| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.08924470109587243, |
| "eval_steps": 500, |
| "global_step": 1700, |
| "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 |
| } |
| ], |
| "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": 2.55681176383488e+16, |
| "train_batch_size": 8, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
