| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 175, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005714285714285714, | |
| "grad_norm": 70.73185729980469, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 7.814, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.011428571428571429, | |
| "grad_norm": 67.59354400634766, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 7.4984, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.017142857142857144, | |
| "grad_norm": 33.69651794433594, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 5.9383, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.022857142857142857, | |
| "grad_norm": 13.683563232421875, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 5.1633, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.02857142857142857, | |
| "grad_norm": 14.047224044799805, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 5.2954, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03428571428571429, | |
| "grad_norm": 13.862848281860352, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 5.2335, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 11.201601028442383, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 5.1067, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.045714285714285714, | |
| "grad_norm": 9.068086624145508, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 4.8728, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05142857142857143, | |
| "grad_norm": 26.591472625732422, | |
| "learning_rate": 5e-05, | |
| "loss": 4.9554, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.05714285714285714, | |
| "grad_norm": 11.889716148376465, | |
| "learning_rate": 5.555555555555556e-05, | |
| "loss": 4.7776, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06285714285714286, | |
| "grad_norm": 8.594311714172363, | |
| "learning_rate": 6.111111111111112e-05, | |
| "loss": 4.711, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.06857142857142857, | |
| "grad_norm": 6.407460689544678, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 4.5297, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.07428571428571429, | |
| "grad_norm": 5.835541248321533, | |
| "learning_rate": 7.222222222222222e-05, | |
| "loss": 4.5921, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 5.524861812591553, | |
| "learning_rate": 7.777777777777778e-05, | |
| "loss": 4.6319, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.08571428571428572, | |
| "grad_norm": 5.910550117492676, | |
| "learning_rate": 8.333333333333334e-05, | |
| "loss": 4.5386, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.09142857142857143, | |
| "grad_norm": 3.423496723175049, | |
| "learning_rate": 8.888888888888889e-05, | |
| "loss": 4.481, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.09714285714285714, | |
| "grad_norm": 14.430135726928711, | |
| "learning_rate": 9.444444444444444e-05, | |
| "loss": 4.2537, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.10285714285714286, | |
| "grad_norm": 4.56686544418335, | |
| "learning_rate": 0.0001, | |
| "loss": 4.5151, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.10857142857142857, | |
| "grad_norm": 3.4966490268707275, | |
| "learning_rate": 9.999776148326216e-05, | |
| "loss": 4.332, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.11428571428571428, | |
| "grad_norm": 7.750316619873047, | |
| "learning_rate": 9.999104613348688e-05, | |
| "loss": 4.361, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 4.739988803863525, | |
| "learning_rate": 9.997985455197114e-05, | |
| "loss": 4.3876, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.12571428571428572, | |
| "grad_norm": 6.870810508728027, | |
| "learning_rate": 9.996418774081658e-05, | |
| "loss": 3.8275, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.13142857142857142, | |
| "grad_norm": 3.8571834564208984, | |
| "learning_rate": 9.994404710283998e-05, | |
| "loss": 4.3033, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.13714285714285715, | |
| "grad_norm": 2.915670156478882, | |
| "learning_rate": 9.991943444144757e-05, | |
| "loss": 3.7222, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 3.343397855758667, | |
| "learning_rate": 9.98903519604735e-05, | |
| "loss": 4.4069, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.14857142857142858, | |
| "grad_norm": 4.528430461883545, | |
| "learning_rate": 9.985680226398261e-05, | |
| "loss": 4.5253, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.15428571428571428, | |
| "grad_norm": 2.878030300140381, | |
| "learning_rate": 9.981878835603717e-05, | |
| "loss": 4.2524, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 4.2011637687683105, | |
| "learning_rate": 9.977631364042795e-05, | |
| "loss": 4.5261, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.1657142857142857, | |
| "grad_norm": 3.751654624938965, | |
| "learning_rate": 9.972938192036944e-05, | |
| "loss": 4.1648, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.17142857142857143, | |
| "grad_norm": 5.761055946350098, | |
| "learning_rate": 9.967799739815925e-05, | |
| "loss": 4.2666, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.17714285714285713, | |
| "grad_norm": 2.6174347400665283, | |
| "learning_rate": 9.962216467480193e-05, | |
| "loss": 4.4525, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.18285714285714286, | |
| "grad_norm": 3.8841474056243896, | |
| "learning_rate": 9.956188874959687e-05, | |
| "loss": 4.0747, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.18857142857142858, | |
| "grad_norm": 2.7967708110809326, | |
| "learning_rate": 9.94971750196908e-05, | |
| "loss": 4.4146, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.19428571428571428, | |
| "grad_norm": 3.007628917694092, | |
| "learning_rate": 9.942802927959443e-05, | |
| "loss": 4.2302, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 3.3915443420410156, | |
| "learning_rate": 9.93544577206636e-05, | |
| "loss": 4.3599, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2057142857142857, | |
| "grad_norm": 3.1952192783355713, | |
| "learning_rate": 9.927646693054496e-05, | |
| "loss": 3.9553, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.21142857142857144, | |
| "grad_norm": 3.318091630935669, | |
| "learning_rate": 9.919406389258607e-05, | |
| "loss": 4.4174, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.21714285714285714, | |
| "grad_norm": 3.1411001682281494, | |
| "learning_rate": 9.910725598521013e-05, | |
| "loss": 4.3758, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.22285714285714286, | |
| "grad_norm": 2.8873131275177, | |
| "learning_rate": 9.901605098125528e-05, | |
| "loss": 4.0678, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.22857142857142856, | |
| "grad_norm": 2.7040855884552, | |
| "learning_rate": 9.892045704727864e-05, | |
| "loss": 4.1704, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2342857142857143, | |
| "grad_norm": 2.8548882007598877, | |
| "learning_rate": 9.882048274282505e-05, | |
| "loss": 4.3153, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 2.8522284030914307, | |
| "learning_rate": 9.871613701966067e-05, | |
| "loss": 4.2952, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.24571428571428572, | |
| "grad_norm": 3.3185341358184814, | |
| "learning_rate": 9.860742922097141e-05, | |
| "loss": 4.292, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.25142857142857145, | |
| "grad_norm": 2.78251051902771, | |
| "learning_rate": 9.849436908052636e-05, | |
| "loss": 4.0693, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.2571428571428571, | |
| "grad_norm": 3.5805132389068604, | |
| "learning_rate": 9.837696672180618e-05, | |
| "loss": 4.2146, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.26285714285714284, | |
| "grad_norm": 3.7008020877838135, | |
| "learning_rate": 9.825523265709666e-05, | |
| "loss": 4.2144, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.26857142857142857, | |
| "grad_norm": 3.876337766647339, | |
| "learning_rate": 9.812917778654748e-05, | |
| "loss": 4.3116, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.2742857142857143, | |
| "grad_norm": 3.8221843242645264, | |
| "learning_rate": 9.799881339719615e-05, | |
| "loss": 3.788, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 3.6571123600006104, | |
| "learning_rate": 9.786415116195732e-05, | |
| "loss": 4.185, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 3.0925450325012207, | |
| "learning_rate": 9.772520313857775e-05, | |
| "loss": 3.9539, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2914285714285714, | |
| "grad_norm": 2.5549099445343018, | |
| "learning_rate": 9.758198176855648e-05, | |
| "loss": 3.8761, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.29714285714285715, | |
| "grad_norm": 3.2105484008789062, | |
| "learning_rate": 9.743449987603083e-05, | |
| "loss": 4.0286, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3028571428571429, | |
| "grad_norm": 3.4165728092193604, | |
| "learning_rate": 9.72827706666282e-05, | |
| "loss": 3.9907, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.30857142857142855, | |
| "grad_norm": 3.0576295852661133, | |
| "learning_rate": 9.712680772628364e-05, | |
| "loss": 3.9951, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.3142857142857143, | |
| "grad_norm": 6.778007507324219, | |
| "learning_rate": 9.69666250200232e-05, | |
| "loss": 4.0592, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 5.007216453552246, | |
| "learning_rate": 9.680223689071364e-05, | |
| "loss": 4.1202, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.32571428571428573, | |
| "grad_norm": 4.383210182189941, | |
| "learning_rate": 9.663365805777814e-05, | |
| "loss": 4.1462, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.3314285714285714, | |
| "grad_norm": 3.4334166049957275, | |
| "learning_rate": 9.646090361587827e-05, | |
| "loss": 3.9832, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.33714285714285713, | |
| "grad_norm": 3.4360697269439697, | |
| "learning_rate": 9.628398903356239e-05, | |
| "loss": 3.9158, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.34285714285714286, | |
| "grad_norm": 2.9482150077819824, | |
| "learning_rate": 9.610293015188067e-05, | |
| "loss": 3.6752, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3485714285714286, | |
| "grad_norm": 4.406130790710449, | |
| "learning_rate": 9.591774318296661e-05, | |
| "loss": 4.1023, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.35428571428571426, | |
| "grad_norm": 3.4706637859344482, | |
| "learning_rate": 9.572844470858537e-05, | |
| "loss": 3.9855, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 4.023850917816162, | |
| "learning_rate": 9.553505167864908e-05, | |
| "loss": 3.948, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.3657142857142857, | |
| "grad_norm": 2.9235880374908447, | |
| "learning_rate": 9.533758140969912e-05, | |
| "loss": 3.5046, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.37142857142857144, | |
| "grad_norm": 3.7228496074676514, | |
| "learning_rate": 9.513605158335562e-05, | |
| "loss": 3.6147, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.37714285714285717, | |
| "grad_norm": 3.482706308364868, | |
| "learning_rate": 9.493048024473412e-05, | |
| "loss": 3.5001, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.38285714285714284, | |
| "grad_norm": 4.889612674713135, | |
| "learning_rate": 9.47208858008299e-05, | |
| "loss": 3.941, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.38857142857142857, | |
| "grad_norm": 3.4616644382476807, | |
| "learning_rate": 9.450728701886983e-05, | |
| "loss": 3.7212, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.3942857142857143, | |
| "grad_norm": 3.7083253860473633, | |
| "learning_rate": 9.428970302463185e-05, | |
| "loss": 3.6287, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 3.255087375640869, | |
| "learning_rate": 9.406815330073244e-05, | |
| "loss": 3.7913, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4057142857142857, | |
| "grad_norm": 3.504664897918701, | |
| "learning_rate": 9.384265768488225e-05, | |
| "loss": 3.892, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4114285714285714, | |
| "grad_norm": 3.181671380996704, | |
| "learning_rate": 9.36132363681097e-05, | |
| "loss": 3.7464, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.41714285714285715, | |
| "grad_norm": 5.243124961853027, | |
| "learning_rate": 9.337990989295306e-05, | |
| "loss": 3.8959, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4228571428571429, | |
| "grad_norm": 3.619094133377075, | |
| "learning_rate": 9.314269915162114e-05, | |
| "loss": 3.8042, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 4.521401405334473, | |
| "learning_rate": 9.290162538412256e-05, | |
| "loss": 3.7167, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4342857142857143, | |
| "grad_norm": 5.233442306518555, | |
| "learning_rate": 9.265671017636383e-05, | |
| "loss": 3.8335, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 4.076780796051025, | |
| "learning_rate": 9.240797545821667e-05, | |
| "loss": 3.7429, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.44571428571428573, | |
| "grad_norm": 3.2406275272369385, | |
| "learning_rate": 9.215544350155422e-05, | |
| "loss": 3.6701, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.4514285714285714, | |
| "grad_norm": 3.954871654510498, | |
| "learning_rate": 9.1899136918257e-05, | |
| "loss": 3.7933, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 3.645928144454956, | |
| "learning_rate": 9.163907865818806e-05, | |
| "loss": 3.776, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.46285714285714286, | |
| "grad_norm": 2.8732380867004395, | |
| "learning_rate": 9.13752920071381e-05, | |
| "loss": 3.613, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.4685714285714286, | |
| "grad_norm": 2.751926898956299, | |
| "learning_rate": 9.110780058474052e-05, | |
| "loss": 3.5343, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.4742857142857143, | |
| "grad_norm": 3.1919469833374023, | |
| "learning_rate": 9.08366283423563e-05, | |
| "loss": 3.6823, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 3.450768232345581, | |
| "learning_rate": 9.056179956092962e-05, | |
| "loss": 3.6727, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.4857142857142857, | |
| "grad_norm": 4.474507808685303, | |
| "learning_rate": 9.028333884881357e-05, | |
| "loss": 3.6091, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.49142857142857144, | |
| "grad_norm": 4.046672344207764, | |
| "learning_rate": 9.000127113956674e-05, | |
| "loss": 3.6397, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.49714285714285716, | |
| "grad_norm": 3.351008176803589, | |
| "learning_rate": 8.971562168972064e-05, | |
| "loss": 3.6661, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5028571428571429, | |
| "grad_norm": 4.021505355834961, | |
| "learning_rate": 8.94264160765183e-05, | |
| "loss": 3.5036, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5085714285714286, | |
| "grad_norm": 2.808837413787842, | |
| "learning_rate": 8.913368019562391e-05, | |
| "loss": 3.4524, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5142857142857142, | |
| "grad_norm": 2.7998015880584717, | |
| "learning_rate": 8.883744025880428e-05, | |
| "loss": 3.6086, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 2.3896753787994385, | |
| "learning_rate": 8.853772279158166e-05, | |
| "loss": 3.4425, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5257142857142857, | |
| "grad_norm": 3.2213077545166016, | |
| "learning_rate": 8.823455463085873e-05, | |
| "loss": 3.5151, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5314285714285715, | |
| "grad_norm": 4.131877422332764, | |
| "learning_rate": 8.79279629225156e-05, | |
| "loss": 3.6083, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.5371428571428571, | |
| "grad_norm": 3.334186553955078, | |
| "learning_rate": 8.761797511897906e-05, | |
| "loss": 3.488, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.5428571428571428, | |
| "grad_norm": 3.3978617191314697, | |
| "learning_rate": 8.730461897676464e-05, | |
| "loss": 3.7274, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.5485714285714286, | |
| "grad_norm": 3.625830888748169, | |
| "learning_rate": 8.698792255399104e-05, | |
| "loss": 3.6878, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.5542857142857143, | |
| "grad_norm": 3.0797600746154785, | |
| "learning_rate": 8.666791420786803e-05, | |
| "loss": 3.3222, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 3.047025442123413, | |
| "learning_rate": 8.634462259215719e-05, | |
| "loss": 3.6581, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.5657142857142857, | |
| "grad_norm": 3.139619827270508, | |
| "learning_rate": 8.60180766546062e-05, | |
| "loss": 3.655, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 3.2921836376190186, | |
| "learning_rate": 8.568830563435694e-05, | |
| "loss": 3.5063, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5771428571428572, | |
| "grad_norm": 3.49196195602417, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 3.5654, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.5828571428571429, | |
| "grad_norm": 3.840716600418091, | |
| "learning_rate": 8.501920674356754e-05, | |
| "loss": 3.4673, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.5885714285714285, | |
| "grad_norm": 2.9666836261749268, | |
| "learning_rate": 8.467993878459004e-05, | |
| "loss": 3.6019, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.5942857142857143, | |
| "grad_norm": 2.98416805267334, | |
| "learning_rate": 8.433756556067506e-05, | |
| "loss": 3.403, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.2488508224487305, | |
| "learning_rate": 8.39921177281503e-05, | |
| "loss": 3.4254, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6057142857142858, | |
| "grad_norm": 4.834604263305664, | |
| "learning_rate": 8.364362621864595e-05, | |
| "loss": 3.6265, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6114285714285714, | |
| "grad_norm": 3.29841947555542, | |
| "learning_rate": 8.329212223632511e-05, | |
| "loss": 3.5551, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6171428571428571, | |
| "grad_norm": 3.66792631149292, | |
| "learning_rate": 8.293763725508969e-05, | |
| "loss": 3.5378, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6228571428571429, | |
| "grad_norm": 3.1973721981048584, | |
| "learning_rate": 8.258020301576224e-05, | |
| "loss": 3.5083, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.6285714285714286, | |
| "grad_norm": 3.9324676990509033, | |
| "learning_rate": 8.221985152324385e-05, | |
| "loss": 3.4797, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6342857142857142, | |
| "grad_norm": 3.004668951034546, | |
| "learning_rate": 8.185661504364844e-05, | |
| "loss": 3.5164, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.747321605682373, | |
| "learning_rate": 8.149052610141357e-05, | |
| "loss": 3.2127, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.6457142857142857, | |
| "grad_norm": 4.0046162605285645, | |
| "learning_rate": 8.112161747638823e-05, | |
| "loss": 3.5309, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.6514285714285715, | |
| "grad_norm": 2.192357063293457, | |
| "learning_rate": 8.074992220089769e-05, | |
| "loss": 3.0268, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.6571428571428571, | |
| "grad_norm": 3.4939377307891846, | |
| "learning_rate": 8.037547355678577e-05, | |
| "loss": 3.5529, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.6628571428571428, | |
| "grad_norm": 2.790357828140259, | |
| "learning_rate": 7.999830507243478e-05, | |
| "loss": 3.3849, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.6685714285714286, | |
| "grad_norm": 3.2865593433380127, | |
| "learning_rate": 7.961845051976334e-05, | |
| "loss": 3.4702, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.6742857142857143, | |
| "grad_norm": 3.31089186668396, | |
| "learning_rate": 7.923594391120236e-05, | |
| "loss": 3.4549, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 3.0583114624023438, | |
| "learning_rate": 7.88508194966497e-05, | |
| "loss": 3.496, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.6857142857142857, | |
| "grad_norm": 2.981785297393799, | |
| "learning_rate": 7.846311176040331e-05, | |
| "loss": 3.4234, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6914285714285714, | |
| "grad_norm": 2.7434380054473877, | |
| "learning_rate": 7.80728554180734e-05, | |
| "loss": 3.3552, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.6971428571428572, | |
| "grad_norm": 9.255672454833984, | |
| "learning_rate": 7.768008541347423e-05, | |
| "loss": 3.5707, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7028571428571428, | |
| "grad_norm": 3.796801805496216, | |
| "learning_rate": 7.728483691549491e-05, | |
| "loss": 3.4186, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.7085714285714285, | |
| "grad_norm": 2.910283327102661, | |
| "learning_rate": 7.688714531495061e-05, | |
| "loss": 3.3844, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 2.9970083236694336, | |
| "learning_rate": 7.648704622141347e-05, | |
| "loss": 3.5279, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 2.8664133548736572, | |
| "learning_rate": 7.608457546002424e-05, | |
| "loss": 3.4568, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.7257142857142858, | |
| "grad_norm": 3.1223716735839844, | |
| "learning_rate": 7.567976906828431e-05, | |
| "loss": 3.2902, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.7314285714285714, | |
| "grad_norm": 2.87223482131958, | |
| "learning_rate": 7.527266329282905e-05, | |
| "loss": 3.2975, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.7371428571428571, | |
| "grad_norm": 2.5925962924957275, | |
| "learning_rate": 7.486329458618215e-05, | |
| "loss": 3.3436, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.7428571428571429, | |
| "grad_norm": 3.0222792625427246, | |
| "learning_rate": 7.445169960349167e-05, | |
| "loss": 3.4291, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7485714285714286, | |
| "grad_norm": 3.4972734451293945, | |
| "learning_rate": 7.403791519924794e-05, | |
| "loss": 3.3198, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.7542857142857143, | |
| "grad_norm": 2.552716016769409, | |
| "learning_rate": 7.362197842398355e-05, | |
| "loss": 3.4851, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 3.033993721008301, | |
| "learning_rate": 7.320392652095585e-05, | |
| "loss": 3.2181, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.7657142857142857, | |
| "grad_norm": 3.41973876953125, | |
| "learning_rate": 7.278379692281208e-05, | |
| "loss": 3.319, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.7714285714285715, | |
| "grad_norm": 3.8678910732269287, | |
| "learning_rate": 7.23616272482378e-05, | |
| "loss": 3.4172, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.7771428571428571, | |
| "grad_norm": 2.930457830429077, | |
| "learning_rate": 7.193745529858826e-05, | |
| "loss": 3.3255, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.7828571428571428, | |
| "grad_norm": 2.9811675548553467, | |
| "learning_rate": 7.151131905450386e-05, | |
| "loss": 3.4066, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.7885714285714286, | |
| "grad_norm": 2.4796462059020996, | |
| "learning_rate": 7.10832566725092e-05, | |
| "loss": 3.4391, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.7942857142857143, | |
| "grad_norm": 3.213162899017334, | |
| "learning_rate": 7.065330648159656e-05, | |
| "loss": 3.2553, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 4.176576614379883, | |
| "learning_rate": 7.022150697979384e-05, | |
| "loss": 3.1612, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.8057142857142857, | |
| "grad_norm": 2.9058704376220703, | |
| "learning_rate": 6.97878968307176e-05, | |
| "loss": 3.2411, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.8114285714285714, | |
| "grad_norm": 2.4541189670562744, | |
| "learning_rate": 6.935251486011087e-05, | |
| "loss": 3.2884, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.8171428571428572, | |
| "grad_norm": 2.880113363265991, | |
| "learning_rate": 6.891540005236675e-05, | |
| "loss": 3.2428, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.8228571428571428, | |
| "grad_norm": 2.3897929191589355, | |
| "learning_rate": 6.847659154703785e-05, | |
| "loss": 3.3205, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.8285714285714286, | |
| "grad_norm": 2.817380905151367, | |
| "learning_rate": 6.803612863533148e-05, | |
| "loss": 3.1432, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.8342857142857143, | |
| "grad_norm": 2.5644307136535645, | |
| "learning_rate": 6.759405075659166e-05, | |
| "loss": 3.1011, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 3.0853381156921387, | |
| "learning_rate": 6.715039749476763e-05, | |
| "loss": 3.3309, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.8457142857142858, | |
| "grad_norm": 2.8046741485595703, | |
| "learning_rate": 6.67052085748695e-05, | |
| "loss": 3.2116, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.8514285714285714, | |
| "grad_norm": 2.470891237258911, | |
| "learning_rate": 6.625852385941119e-05, | |
| "loss": 3.013, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 2.9108824729919434, | |
| "learning_rate": 6.58103833448412e-05, | |
| "loss": 3.3611, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8628571428571429, | |
| "grad_norm": 2.715352773666382, | |
| "learning_rate": 6.536082715796125e-05, | |
| "loss": 3.2395, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.8685714285714285, | |
| "grad_norm": 3.0429553985595703, | |
| "learning_rate": 6.490989555233327e-05, | |
| "loss": 3.2163, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.8742857142857143, | |
| "grad_norm": 3.281540870666504, | |
| "learning_rate": 6.445762890467517e-05, | |
| "loss": 3.2273, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 2.3502442836761475, | |
| "learning_rate": 6.400406771124536e-05, | |
| "loss": 3.1729, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.8857142857142857, | |
| "grad_norm": 2.791484832763672, | |
| "learning_rate": 6.354925258421675e-05, | |
| "loss": 3.1035, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.8914285714285715, | |
| "grad_norm": 2.831284523010254, | |
| "learning_rate": 6.309322424804034e-05, | |
| "loss": 3.2658, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.8971428571428571, | |
| "grad_norm": 2.340580940246582, | |
| "learning_rate": 6.263602353579868e-05, | |
| "loss": 3.186, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.9028571428571428, | |
| "grad_norm": 2.8199057579040527, | |
| "learning_rate": 6.21776913855496e-05, | |
| "loss": 3.0878, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.9085714285714286, | |
| "grad_norm": 2.6800119876861572, | |
| "learning_rate": 6.171826883666074e-05, | |
| "loss": 3.2225, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.9142857142857143, | |
| "grad_norm": 3.132681369781494, | |
| "learning_rate": 6.125779702613471e-05, | |
| "loss": 3.0903, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 3.0272161960601807, | |
| "learning_rate": 6.079631718492569e-05, | |
| "loss": 3.2235, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.9257142857142857, | |
| "grad_norm": 2.942263603210449, | |
| "learning_rate": 6.0333870634247645e-05, | |
| "loss": 3.2786, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.9314285714285714, | |
| "grad_norm": 2.588752269744873, | |
| "learning_rate": 5.9870498781874365e-05, | |
| "loss": 3.1267, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.9371428571428572, | |
| "grad_norm": 2.386512279510498, | |
| "learning_rate": 5.940624311843169e-05, | |
| "loss": 3.1636, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.9428571428571428, | |
| "grad_norm": 2.2246899604797363, | |
| "learning_rate": 5.8941145213682594e-05, | |
| "loss": 3.1075, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.9485714285714286, | |
| "grad_norm": 2.6086435317993164, | |
| "learning_rate": 5.847524671280484e-05, | |
| "loss": 3.2346, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.9542857142857143, | |
| "grad_norm": 2.297578811645508, | |
| "learning_rate": 5.8008589332662136e-05, | |
| "loss": 3.1868, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 2.551158905029297, | |
| "learning_rate": 5.7541214858068705e-05, | |
| "loss": 3.0412, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.9657142857142857, | |
| "grad_norm": 2.5748233795166016, | |
| "learning_rate": 5.7073165138047924e-05, | |
| "loss": 3.0838, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.9714285714285714, | |
| "grad_norm": 2.5943634510040283, | |
| "learning_rate": 5.660448208208513e-05, | |
| "loss": 3.0985, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.9771428571428571, | |
| "grad_norm": 2.3920516967773438, | |
| "learning_rate": 5.613520765637489e-05, | |
| "loss": 3.1383, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.9828571428571429, | |
| "grad_norm": 2.8342676162719727, | |
| "learning_rate": 5.56653838800635e-05, | |
| "loss": 3.0972, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.9885714285714285, | |
| "grad_norm": 2.9993340969085693, | |
| "learning_rate": 5.519505282148644e-05, | |
| "loss": 3.0928, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.9942857142857143, | |
| "grad_norm": 2.4510772228240967, | |
| "learning_rate": 5.472425659440157e-05, | |
| "loss": 3.0435, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.8540945053100586, | |
| "learning_rate": 5.425303735421828e-05, | |
| "loss": 2.9636, | |
| "step": 175 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 350, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 175, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.9559171544252416e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |