7.2136, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.006824594789684363, | |
| "grad_norm": 95.49678039550781, | |
| "learning_rate": 3.6899999999999996e-05, | |
| "loss": 7.2046, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.007349563619660083, | |
| "grad_norm": 70.33267974853516, | |
| "learning_rate": 3.99e-05, | |
| "loss": 7.0895, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.007874532449635803, | |
| "grad_norm": 42.994972229003906, | |
| "learning_rate": 4.289999999999999e-05, | |
| "loss": 7.1157, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.008399501279611523, | |
| "grad_norm": 87.23907470703125, | |
| "learning_rate": 4.59e-05, | |
| "loss": 7.1537, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.008924470109587244, | |
| "grad_norm": 53.14045333862305, | |
| "learning_rate": 4.8899999999999996e-05, | |
| "loss": 7.0126, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.009449438939562963, | |
| "grad_norm": 44.222225189208984, | |
| "learning_rate": 5.1899999999999994e-05, | |
| "loss": 7.1067, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.009974407769538684, | |
| "grad_norm": 65.67877197265625, | |
| "learning_rate": 5.489999999999999e-05, | |
| "loss": 7.0359, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.010499376599514405, | |
| "grad_norm": 34.80784225463867, | |
| "learning_rate": 5.79e-05, | |
| "loss": 7.0124, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.011024345429490124, | |
| "grad_norm": 80.05706024169922, | |
| "learning_rate": 6.0899999999999996e-05, | |
| "loss": 6.9882, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.011549314259465844, | |
| "grad_norm": 18.072044372558594, | |
| "learning_rate": 6.39e-05, | |
| "loss": 7.082, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.012074283089441565, | |
| "grad_norm": 27.460533142089844, | |
| "learning_rate": 6.69e-05, | |
| "loss": 7.0056, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.012599251919417284, | |
| "grad_norm": 20.942256927490234, | |
| "learning_rate": 6.989999999999999e-05, | |
| "loss": 7.0385, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.013124220749393005, | |
| "grad_norm": 21.205760955810547, | |
| "learning_rate": 7.29e-05, | |
| "loss": 6.9065, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.013649189579368726, | |
| "grad_norm": 30.357505798339844, | |
| "learning_rate": 7.589999999999999e-05, | |
| "loss": 6.9887, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.014174158409344445, | |
| "grad_norm": 21.6142520904541, | |
| "learning_rate": 7.89e-05, | |
| "loss": 7.0671, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.014699127239320165, | |
| "grad_norm": 34.31865310668945, | |
| "learning_rate": 8.19e-05, | |
| "loss": 6.9352, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.015224096069295886, | |
| "grad_norm": 31.643892288208008, | |
| "learning_rate": 8.489999999999999e-05, | |
| "loss": 7.0615, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.015749064899271605, | |
| "grad_norm": 21.246601104736328, | |
| "learning_rate": 8.789999999999998e-05, | |
| "loss": 6.8747, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.016274033729247326, | |
| "grad_norm": 17.99869155883789, | |
| "learning_rate": 9.089999999999999e-05, | |
| "loss": 6.9325, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.016799002559223047, | |
| "grad_norm": 10.438518524169922, | |
| "learning_rate": 9.389999999999999e-05, | |
| "loss": 6.975, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.017323971389198767, | |
| "grad_norm": 38.324283599853516, | |
| "learning_rate": 9.69e-05, | |
| "loss": 6.8833, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.017848940219174488, | |
| "grad_norm": 25.789045333862305, | |
| "learning_rate": 9.99e-05, | |
| "loss": 6.869, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.018373909049150206, | |
| "grad_norm": 26.649545669555664, | |
| "learning_rate": 0.0001029, | |
| "loss": 6.8773, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.018898877879125926, | |
| "grad_norm": 24.452266693115234, | |
| "learning_rate": 0.00010589999999999999, | |
| "loss": 6.7558, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.019423846709101647, | |
| "grad_norm": 14.234713554382324, | |
| "learning_rate": 0.00010889999999999999, | |
| "loss": 6.9624, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.019948815539077368, | |
| "grad_norm": 16.222957611083984, | |
| "learning_rate": 0.0001119, | |
| "loss": 6.9116, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.02047378436905309, | |
| "grad_norm": 14.955327033996582, | |
| "learning_rate": 0.00011489999999999999, | |
| "loss": 6.8227, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.02099875319902881, | |
| "grad_norm": 13.272610664367676, | |
| "learning_rate": 0.00011789999999999999, | |
| "loss": 6.8292, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.021523722029004527, | |
| "grad_norm": 14.966605186462402, | |
| "learning_rate": 0.0001209, | |
| "loss": 6.8364, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.022048690858980247, | |
| "grad_norm": 11.822840690612793, | |
| "learning_rate": 0.00012389999999999998, | |
| "loss": 6.668, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.022573659688955968, | |
| "grad_norm": 12.50557804107666, | |
| "learning_rate": 0.0001269, | |
| "loss": 6.6604, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.02309862851893169, | |
| "grad_norm": 13.769299507141113, | |
| "learning_rate": 0.00012989999999999999, | |
| "loss": 6.6969, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.02362359734890741, | |
| "grad_norm": 9.62465763092041, | |
| "learning_rate": 0.00013289999999999998, | |
| "loss": 6.7391, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.02414856617888313, | |
| "grad_norm": 8.241267204284668, | |
| "learning_rate": 0.0001359, | |
| "loss": 6.7064, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.024673535008858848, | |
| "grad_norm": 14.394071578979492, | |
| "learning_rate": 0.0001389, | |
| "loss": 6.7164, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.02519850383883457, | |
| "grad_norm": 8.198172569274902, | |
| "learning_rate": 0.00014189999999999998, | |
| "loss": 6.7228, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.02572347266881029, | |
| "grad_norm": 17.51156997680664, | |
| "learning_rate": 0.00014489999999999997, | |
| "loss": 6.6933, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.02624844149878601, | |
| "grad_norm": 14.444514274597168, | |
| "learning_rate": 0.0001479, | |
| "loss": 6.5789, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.02677341032876173, | |
| "grad_norm": 8.710993766784668, | |
| "learning_rate": 0.00015089999999999998, | |
| "loss": 6.6053, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.02729837915873745, | |
| "grad_norm": 11.155731201171875, | |
| "learning_rate": 0.0001539, | |
| "loss": 6.6003, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.02782334798871317, | |
| "grad_norm": 9.0637845993042, | |
| "learning_rate": 0.0001569, | |
| "loss": 6.5253, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.02834831681868889, | |
| "grad_norm": 7.813695907592773, | |
| "learning_rate": 0.00015989999999999998, | |
| "loss": 6.5777, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.02887328564866461, | |
| "grad_norm": 9.640409469604492, | |
| "learning_rate": 0.0001629, | |
| "loss": 6.5582, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.02939825447864033, | |
| "grad_norm": 10.287951469421387, | |
| "learning_rate": 0.0001659, | |
| "loss": 6.5429, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.02992322330861605, | |
| "grad_norm": 8.388657569885254, | |
| "learning_rate": 0.00016889999999999996, | |
| "loss": 6.4797, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.030448192138591772, | |
| "grad_norm": 6.651219844818115, | |
| "learning_rate": 0.00017189999999999998, | |
| "loss": 6.472, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.03097316096856749, | |
| "grad_norm": 6.32115364074707, | |
| "learning_rate": 0.00017489999999999997, | |
| "loss": 6.4754, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.03149812979854321, | |
| "grad_norm": 10.339417457580566, | |
| "learning_rate": 0.0001779, | |
| "loss": 6.4067, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.03202309862851893, | |
| "grad_norm": 7.705801963806152, | |
| "learning_rate": 0.00018089999999999998, | |
| "loss": 6.4035, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.03254806745849465, | |
| "grad_norm": 6.596720218658447, | |
| "learning_rate": 0.00018389999999999997, | |
| "loss": 6.3301, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.03307303628847037, | |
| "grad_norm": 10.351055145263672, | |
| "learning_rate": 0.0001869, | |
| "loss": 6.3362, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.03359800511844609, | |
| "grad_norm": 8.058084487915039, | |
| "learning_rate": 0.00018989999999999998, | |
| "loss": 6.4833, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.034122973948421814, | |
| "grad_norm": 8.515560150146484, | |
| "learning_rate": 0.0001929, | |
| "loss": 6.2454, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.034647942778397535, | |
| "grad_norm": 7.069362640380859, | |
| "learning_rate": 0.0001959, | |
| "loss": 6.2539, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.035172911608373256, | |
| "grad_norm": 6.835115909576416, | |
| "learning_rate": 0.00019889999999999998, | |
| "loss": 6.2121, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.035697880438348976, | |
| "grad_norm": 9.131962776184082, | |
| "learning_rate": 0.0002019, | |
| "loss": 6.1163, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.03622284926832469, | |
| "grad_norm": 6.500721454620361, | |
| "learning_rate": 0.0002049, | |
| "loss": 6.2233, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.03674781809830041, | |
| "grad_norm": 6.955876350402832, | |
| "learning_rate": 0.00020789999999999996, | |
| "loss": 6.2746, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.03727278692827613, | |
| "grad_norm": 12.63397216796875, | |
| "learning_rate": 0.00021089999999999998, | |
| "loss": 6.2639, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.03779775575825185, | |
| "grad_norm": 6.766085624694824, | |
| "learning_rate": 0.00021389999999999997, | |
| "loss": 6.2185, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.03832272458822757, | |
| "grad_norm": 7.125621795654297, | |
| "learning_rate": 0.0002169, | |
| "loss": 6.2077, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.038847693418203294, | |
| "grad_norm": 6.098615646362305, | |
| "learning_rate": 0.00021989999999999998, | |
| "loss": 6.1988, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.039372662248179015, | |
| "grad_norm": 6.724160671234131, | |
| "learning_rate": 0.00022289999999999997, | |
| "loss": 6.1645, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.039897631078154736, | |
| "grad_norm": 6.864727973937988, | |
| "learning_rate": 0.0002259, | |
| "loss": 5.9853, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.040422599908130456, | |
| "grad_norm": 7.888518810272217, | |
| "learning_rate": 0.00022889999999999998, | |
| "loss": 6.0459, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.04094756873810618, | |
| "grad_norm": 5.348147392272949, | |
| "learning_rate": 0.0002319, | |
| "loss": 6.0306, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.0414725375680819, | |
| "grad_norm": 7.666118621826172, | |
| "learning_rate": 0.0002349, | |
| "loss": 6.0464, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.04199750639805762, | |
| "grad_norm": 12.02874755859375, | |
| "learning_rate": 0.00023789999999999998, | |
| "loss": 5.9991, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.04252247522803333, | |
| "grad_norm": 6.495911598205566, | |
| "learning_rate": 0.0002409, | |
| "loss": 5.9451, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.04304744405800905, | |
| "grad_norm": 9.461275100708008, | |
| "learning_rate": 0.00024389999999999997, | |
| "loss": 5.9558, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.043572412887984774, | |
| "grad_norm": 6.616198539733887, | |
| "learning_rate": 0.0002469, | |
| "loss": 5.9853, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.044097381717960495, | |
| "grad_norm": 6.1858086585998535, | |
| "learning_rate": 0.00024989999999999995, | |
| "loss": 5.9692, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.044622350547936215, | |
| "grad_norm": 5.398629188537598, | |
| "learning_rate": 0.00025289999999999997, | |
| "loss": 5.9193, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.045147319377911936, | |
| "grad_norm": 7.980245113372803, | |
| "learning_rate": 0.0002559, | |
| "loss": 5.8701, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.04567228820788766, | |
| "grad_norm": 6.517972946166992, | |
| "learning_rate": 0.00025889999999999995, | |
| "loss": 5.905, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.04619725703786338, | |
| "grad_norm": 6.580096244812012, | |
| "learning_rate": 0.00026189999999999997, | |
| "loss": 5.7834, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.0467222258678391, | |
| "grad_norm": 6.371602535247803, | |
| "learning_rate": 0.0002649, | |
| "loss": 5.8452, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.04724719469781482, | |
| "grad_norm": 6.168027400970459, | |
| "learning_rate": 0.0002679, | |
| "loss": 5.6781, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.04777216352779054, | |
| "grad_norm": 6.000701427459717, | |
| "learning_rate": 0.00027089999999999997, | |
| "loss": 5.6675, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.04829713235776626, | |
| "grad_norm": 6.960968971252441, | |
| "learning_rate": 0.0002739, | |
| "loss": 5.5692, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.048822101187741974, | |
| "grad_norm": 10.075114250183105, | |
| "learning_rate": 0.0002769, | |
| "loss": 5.7553, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.049347070017717695, | |
| "grad_norm": 5.733921527862549, | |
| "learning_rate": 0.0002799, | |
| "loss": 5.5125, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.049872038847693416, | |
| "grad_norm": 7.427346706390381, | |
| "learning_rate": 0.00028289999999999994, | |
| "loss": 5.6595, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.05039700767766914, | |
| "grad_norm": 5.955673694610596, | |
| "learning_rate": 0.00028589999999999996, | |
| "loss": 5.5868, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.05092197650764486, | |
| "grad_norm": 7.7476420402526855, | |
| "learning_rate": 0.0002889, | |
| "loss": 5.632, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.05144694533762058, | |
| "grad_norm": 6.524439811706543, | |
| "learning_rate": 0.0002919, | |
| "loss": 5.6333, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.0519719141675963, | |
| "grad_norm": 6.380852699279785, | |
| "learning_rate": 0.00029489999999999996, | |
| "loss": 5.6931, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.05249688299757202, | |
| "grad_norm": 6.871588706970215, | |
| "learning_rate": 0.0002979, | |
| "loss": 5.4705, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.05302185182754774, | |
| "grad_norm": 5.873289585113525, | |
| "learning_rate": 0.00029999999999993335, | |
| "loss": 5.659, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.05354682065752346, | |
| "grad_norm": 6.722020626068115, | |
| "learning_rate": 0.00029999999999874875, | |
| "loss": 5.5263, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.05407178948749918, | |
| "grad_norm": 7.375940799713135, | |
| "learning_rate": 0.00029999999999608346, | |
| "loss": 5.4849, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.0545967583174749, | |
| "grad_norm": 6.465951919555664, | |
| "learning_rate": 0.00029999999999193736, | |
| "loss": 5.4179, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.05512172714745062, | |
| "grad_norm": 6.1253156661987305, | |
| "learning_rate": 0.00029999999998631057, | |
| "loss": 5.415, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.05564669597742634, | |
| "grad_norm": 6.320364475250244, | |
| "learning_rate": 0.000299999999979203, | |
| "loss": 5.2886, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.05617166480740206, | |
| "grad_norm": 5.7049880027771, | |
| "learning_rate": 0.00029999999997061473, | |
| "loss": 5.3008, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.05669663363737778, | |
| "grad_norm": 7.153939247131348, | |
| "learning_rate": 0.00029999999996054575, | |
| "loss": 5.336, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.0572216024673535, | |
| "grad_norm": 6.331596374511719, | |
| "learning_rate": 0.00029999999994899596, | |
| "loss": 5.2242, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.05774657129732922, | |
| "grad_norm": 6.270790100097656, | |
| "learning_rate": 0.00029999999993596553, | |
| "loss": 5.2873, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.05827154012730494, | |
| "grad_norm": 6.87438440322876, | |
| "learning_rate": 0.0002999999999214543, | |
| "loss": 5.2111, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.05879650895728066, | |
| "grad_norm": 6.366969585418701, | |
| "learning_rate": 0.0002999999999054623, | |
| "loss": 5.1685, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.05932147778725638, | |
| "grad_norm": 6.2817816734313965, | |
| "learning_rate": 0.00029999999988798964, | |
| "loss": 5.3457, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.0598464466172321, | |
| "grad_norm": 5.789464950561523, | |
| "learning_rate": 0.0002999999998690362, | |
| "loss": 5.0926, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.060371415447207824, | |
| "grad_norm": 6.459042072296143, | |
| "learning_rate": 0.00029999999984860204, | |
| "loss": 5.0385, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.060896384277183545, | |
| "grad_norm": 6.7204413414001465, | |
| "learning_rate": 0.00029999999982668717, | |
| "loss": 5.2199, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.061421353107159266, | |
| "grad_norm": 6.44543981552124, | |
| "learning_rate": 0.0002999999998032915, | |
| "loss": 5.1449, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.06194632193713498, | |
| "grad_norm": 5.885377407073975, | |
| "learning_rate": 0.0002999999997784151, | |
| "loss": 5.1175, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.0624712907671107, | |
| "grad_norm": 6.293627738952637, | |
| "learning_rate": 0.000299999999752058, | |
| "loss": 5.0933, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.06299625959708642, | |
| "grad_norm": 6.686017990112305, | |
| "learning_rate": 0.0002999999997242202, | |
| "loss": 4.9613, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.06352122842706215, | |
| "grad_norm": 5.605543613433838, | |
| "learning_rate": 0.0002999999996949016, | |
| "loss": 5.0621, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.06404619725703786, | |
| "grad_norm": 6.000121593475342, | |
| "learning_rate": 0.0002999999996641022, | |
| "loss": 5.0555, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.06457116608701359, | |
| "grad_norm": 7.083418369293213, | |
| "learning_rate": 0.00029999999963182215, | |
| "loss": 5.0052, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.0650961349169893, | |
| "grad_norm": 6.544703483581543, | |
| "learning_rate": 0.00029999999959806135, | |
| "loss": 5.0612, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.06562110374696502, | |
| "grad_norm": 6.990364074707031, | |
| "learning_rate": 0.0002999999995628198, | |
| "loss": 5.0725, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.06614607257694075, | |
| "grad_norm": 6.532026290893555, | |
| "learning_rate": 0.0002999999995260976, | |
| "loss": 5.0609, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.06667104140691646, | |
| "grad_norm": 6.3640522956848145, | |
| "learning_rate": 0.00029999999948789453, | |
| "loss": 4.9563, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.06719601023689219, | |
| "grad_norm": 7.006086349487305, | |
| "learning_rate": 0.00029999999944821084, | |
| "loss": 4.9507, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.0677209790668679, | |
| "grad_norm": 6.476901531219482, | |
| "learning_rate": 0.00029999999940704634, | |
| "loss": 4.9325, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.06824594789684363, | |
| "grad_norm": 6.32811975479126, | |
| "learning_rate": 0.00029999999936440114, | |
| "loss": 4.819, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.06877091672681934, | |
| "grad_norm": 5.81315279006958, | |
| "learning_rate": 0.00029999999932027514, | |
| "loss": 4.8838, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.06929588555679507, | |
| "grad_norm": 6.884243488311768, | |
| "learning_rate": 0.00029999999927466845, | |
| "loss": 4.9064, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.06982085438677078, | |
| "grad_norm": 5.8015313148498535, | |
| "learning_rate": 0.00029999999922758106, | |
| "loss": 4.8818, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.07034582321674651, | |
| "grad_norm": 6.7853875160217285, | |
| "learning_rate": 0.00029999999917901287, | |
| "loss": 4.7985, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.07087079204672223, | |
| "grad_norm": 6.297364711761475, | |
| "learning_rate": 0.000299999999128964, | |
| "loss": 4.8261, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.07139576087669795, | |
| "grad_norm": 7.7735514640808105, | |
| "learning_rate": 0.0002999999990774343, | |
| "loss": 4.8545, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.07192072970667367, | |
| "grad_norm": 6.501247406005859, | |
| "learning_rate": 0.00029999999902442396, | |
| "loss": 4.7666, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.07244569853664938, | |
| "grad_norm": 5.721576690673828, | |
| "learning_rate": 0.00029999999896993283, | |
| "loss": 4.6484, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.07297066736662511, | |
| "grad_norm": 6.3349127769470215, | |
| "learning_rate": 0.000299999998913961, | |
| "loss": 4.6921, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.07349563619660082, | |
| "grad_norm": 5.776736259460449, | |
| "learning_rate": 0.0002999999988565084, | |
| "loss": 4.6327, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.07402060502657655, | |
| "grad_norm": 8.254961013793945, | |
| "learning_rate": 0.0002999999987975751, | |
| "loss": 4.7353, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.07454557385655226, | |
| "grad_norm": 7.920860767364502, | |
| "learning_rate": 0.000299999998737161, | |
| "loss": 4.6561, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.07507054268652799, | |
| "grad_norm": 6.328056812286377, | |
| "learning_rate": 0.0002999999986752662, | |
| "loss": 4.556, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.0755955115165037, | |
| "grad_norm": 6.354831695556641, | |
| "learning_rate": 0.0002999999986118907, | |
| "loss": 4.5249, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.07612048034647943, | |
| "grad_norm": 5.605742931365967, | |
| "learning_rate": 0.0002999999985470344, | |
| "loss": 4.5733, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.07664544917645515, | |
| "grad_norm": 7.2531514167785645, | |
| "learning_rate": 0.0002999999984806974, | |
| "loss": 4.6005, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.07717041800643087, | |
| "grad_norm": 6.168049335479736, | |
| "learning_rate": 0.0002999999984128797, | |
| "loss": 4.4608, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.07769538683640659, | |
| "grad_norm": 5.962798595428467, | |
| "learning_rate": 0.00029999999834358116, | |
| "loss": 4.4705, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.0782203556663823, | |
| "grad_norm": 7.6847357749938965, | |
| "learning_rate": 0.00029999999827280195, | |
| "loss": 4.5147, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.07874532449635803, | |
| "grad_norm": 8.098923683166504, | |
| "learning_rate": 0.000299999998200542, | |
| "loss": 4.5436, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.07927029332633374, | |
| "grad_norm": 6.158365249633789, | |
| "learning_rate": 0.00029999999812680133, | |
| "loss": 4.4493, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.07979526215630947, | |
| "grad_norm": 6.325509548187256, | |
| "learning_rate": 0.00029999999805157987, | |
| "loss": 4.4293, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.08032023098628518, | |
| "grad_norm": 6.080830097198486, | |
| "learning_rate": 0.0002999999979748777, | |
| "loss": 4.3919, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.08084519981626091, | |
| "grad_norm": 7.090545654296875, | |
| "learning_rate": 0.0002999999978966948, | |
| "loss": 4.4191, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.08137016864623663, | |
| "grad_norm": 6.284291744232178, | |
| "learning_rate": 0.00029999999781703116, | |
| "loss": 4.3456, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.08189513747621235, | |
| "grad_norm": 6.383409023284912, | |
| "learning_rate": 0.0002999999977358868, | |
| "loss": 4.3264, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.08242010630618807, | |
| "grad_norm": 6.384091854095459, | |
| "learning_rate": 0.00029999999765326166, | |
| "loss": 4.265, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.0829450751361638, | |
| "grad_norm": 6.6042022705078125, | |
| "learning_rate": 0.0002999999975691559, | |
| "loss": 4.4527, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.08347004396613951, | |
| "grad_norm": 5.603631496429443, | |
| "learning_rate": 0.0002999999974835692, | |
| "loss": 4.3473, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.08399501279611524, | |
| "grad_norm": 5.590120315551758, | |
| "learning_rate": 0.00029999999739650194, | |
| "loss": 4.2386, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.08451998162609095, | |
| "grad_norm": 6.289425849914551, | |
| "learning_rate": 0.00029999999730795384, | |
| "loss": 4.2665, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.08504495045606666, | |
| "grad_norm": 6.349957466125488, | |
| "learning_rate": 0.00029999999721792506, | |
| "loss": 4.2492, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.08556991928604239, | |
| "grad_norm": 6.4022722244262695, | |
| "learning_rate": 0.0002999999971264155, | |
| "loss": 4.2056, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.0860948881160181, | |
| "grad_norm": 7.056876182556152, | |
| "learning_rate": 0.00029999999703342523, | |
| "loss": 4.1929, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.08661985694599383, | |
| "grad_norm": 6.141146659851074, | |
| "learning_rate": 0.00029999999693895425, | |
| "loss": 4.2077, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.08714482577596955, | |
| "grad_norm": 6.444318771362305, | |
| "learning_rate": 0.0002999999968430025, | |
| "loss": 4.1848, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.08766979460594528, | |
| "grad_norm": 6.252465724945068, | |
| "learning_rate": 0.00029999999674557005, | |
| "loss": 4.1801, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.08819476343592099, | |
| "grad_norm": 6.427366256713867, | |
| "learning_rate": 0.0002999999966466568, | |
| "loss": 4.2509, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.08871973226589672, | |
| "grad_norm": 6.234042167663574, | |
| "learning_rate": 0.0002999999965462629, | |
| "loss": 4.1133, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.08924470109587243, | |
| "grad_norm": 6.525606155395508, | |
| "learning_rate": 0.00029999999644438817, | |
| "loss": 4.019, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.0005249688299757202, | |
| "grad_norm": 5.570987701416016, | |
| "learning_rate": 0.00029999999634103275, | |
| "loss": 4.1462, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.0010499376599514404, | |
| "grad_norm": 5.909586429595947, | |
| "learning_rate": 0.0002999999962361966, | |
| "loss": 4.126, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.0015749064899271605, | |
| "grad_norm": 7.547698020935059, | |
| "learning_rate": 0.00029999999612987966, | |
| "loss": 4.0706, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.002099875319902881, | |
| "grad_norm": 6.932499408721924, | |
| "learning_rate": 0.00029999999602208204, | |
| "loss": 4.1576, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.002624844149878601, | |
| "grad_norm": 7.4274797439575195, | |
| "learning_rate": 0.0002999999959128037, | |
| "loss": 4.1343, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.003149812979854321, | |
| "grad_norm": 6.038876056671143, | |
| "learning_rate": 0.00029999999580204457, | |
| "loss": 4.1781, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.0036747818098300414, | |
| "grad_norm": 6.615757465362549, | |
| "learning_rate": 0.0002999999956898047, | |
| "loss": 4.0667, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.004199750639805762, | |
| "grad_norm": 6.53090238571167, | |
| "learning_rate": 0.00029999999557608415, | |
| "loss": 4.1289, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.0047247194697814816, | |
| "grad_norm": 5.512113571166992, | |
| "learning_rate": 0.0002999999954608828, | |
| "loss": 3.9358, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.005249688299757202, | |
| "grad_norm": 6.99248743057251, | |
| "learning_rate": 0.00029999999534420074, | |
| "loss": 3.9963, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.005774657129732922, | |
| "grad_norm": 5.911637306213379, | |
| "learning_rate": 0.00029999999522603793, | |
| "loss": 4.017, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.006299625959708642, | |
| "grad_norm": 5.9147562980651855, | |
| "learning_rate": 0.00029999999510639443, | |
| "loss": 4.0137, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.006824594789684363, | |
| "grad_norm": 5.845144271850586, | |
| "learning_rate": 0.00029999999498527013, | |
| "loss": 4.0678, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.007349563619660083, | |
| "grad_norm": 5.957810401916504, | |
| "learning_rate": 0.00029999999486266513, | |
| "loss": 3.891, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.007874532449635803, | |
| "grad_norm": 6.678396224975586, | |
| "learning_rate": 0.0002999999947385794, | |
| "loss": 3.9104, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.008399501279611523, | |
| "grad_norm": 5.920273303985596, | |
| "learning_rate": 0.0002999999946130129, | |
| "loss": 3.9749, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.008924470109587244, | |
| "grad_norm": 6.2641777992248535, | |
| "learning_rate": 0.0002999999944859657, | |
| "loss": 3.8296, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.009449438939562963, | |
| "grad_norm": 6.42849588394165, | |
| "learning_rate": 0.00029999999435743776, | |
| "loss": 3.9217, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.009974407769538684, | |
| "grad_norm": 5.971102714538574, | |
| "learning_rate": 0.0002999999942274291, | |
| "loss": 3.8751, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.010499376599514405, | |
| "grad_norm": 6.6937408447265625, | |
| "learning_rate": 0.00029999999409593964, | |
| "loss": 3.8184, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.011024345429490124, | |
| "grad_norm": 6.671769618988037, | |
| "learning_rate": 0.0002999999939629695, | |
| "loss": 3.7852, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.011549314259465844, | |
| "grad_norm": 6.817453861236572, | |
| "learning_rate": 0.00029999999382851857, | |
| "loss": 3.8804, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.012074283089441565, | |
| "grad_norm": 7.681510925292969, | |
| "learning_rate": 0.00029999999369258694, | |
| "loss": 3.8054, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.012599251919417284, | |
| "grad_norm": 6.021421432495117, | |
| "learning_rate": 0.00029999999355517456, | |
| "loss": 3.8191, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.013124220749393005, | |
| "grad_norm": 6.368685245513916, | |
| "learning_rate": 0.0002999999934162815, | |
| "loss": 3.671, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.013649189579368726, | |
| "grad_norm": 6.735239505767822, | |
| "learning_rate": 0.0002999999932759076, | |
| "loss": 3.8424, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.014174158409344445, | |
| "grad_norm": 6.516197681427002, | |
| "learning_rate": 0.00029999999313405304, | |
| "loss": 3.9186, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.014699127239320165, | |
| "grad_norm": 6.4768476486206055, | |
| "learning_rate": 0.0002999999929907177, | |
| "loss": 3.7679, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.015224096069295886, | |
| "grad_norm": 6.700578212738037, | |
| "learning_rate": 0.00029999999284590164, | |
| "loss": 3.904, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.015749064899271605, | |
| "grad_norm": 6.943190097808838, | |
| "learning_rate": 0.0002999999926996049, | |
| "loss": 3.7242, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.016274033729247326, | |
| "grad_norm": 5.892586708068848, | |
| "learning_rate": 0.0002999999925518273, | |
| "loss": 3.7425, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.016799002559223047, | |
| "grad_norm": 6.512500286102295, | |
| "learning_rate": 0.0002999999924025691, | |
| "loss": 3.797, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.017323971389198767, | |
| "grad_norm": 6.24969482421875, | |
| "learning_rate": 0.0002999999922518301, | |
| "loss": 3.7444, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.017848940219174488, | |
| "grad_norm": 6.845114707946777, | |
| "learning_rate": 0.0002999999920996103, | |
| "loss": 3.6969, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.018373909049150206, | |
| "grad_norm": 6.944314002990723, | |
| "learning_rate": 0.00029999999194590986, | |
| "loss": 3.731, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.018898877879125926, | |
| "grad_norm": 6.697751998901367, | |
| "learning_rate": 0.00029999999179072866, | |
| "loss": 3.5394, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.019423846709101647, | |
| "grad_norm": 6.382928848266602, | |
| "learning_rate": 0.0002999999916340667, | |
| "loss": 3.8336, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.019948815539077368, | |
| "grad_norm": 6.065791130065918, | |
| "learning_rate": 0.000299999991475924, | |
| "loss": 3.746, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.02047378436905309, | |
| "grad_norm": 6.519564151763916, | |
| "learning_rate": 0.0002999999913163006, | |
| "loss": 3.6768, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.02099875319902881, | |
| "grad_norm": 6.374921798706055, | |
| "learning_rate": 0.0002999999911551964, | |
| "loss": 3.6625, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.021523722029004527, | |
| "grad_norm": 6.67402982711792, | |
| "learning_rate": 0.00029999999099261155, | |
| "loss": 3.7065, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.022048690858980247, | |
| "grad_norm": 6.103598594665527, | |
| "learning_rate": 0.0002999999908285459, | |
| "loss": 3.5128, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.022573659688955968, | |
| "grad_norm": 6.75986909866333, | |
| "learning_rate": 0.0002999999906629995, | |
| "loss": 3.4612, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.02309862851893169, | |
| "grad_norm": 6.420792102813721, | |
| "learning_rate": 0.0002999999904959724, | |
| "loss": 3.5584, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.02362359734890741, | |
| "grad_norm": 8.478610038757324, | |
| "learning_rate": 0.0002999999903274646, | |
| "loss": 3.6476, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.02414856617888313, | |
| "grad_norm": 7.021547317504883, | |
| "learning_rate": 0.000299999990157476, | |
| "loss": 3.5927, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.024673535008858848, | |
| "grad_norm": 6.317281723022461, | |
| "learning_rate": 0.00029999998998600665, | |
| "loss": 3.6241, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.02519850383883457, | |
| "grad_norm": 5.901510715484619, | |
| "learning_rate": 0.0002999999898130566, | |
| "loss": 3.6491, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.02572347266881029, | |
| "grad_norm": 6.88697624206543, | |
| "learning_rate": 0.00029999998963862583, | |
| "loss": 3.619, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.02624844149878601, | |
| "grad_norm": 6.949872016906738, | |
| "learning_rate": 0.0002999999894627143, | |
| "loss": 3.4568, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.02677341032876173, | |
| "grad_norm": 6.634612560272217, | |
| "learning_rate": 0.000299999989285322, | |
| "loss": 3.5583, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.02729837915873745, | |
| "grad_norm": 8.479588508605957, | |
| "learning_rate": 0.00029999998910644905, | |
| "loss": 3.5644, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.02782334798871317, | |
| "grad_norm": 6.708542346954346, | |
| "learning_rate": 0.0002999999889260953, | |
| "loss": 3.4792, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.02834831681868889, | |
| "grad_norm": 7.055044174194336, | |
| "learning_rate": 0.00029999998874426085, | |
| "loss": 3.5505, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.02887328564866461, | |
| "grad_norm": 6.504602432250977, | |
| "learning_rate": 0.0002999999885609456, | |
| "loss": 3.5346, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.02939825447864033, | |
| "grad_norm": 6.789810657501221, | |
| "learning_rate": 0.0002999999883761497, | |
| "loss": 3.5312, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.02992322330861605, | |
| "grad_norm": 6.327883720397949, | |
| "learning_rate": 0.000299999988189873, | |
| "loss": 3.4453, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.030448192138591772, | |
| "grad_norm": 6.850128173828125, | |
| "learning_rate": 0.00029999998800211557, | |
| "loss": 3.4946, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.03097316096856749, | |
| "grad_norm": 6.40550422668457, | |
| "learning_rate": 0.0002999999878128774, | |
| "loss": 3.4903, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.03149812979854321, | |
| "grad_norm": 6.313216209411621, | |
| "learning_rate": 0.0002999999876221585, | |
| "loss": 3.4033, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.03202309862851893, | |
| "grad_norm": 7.214779376983643, | |
| "learning_rate": 0.0002999999874299589, | |
| "loss": 3.4198, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.03254806745849465, | |
| "grad_norm": 6.715667247772217, | |
| "learning_rate": 0.00029999998723627847, | |
| "loss": 3.3777, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.03307303628847037, | |
| "grad_norm": 5.828542709350586, | |
| "learning_rate": 0.0002999999870411174, | |
| "loss": 3.405, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.03359800511844609, | |
| "grad_norm": 7.116888999938965, | |
| "learning_rate": 0.00029999998684447556, | |
| "loss": 3.6236, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.034122973948421814, | |
| "grad_norm": 6.714303016662598, | |
| "learning_rate": 0.000299999986646353, | |
| "loss": 3.2953, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.034647942778397535, | |
| "grad_norm": 6.0912089347839355, | |
| "learning_rate": 0.0002999999864467497, | |
| "loss": 3.3727, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.035172911608373256, | |
| "grad_norm": 6.161454677581787, | |
| "learning_rate": 0.00029999998624566566, | |
| "loss": 3.2791, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.035697880438348976, | |
| "grad_norm": 6.854067802429199, | |
| "learning_rate": 0.00029999998604310086, | |
| "loss": 3.2219, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.03622284926832469, | |
| "grad_norm": 6.076297760009766, | |
| "learning_rate": 0.0002999999858390553, | |
| "loss": 3.3636, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.03674781809830041, | |
| "grad_norm": 6.223069190979004, | |
| "learning_rate": 0.00029999998563352907, | |
| "loss": 3.457, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.03727278692827613, | |
| "grad_norm": 6.591337203979492, | |
| "learning_rate": 0.0002999999854265221, | |
| "loss": 3.4884, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.03779775575825185, | |
| "grad_norm": 9.24142837524414, | |
| "learning_rate": 0.00029999998521803434, | |
| "loss": 3.419, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.03832272458822757, | |
| "grad_norm": 7.019595146179199, | |
| "learning_rate": 0.0002999999850080659, | |
| "loss": 3.427, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.038847693418203294, | |
| "grad_norm": 6.758765697479248, | |
| "learning_rate": 0.00029999998479661666, | |
| "loss": 3.4453, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.039372662248179015, | |
| "grad_norm": 7.150947570800781, | |
| "learning_rate": 0.00029999998458368673, | |
| "loss": 3.4291, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.039897631078154736, | |
| "grad_norm": 6.2172136306762695, | |
| "learning_rate": 0.00029999998436927604, | |
| "loss": 3.1965, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.040422599908130456, | |
| "grad_norm": 6.834853649139404, | |
| "learning_rate": 0.0002999999841533846, | |
| "loss": 3.3453, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.04094756873810618, | |
| "grad_norm": 7.9789137840271, | |
| "learning_rate": 0.00029999998393601243, | |
| "loss": 3.3349, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.0414725375680819, | |
| "grad_norm": 6.828987121582031, | |
| "learning_rate": 0.00029999998371715955, | |
| "loss": 3.393, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.04199750639805762, | |
| "grad_norm": 7.442930221557617, | |
| "learning_rate": 0.00029999998349682593, | |
| "loss": 3.3438, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.04252247522803333, | |
| "grad_norm": 7.22468376159668, | |
| "learning_rate": 0.00029999998327501155, | |
| "loss": 3.3022, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.04304744405800905, | |
| "grad_norm": 7.090549468994141, | |
| "learning_rate": 0.0002999999830517165, | |
| "loss": 3.3002, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.043572412887984774, | |
| "grad_norm": 7.341608047485352, | |
| "learning_rate": 0.00029999998282694067, | |
| "loss": 3.351, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.044097381717960495, | |
| "grad_norm": 6.907212257385254, | |
| "learning_rate": 0.00029999998260068404, | |
| "loss": 3.3846, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.044622350547936215, | |
| "grad_norm": 6.958861351013184, | |
| "learning_rate": 0.0002999999823729468, | |
| "loss": 3.3186, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.045147319377911936, | |
| "grad_norm": 7.10768461227417, | |
| "learning_rate": 0.0002999999821437287, | |
| "loss": 3.3192, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.04567228820788766, | |
| "grad_norm": 6.437366962432861, | |
| "learning_rate": 0.0002999999819130299, | |
| "loss": 3.3617, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.04619725703786338, | |
| "grad_norm": 7.021583557128906, | |
| "learning_rate": 0.0002999999816808504, | |
| "loss": 3.2424, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.0467222258678391, | |
| "grad_norm": 7.078067302703857, | |
| "learning_rate": 0.00029999998144719013, | |
| "loss": 3.3379, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.04724719469781482, | |
| "grad_norm": 6.58331823348999, | |
| "learning_rate": 0.0002999999812120492, | |
| "loss": 3.1476, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.04777216352779054, | |
| "grad_norm": 7.511198997497559, | |
| "learning_rate": 0.0002999999809754274, | |
| "loss": 3.1552, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.04829713235776626, | |
| "grad_norm": 6.889819622039795, | |
| "learning_rate": 0.00029999998073732497, | |
| "loss": 3.0854, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.048822101187741974, | |
| "grad_norm": 6.780309200286865, | |
| "learning_rate": 0.00029999998049774177, | |
| "loss": 3.3289, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.049347070017717695, | |
| "grad_norm": 6.212784290313721, | |
| "learning_rate": 0.0002999999802566778, | |
| "loss": 3.0708, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.049872038847693416, | |
| "grad_norm": 6.844520092010498, | |
| "learning_rate": 0.00029999998001413317, | |
| "loss": 3.2822, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.05039700767766914, | |
| "grad_norm": 6.3941521644592285, | |
| "learning_rate": 0.0002999999797701077, | |
| "loss": 3.2076, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.05092197650764486, | |
| "grad_norm": 6.843677997589111, | |
| "learning_rate": 0.0002999999795246016, | |
| "loss": 3.2694, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.05144694533762058, | |
| "grad_norm": 6.763852119445801, | |
| "learning_rate": 0.0002999999792776147, | |
| "loss": 3.2928, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.0519719141675963, | |
| "grad_norm": 7.747461795806885, | |
| "learning_rate": 0.00029999997902914704, | |
| "loss": 3.3835, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.05249688299757202, | |
| "grad_norm": 6.702919006347656, | |
| "learning_rate": 0.0002999999787791987, | |
| "loss": 3.1438, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.05302185182754774, | |
| "grad_norm": 7.283872604370117, | |
| "learning_rate": 0.0002999999785277696, | |
| "loss": 3.3725, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.05354682065752346, | |
| "grad_norm": 7.523728847503662, | |
| "learning_rate": 0.0002999999782748598, | |
| "loss": 3.2487, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.05407178948749918, | |
| "grad_norm": 7.731942653656006, | |
| "learning_rate": 0.00029999997802046926, | |
| "loss": 3.214, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.0545967583174749, | |
| "grad_norm": 7.324973106384277, | |
| "learning_rate": 0.0002999999777645979, | |
| "loss": 3.1839, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.05512172714745062, | |
| "grad_norm": 7.510282516479492, | |
| "learning_rate": 0.0002999999775072459, | |
| "loss": 3.2158, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.05564669597742634, | |
| "grad_norm": 6.5173211097717285, | |
| "learning_rate": 0.0002999999772484131, | |
| "loss": 3.0809, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.05617166480740206, | |
| "grad_norm": 6.481199741363525, | |
| "learning_rate": 0.0002999999769880996, | |
| "loss": 3.0955, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.05669663363737778, | |
| "grad_norm": 6.510835647583008, | |
| "learning_rate": 0.0002999999767263053, | |
| "loss": 3.1873, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.0572216024673535, | |
| "grad_norm": 7.291550159454346, | |
| "learning_rate": 0.00029999997646303035, | |
| "loss": 3.0941, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.05774657129732922, | |
| "grad_norm": 7.231327056884766, | |
| "learning_rate": 0.0002999999761982746, | |
| "loss": 3.1549, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.0005249688299757202, | |
| "grad_norm": 6.586388111114502, | |
| "learning_rate": 0.00029999997593203816, | |
| "loss": 3.161, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.0010499376599514404, | |
| "grad_norm": 5.895412445068359, | |
| "learning_rate": 0.000299999975664321, | |
| "loss": 3.1409, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.0015749064899271605, | |
| "grad_norm": 7.71649694442749, | |
| "learning_rate": 0.000299999975395123, | |
| "loss": 3.1125, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.002099875319902881, | |
| "grad_norm": 6.9331512451171875, | |
| "learning_rate": 0.00029999997512444437, | |
| "loss": 3.2067, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.002624844149878601, | |
| "grad_norm": 6.3300676345825195, | |
| "learning_rate": 0.00029999997485228495, | |
| "loss": 3.1829, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.003149812979854321, | |
| "grad_norm": 6.470039367675781, | |
| "learning_rate": 0.0002999999745786448, | |
| "loss": 3.2371, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.0036747818098300414, | |
| "grad_norm": 6.64292049407959, | |
| "learning_rate": 0.00029999997430352394, | |
| "loss": 3.153, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.004199750639805762, | |
| "grad_norm": 7.108335018157959, | |
| "learning_rate": 0.0002999999740269223, | |
| "loss": 3.2237, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.0047247194697814816, | |
| "grad_norm": 6.253315448760986, | |
| "learning_rate": 0.00029999997374883993, | |
| "loss": 3.0427, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.005249688299757202, | |
| "grad_norm": 8.03952407836914, | |
| "learning_rate": 0.0002999999734692769, | |
| "loss": 3.1039, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.005774657129732922, | |
| "grad_norm": 6.400339126586914, | |
| "learning_rate": 0.00029999997318823303, | |
| "loss": 3.1374, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.006299625959708642, | |
| "grad_norm": 7.349103927612305, | |
| "learning_rate": 0.0002999999729057085, | |
| "loss": 3.124, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.006824594789684363, | |
| "grad_norm": 6.341732025146484, | |
| "learning_rate": 0.0002999999726217032, | |
| "loss": 3.1996, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.007349563619660083, | |
| "grad_norm": 7.837357044219971, | |
| "learning_rate": 0.00029999997233621714, | |
| "loss": 3.0418, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.007874532449635803, | |
| "grad_norm": 7.084399700164795, | |
| "learning_rate": 0.00029999997204925035, | |
| "loss": 3.0537, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.008399501279611523, | |
| "grad_norm": 7.049639701843262, | |
| "learning_rate": 0.00029999997176080286, | |
| "loss": 3.1337, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.008924470109587244, | |
| "grad_norm": 7.964977264404297, | |
| "learning_rate": 0.0002999999714708746, | |
| "loss": 3.0249, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.009449438939562963, | |
| "grad_norm": 6.827041149139404, | |
| "learning_rate": 0.00029999997117946564, | |
| "loss": 3.0854, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.009974407769538684, | |
| "grad_norm": 6.784083366394043, | |
| "learning_rate": 0.0002999999708865759, | |
| "loss": 3.0744, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.010499376599514405, | |
| "grad_norm": 7.1898369789123535, | |
| "learning_rate": 0.0002999999705922054, | |
| "loss": 3.0265, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.011024345429490124, | |
| "grad_norm": 6.498958110809326, | |
| "learning_rate": 0.0002999999702963543, | |
| "loss": 2.9887, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 0.011549314259465844, | |
| "grad_norm": 8.748756408691406, | |
| "learning_rate": 0.00029999996999902236, | |
| "loss": 3.0802, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.012074283089441565, | |
| "grad_norm": 7.223897457122803, | |
| "learning_rate": 0.0002999999697002097, | |
| "loss": 3.034, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.012599251919417284, | |
| "grad_norm": 6.689151287078857, | |
| "learning_rate": 0.00029999996939991625, | |
| "loss": 3.04, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.013124220749393005, | |
| "grad_norm": 7.291615009307861, | |
| "learning_rate": 0.00029999996909814213, | |
| "loss": 2.9241, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.013649189579368726, | |
| "grad_norm": 7.096383571624756, | |
| "learning_rate": 0.00029999996879488726, | |
| "loss": 3.0883, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.014174158409344445, | |
| "grad_norm": 8.031325340270996, | |
| "learning_rate": 0.00029999996849015164, | |
| "loss": 3.1605, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 0.014699127239320165, | |
| "grad_norm": 6.762740135192871, | |
| "learning_rate": 0.00029999996818393527, | |
| "loss": 3.0279, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.015224096069295886, | |
| "grad_norm": 7.445662498474121, | |
| "learning_rate": 0.0002999999678762382, | |
| "loss": 3.1602, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.015749064899271605, | |
| "grad_norm": 7.1995697021484375, | |
| "learning_rate": 0.0002999999675670604, | |
| "loss": 3.0274, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.016274033729247326, | |
| "grad_norm": 6.132891654968262, | |
| "learning_rate": 0.00029999996725640183, | |
| "loss": 3.0188, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 0.016799002559223047, | |
| "grad_norm": 7.464890956878662, | |
| "learning_rate": 0.0002999999669442626, | |
| "loss": 3.0844, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.017323971389198767, | |
| "grad_norm": 6.418426513671875, | |
| "learning_rate": 0.0002999999666306425, | |
| "loss": 3.0466, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 0.017848940219174488, | |
| "grad_norm": 6.641352653503418, | |
| "learning_rate": 0.00029999996631554176, | |
| "loss": 3.0089, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.018373909049150206, | |
| "grad_norm": 7.779778003692627, | |
| "learning_rate": 0.00029999996599896025, | |
| "loss": 3.0565, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.018898877879125926, | |
| "grad_norm": 7.073680877685547, | |
| "learning_rate": 0.000299999965680898, | |
| "loss": 2.8763, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.019423846709101647, | |
| "grad_norm": 6.633608341217041, | |
| "learning_rate": 0.00029999996536135505, | |
| "loss": 3.147, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 0.019948815539077368, | |
| "grad_norm": 6.837402820587158, | |
| "learning_rate": 0.0002999999650403313, | |
| "loss": 3.0738, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.02047378436905309, | |
| "grad_norm": 7.128554344177246, | |
| "learning_rate": 0.00029999996471782685, | |
| "loss": 3.0307, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 0.02099875319902881, | |
| "grad_norm": 9.527042388916016, | |
| "learning_rate": 0.0002999999643938417, | |
| "loss": 3.0048, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.021523722029004527, | |
| "grad_norm": 6.489922523498535, | |
| "learning_rate": 0.00029999996406837576, | |
| "loss": 3.047, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.022048690858980247, | |
| "grad_norm": 6.500357151031494, | |
| "learning_rate": 0.0002999999637414291, | |
| "loss": 2.8873, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.022573659688955968, | |
| "grad_norm": 7.056972026824951, | |
| "learning_rate": 0.00029999996341300173, | |
| "loss": 2.8347, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 0.02309862851893169, | |
| "grad_norm": 6.988775730133057, | |
| "learning_rate": 0.0002999999630830936, | |
| "loss": 2.9194, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.02362359734890741, | |
| "grad_norm": 7.631206035614014, | |
| "learning_rate": 0.0002999999627517047, | |
| "loss": 3.0304, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.02414856617888313, | |
| "grad_norm": 8.399604797363281, | |
| "learning_rate": 0.0002999999624188351, | |
| "loss": 2.9789, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.024673535008858848, | |
| "grad_norm": 6.7663750648498535, | |
| "learning_rate": 0.0002999999620844848, | |
| "loss": 3.0153, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.02519850383883457, | |
| "grad_norm": 7.070738792419434, | |
| "learning_rate": 0.0002999999617486537, | |
| "loss": 3.0536, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.02572347266881029, | |
| "grad_norm": 6.313982963562012, | |
| "learning_rate": 0.0002999999614113419, | |
| "loss": 3.0068, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 0.02624844149878601, | |
| "grad_norm": 7.207414150238037, | |
| "learning_rate": 0.0002999999610725493, | |
| "loss": 2.8757, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.02677341032876173, | |
| "grad_norm": 6.756102085113525, | |
| "learning_rate": 0.00029999996073227603, | |
| "loss": 2.9751, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 0.02729837915873745, | |
| "grad_norm": 7.889659404754639, | |
| "learning_rate": 0.000299999960390522, | |
| "loss": 2.9983, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.02782334798871317, | |
| "grad_norm": 7.308917045593262, | |
| "learning_rate": 0.0002999999600472873, | |
| "loss": 2.9138, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 0.02834831681868889, | |
| "grad_norm": 7.27027702331543, | |
| "learning_rate": 0.00029999995970257176, | |
| "loss": 2.9816, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.02887328564866461, | |
| "grad_norm": 7.231856822967529, | |
| "learning_rate": 0.00029999995935637555, | |
| "loss": 2.9656, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.02939825447864033, | |
| "grad_norm": 6.888019561767578, | |
| "learning_rate": 0.00029999995900869853, | |
| "loss": 2.9676, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.02992322330861605, | |
| "grad_norm": 6.69609260559082, | |
| "learning_rate": 0.00029999995865954087, | |
| "loss": 2.8967, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 0.030448192138591772, | |
| "grad_norm": 7.686691761016846, | |
| "learning_rate": 0.00029999995830890246, | |
| "loss": 2.9528, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.03097316096856749, | |
| "grad_norm": 6.648905277252197, | |
| "learning_rate": 0.00029999995795678325, | |
| "loss": 2.9439, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 0.03149812979854321, | |
| "grad_norm": 7.733336925506592, | |
| "learning_rate": 0.00029999995760318334, | |
| "loss": 2.8676, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.03202309862851893, | |
| "grad_norm": 7.063968181610107, | |
| "learning_rate": 0.0002999999572481027, | |
| "loss": 2.8855, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 0.03254806745849465, | |
| "grad_norm": 6.942965984344482, | |
| "learning_rate": 0.00029999995689154133, | |
| "loss": 2.8601, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.03307303628847037, | |
| "grad_norm": 6.782042980194092, | |
| "learning_rate": 0.0002999999565334992, | |
| "loss": 2.8796, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 0.03359800511844609, | |
| "grad_norm": 7.013416290283203, | |
| "learning_rate": 0.0002999999561739763, | |
| "loss": 3.0813, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.034122973948421814, | |
| "grad_norm": 6.523316383361816, | |
| "learning_rate": 0.0002999999558129727, | |
| "loss": 2.7886, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.034647942778397535, | |
| "grad_norm": 7.136838436126709, | |
| "learning_rate": 0.0002999999554504884, | |
| "loss": 2.8674, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.035172911608373256, | |
| "grad_norm": 6.920617580413818, | |
| "learning_rate": 0.00029999995508652334, | |
| "loss": 2.778, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 0.035697880438348976, | |
| "grad_norm": 7.3091912269592285, | |
| "learning_rate": 0.00029999995472107754, | |
| "loss": 2.7434, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 0.03622284926832469, | |
| "grad_norm": 6.648264408111572, | |
| "learning_rate": 0.000299999954354151, | |
| "loss": 2.8715, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 0.03674781809830041, | |
| "grad_norm": 6.835089683532715, | |
| "learning_rate": 0.0002999999539857437, | |
| "loss": 2.9585, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.03727278692827613, | |
| "grad_norm": 6.782878875732422, | |
| "learning_rate": 0.0002999999536158557, | |
| "loss": 2.9946, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 0.03779775575825185, | |
| "grad_norm": 7.893862724304199, | |
| "learning_rate": 0.00029999995324448694, | |
| "loss": 2.9384, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.03832272458822757, | |
| "grad_norm": 7.523046016693115, | |
| "learning_rate": 0.00029999995287163745, | |
| "loss": 2.9467, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 0.038847693418203294, | |
| "grad_norm": 7.165585517883301, | |
| "learning_rate": 0.0002999999524973073, | |
| "loss": 2.9622, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 0.039372662248179015, | |
| "grad_norm": 7.151083469390869, | |
| "learning_rate": 0.0002999999521214963, | |
| "loss": 2.9499, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.039897631078154736, | |
| "grad_norm": 6.360107898712158, | |
| "learning_rate": 0.0002999999517442046, | |
| "loss": 2.7351, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 0.040422599908130456, | |
| "grad_norm": 7.411434173583984, | |
| "learning_rate": 0.0002999999513654322, | |
| "loss": 2.8847, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 0.04094756873810618, | |
| "grad_norm": 9.267210960388184, | |
| "learning_rate": 0.00029999995098517896, | |
| "loss": 2.8775, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 0.0414725375680819, | |
| "grad_norm": 7.2229084968566895, | |
| "learning_rate": 0.0002999999506034451, | |
| "loss": 2.9468, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 0.04199750639805762, | |
| "grad_norm": 8.308311462402344, | |
| "learning_rate": 0.0002999999502202305, | |
| "loss": 2.8883, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.04252247522803333, | |
| "grad_norm": 7.181619167327881, | |
| "learning_rate": 0.00029999994983553505, | |
| "loss": 2.856, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 0.04304744405800905, | |
| "grad_norm": 8.149490356445312, | |
| "learning_rate": 0.000299999949449359, | |
| "loss": 2.8516, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 0.043572412887984774, | |
| "grad_norm": 7.3903679847717285, | |
| "learning_rate": 0.00029999994910053445, | |
| "loss": 2.9047, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 0.044097381717960495, | |
| "grad_norm": 6.560291767120361, | |
| "learning_rate": 0.00029999994871154494, | |
| "loss": 2.9442, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 0.044622350547936215, | |
| "grad_norm": 6.807400703430176, | |
| "learning_rate": 0.0002999999483210747, | |
| "loss": 2.8794, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.045147319377911936, | |
| "grad_norm": 10.622593879699707, | |
| "learning_rate": 0.0002999999479291237, | |
| "loss": 2.8952, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 0.04567228820788766, | |
| "grad_norm": 6.817701816558838, | |
| "learning_rate": 0.00029999994753569196, | |
| "loss": 2.9354, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 0.04619725703786338, | |
| "grad_norm": 7.763261318206787, | |
| "learning_rate": 0.0002999999471407795, | |
| "loss": 2.8219, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.0467222258678391, | |
| "grad_norm": 7.670142650604248, | |
| "learning_rate": 0.0002999999467443863, | |
| "loss": 2.9176, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 0.04724719469781482, | |
| "grad_norm": 6.744191646575928, | |
| "learning_rate": 0.0002999999463465124, | |
| "loss": 2.7404, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.04777216352779054, | |
| "grad_norm": 7.410317420959473, | |
| "learning_rate": 0.0002999999459471577, | |
| "loss": 2.7413, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 0.04829713235776626, | |
| "grad_norm": 6.410220146179199, | |
| "learning_rate": 0.00029999994554632235, | |
| "loss": 2.6918, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 0.048822101187741974, | |
| "grad_norm": 6.463805675506592, | |
| "learning_rate": 0.0002999999451440062, | |
| "loss": 2.919, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 0.049347070017717695, | |
| "grad_norm": 6.245724678039551, | |
| "learning_rate": 0.0002999999447402093, | |
| "loss": 2.6892, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 0.049872038847693416, | |
| "grad_norm": 7.3049845695495605, | |
| "learning_rate": 0.00029999994433493166, | |
| "loss": 2.8827, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.05039700767766914, | |
| "grad_norm": 6.892119884490967, | |
| "learning_rate": 0.0002999999439281734, | |
| "loss": 2.8214, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 0.05092197650764486, | |
| "grad_norm": 6.675173282623291, | |
| "learning_rate": 0.0002999999435199343, | |
| "loss": 2.8678, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 0.05144694533762058, | |
| "grad_norm": 7.2190351486206055, | |
| "learning_rate": 0.00029999994311021445, | |
| "loss": 2.8919, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 0.0519719141675963, | |
| "grad_norm": 8.603300094604492, | |
| "learning_rate": 0.00029999994269901386, | |
| "loss": 2.9827, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 0.05249688299757202, | |
| "grad_norm": 7.088057994842529, | |
| "learning_rate": 0.0002999999422863326, | |
| "loss": 2.7665, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.05302185182754774, | |
| "grad_norm": 7.23917818069458, | |
| "learning_rate": 0.00029999994187217055, | |
| "loss": 2.9705, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 0.05354682065752346, | |
| "grad_norm": 7.654670238494873, | |
| "learning_rate": 0.0002999999414565278, | |
| "loss": 2.8659, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 0.05407178948749918, | |
| "grad_norm": 8.260897636413574, | |
| "learning_rate": 0.0002999999410394043, | |
| "loss": 2.8339, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 0.0545967583174749, | |
| "grad_norm": 7.08974552154541, | |
| "learning_rate": 0.00029999994062080007, | |
| "loss": 2.8156, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.05512172714745062, | |
| "grad_norm": 7.183873653411865, | |
| "learning_rate": 0.0002999999402007151, | |
| "loss": 2.8419, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.05564669597742634, | |
| "grad_norm": 6.617881774902344, | |
| "learning_rate": 0.0002999999397791494, | |
| "loss": 2.7173, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 0.05617166480740206, | |
| "grad_norm": 6.371913433074951, | |
| "learning_rate": 0.00029999993935610295, | |
| "loss": 2.7348, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 0.05669663363737778, | |
| "grad_norm": 6.877618789672852, | |
| "learning_rate": 0.0002999999389315758, | |
| "loss": 2.8216, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 0.0572216024673535, | |
| "grad_norm": 8.805124282836914, | |
| "learning_rate": 0.00029999993850556787, | |
| "loss": 2.7467, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 0.05774657129732922, | |
| "grad_norm": 7.108561992645264, | |
| "learning_rate": 0.0002999999380780792, | |
| "loss": 2.7983, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.05827154012730494, | |
| "grad_norm": 7.73663854598999, | |
| "learning_rate": 0.0002999999376491098, | |
| "loss": 2.7633, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 0.05879650895728066, | |
| "grad_norm": 8.234906196594238, | |
| "learning_rate": 0.0002999999372186597, | |
| "loss": 2.7472, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 0.05932147778725638, | |
| "grad_norm": 7.107717514038086, | |
| "learning_rate": 0.0002999999367867288, | |
| "loss": 2.9495, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 0.0598464466172321, | |
| "grad_norm": 6.544136047363281, | |
| "learning_rate": 0.00029999993635331727, | |
| "loss": 2.739, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 0.060371415447207824, | |
| "grad_norm": 6.966454982757568, | |
| "learning_rate": 0.0002999999359184249, | |
| "loss": 2.6947, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.060896384277183545, | |
| "grad_norm": 7.438467979431152, | |
| "learning_rate": 0.0002999999354820518, | |
| "loss": 2.8582, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 0.061421353107159266, | |
| "grad_norm": 6.687982082366943, | |
| "learning_rate": 0.00029999993504419806, | |
| "loss": 2.8573, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 0.06194632193713498, | |
| "grad_norm": 6.590853691101074, | |
| "learning_rate": 0.00029999993460486346, | |
| "loss": 2.7906, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 0.0624712907671107, | |
| "grad_norm": 6.296030044555664, | |
| "learning_rate": 0.0002999999341640482, | |
| "loss": 2.828, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 0.06299625959708642, | |
| "grad_norm": 6.9310712814331055, | |
| "learning_rate": 0.00029999993372175216, | |
| "loss": 2.7046, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.06352122842706215, | |
| "grad_norm": 7.582882404327393, | |
| "learning_rate": 0.0002999999332779754, | |
| "loss": 2.8088, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 0.06404619725703786, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.00029999993283271793, | |
| "loss": 2.8322, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 0.06457116608701359, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.00029999993243072016, | |
| "loss": 2.8203, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 0.0650961349169893, | |
| "grad_norm": 8.013832092285156, | |
| "learning_rate": 0.000299999932027523, | |
| "loss": 2.8994, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 0.06562110374696502, | |
| "grad_norm": 6.29891300201416, | |
| "learning_rate": 0.00029999993157811945, | |
| "loss": 2.9078, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.06614607257694075, | |
| "grad_norm": 7.185868263244629, | |
| "learning_rate": 0.00029999993112723517, | |
| "loss": 2.9262, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 0.06667104140691646, | |
| "grad_norm": 6.5561604499816895, | |
| "learning_rate": 0.00029999993067487014, | |
| "loss": 2.8206, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 0.06719601023689219, | |
| "grad_norm": 8.136852264404297, | |
| "learning_rate": 0.00029999993022102435, | |
| "loss": 2.8447, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 0.0677209790668679, | |
| "grad_norm": 6.196045875549316, | |
| "learning_rate": 0.0002999999297656979, | |
| "loss": 2.8216, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 0.06824594789684363, | |
| "grad_norm": 6.700967311859131, | |
| "learning_rate": 0.00029999992930889065, | |
| "loss": 2.7538, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.06877091672681934, | |
| "grad_norm": 7.149362087249756, | |
| "learning_rate": 0.00029999992885060273, | |
| "loss": 2.8565, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 0.06929588555679507, | |
| "grad_norm": 7.103426933288574, | |
| "learning_rate": 0.000299999928390834, | |
| "loss": 2.835, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 0.06982085438677078, | |
| "grad_norm": 7.073822021484375, | |
| "learning_rate": 0.0002999999279295846, | |
| "loss": 2.8655, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 0.07034582321674651, | |
| "grad_norm": 6.275856971740723, | |
| "learning_rate": 0.0002999999274668544, | |
| "loss": 2.8021, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 0.07087079204672223, | |
| "grad_norm": 7.060281276702881, | |
| "learning_rate": 0.00029999992709560413, | |
| "loss": 2.8474, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.07139576087669795, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.0002999999266302086, | |
| "loss": 2.8978, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 0.07192072970667367, | |
| "grad_norm": 6.629709720611572, | |
| "learning_rate": 0.0002999999262568261, | |
| "loss": 2.7958, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 0.07244569853664938, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.000299999925835638, | |
| "loss": 2.719, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 0.07297066736662511, | |
| "grad_norm": 9.661650657653809, | |
| "learning_rate": 0.0002999999255541796, | |
| "loss": 2.7942, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 0.07349563619660082, | |
| "grad_norm": 8.854543685913086, | |
| "learning_rate": 0.0002999999251309925, | |
| "loss": 2.7819, | |
| "step": 4200 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 10000000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 525, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.31839756017664e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |