| { | |
| "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": 129.30018615722656, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 15.8667, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.011428571428571429, | |
| "grad_norm": 110.85983276367188, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 14.2511, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.017142857142857144, | |
| "grad_norm": 91.92308807373047, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 10.6953, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.022857142857142857, | |
| "grad_norm": 47.65685272216797, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 6.6375, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.02857142857142857, | |
| "grad_norm": 10.29277229309082, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 5.6765, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03428571428571429, | |
| "grad_norm": 13.255165100097656, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 5.9722, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 10.995014190673828, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 5.498, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.045714285714285714, | |
| "grad_norm": 71.67640686035156, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 5.716, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05142857142857143, | |
| "grad_norm": 33.77011489868164, | |
| "learning_rate": 5e-05, | |
| "loss": 5.9213, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.05714285714285714, | |
| "grad_norm": 54.625999450683594, | |
| "learning_rate": 5.555555555555556e-05, | |
| "loss": 5.6055, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06285714285714286, | |
| "grad_norm": 10.449620246887207, | |
| "learning_rate": 6.111111111111112e-05, | |
| "loss": 5.294, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.06857142857142857, | |
| "grad_norm": 6.449913501739502, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 4.762, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.07428571428571429, | |
| "grad_norm": 48.93703079223633, | |
| "learning_rate": 7.222222222222222e-05, | |
| "loss": 6.2383, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 26.003328323364258, | |
| "learning_rate": 7.777777777777778e-05, | |
| "loss": 5.2407, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.08571428571428572, | |
| "grad_norm": 7.427219867706299, | |
| "learning_rate": 8.333333333333334e-05, | |
| "loss": 4.804, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.09142857142857143, | |
| "grad_norm": 5.895601272583008, | |
| "learning_rate": 8.888888888888889e-05, | |
| "loss": 4.7358, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.09714285714285714, | |
| "grad_norm": 6.431394577026367, | |
| "learning_rate": 9.444444444444444e-05, | |
| "loss": 4.5076, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.10285714285714286, | |
| "grad_norm": 4.353723526000977, | |
| "learning_rate": 0.0001, | |
| "loss": 4.687, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.10857142857142857, | |
| "grad_norm": 4.167672157287598, | |
| "learning_rate": 9.999776148326216e-05, | |
| "loss": 4.4169, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.11428571428571428, | |
| "grad_norm": 5.505213737487793, | |
| "learning_rate": 9.999104613348688e-05, | |
| "loss": 4.4702, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 6.18765926361084, | |
| "learning_rate": 9.997985455197114e-05, | |
| "loss": 4.4982, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.12571428571428572, | |
| "grad_norm": 3.865807294845581, | |
| "learning_rate": 9.996418774081658e-05, | |
| "loss": 3.9462, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.13142857142857142, | |
| "grad_norm": 5.731576442718506, | |
| "learning_rate": 9.994404710283998e-05, | |
| "loss": 4.4434, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.13714285714285715, | |
| "grad_norm": 3.568495035171509, | |
| "learning_rate": 9.991943444144757e-05, | |
| "loss": 3.8289, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.14285714285714285, | |
| "grad_norm": 11.101802825927734, | |
| "learning_rate": 9.98903519604735e-05, | |
| "loss": 4.6733, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.14857142857142858, | |
| "grad_norm": 4.277951240539551, | |
| "learning_rate": 9.985680226398261e-05, | |
| "loss": 4.561, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.15428571428571428, | |
| "grad_norm": 4.275538921356201, | |
| "learning_rate": 9.981878835603717e-05, | |
| "loss": 4.3641, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 4.504762649536133, | |
| "learning_rate": 9.977631364042795e-05, | |
| "loss": 4.6264, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.1657142857142857, | |
| "grad_norm": 5.133504390716553, | |
| "learning_rate": 9.972938192036944e-05, | |
| "loss": 4.2641, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.17142857142857143, | |
| "grad_norm": 3.9170451164245605, | |
| "learning_rate": 9.967799739815925e-05, | |
| "loss": 4.3479, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.17714285714285713, | |
| "grad_norm": 3.7615787982940674, | |
| "learning_rate": 9.962216467480193e-05, | |
| "loss": 4.5731, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.18285714285714286, | |
| "grad_norm": 3.1534392833709717, | |
| "learning_rate": 9.956188874959687e-05, | |
| "loss": 4.1598, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.18857142857142858, | |
| "grad_norm": 7.3372297286987305, | |
| "learning_rate": 9.94971750196908e-05, | |
| "loss": 4.6088, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.19428571428571428, | |
| "grad_norm": 3.007223129272461, | |
| "learning_rate": 9.942802927959443e-05, | |
| "loss": 4.3455, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 3.8887863159179688, | |
| "learning_rate": 9.93544577206636e-05, | |
| "loss": 4.5231, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2057142857142857, | |
| "grad_norm": 3.8051180839538574, | |
| "learning_rate": 9.927646693054496e-05, | |
| "loss": 4.1132, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.21142857142857144, | |
| "grad_norm": 4.708549499511719, | |
| "learning_rate": 9.919406389258607e-05, | |
| "loss": 4.5647, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.21714285714285714, | |
| "grad_norm": 3.4854278564453125, | |
| "learning_rate": 9.910725598521013e-05, | |
| "loss": 4.5256, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.22285714285714286, | |
| "grad_norm": 2.6750476360321045, | |
| "learning_rate": 9.901605098125528e-05, | |
| "loss": 4.1944, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.22857142857142856, | |
| "grad_norm": 3.341956377029419, | |
| "learning_rate": 9.892045704727864e-05, | |
| "loss": 4.3231, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2342857142857143, | |
| "grad_norm": 3.4261727333068848, | |
| "learning_rate": 9.882048274282505e-05, | |
| "loss": 4.4948, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 3.1254749298095703, | |
| "learning_rate": 9.871613701966067e-05, | |
| "loss": 4.4643, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.24571428571428572, | |
| "grad_norm": 2.991964340209961, | |
| "learning_rate": 9.860742922097141e-05, | |
| "loss": 4.48, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.25142857142857145, | |
| "grad_norm": 3.125535011291504, | |
| "learning_rate": 9.849436908052636e-05, | |
| "loss": 4.2412, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.2571428571428571, | |
| "grad_norm": 2.7782862186431885, | |
| "learning_rate": 9.837696672180618e-05, | |
| "loss": 4.3415, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.26285714285714284, | |
| "grad_norm": 5.297247409820557, | |
| "learning_rate": 9.825523265709666e-05, | |
| "loss": 4.5001, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.26857142857142857, | |
| "grad_norm": 3.659393072128296, | |
| "learning_rate": 9.812917778654748e-05, | |
| "loss": 4.5451, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.2742857142857143, | |
| "grad_norm": 6.303325176239014, | |
| "learning_rate": 9.799881339719615e-05, | |
| "loss": 4.064, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 4.318917751312256, | |
| "learning_rate": 9.786415116195732e-05, | |
| "loss": 4.3751, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 3.7108237743377686, | |
| "learning_rate": 9.772520313857775e-05, | |
| "loss": 4.22, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2914285714285714, | |
| "grad_norm": 3.1945748329162598, | |
| "learning_rate": 9.758198176855648e-05, | |
| "loss": 4.0659, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.29714285714285715, | |
| "grad_norm": 3.2484543323516846, | |
| "learning_rate": 9.743449987603083e-05, | |
| "loss": 4.2721, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3028571428571429, | |
| "grad_norm": 5.132286548614502, | |
| "learning_rate": 9.72827706666282e-05, | |
| "loss": 4.2387, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.30857142857142855, | |
| "grad_norm": 3.7148561477661133, | |
| "learning_rate": 9.712680772628364e-05, | |
| "loss": 4.2275, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.3142857142857143, | |
| "grad_norm": 4.1826863288879395, | |
| "learning_rate": 9.69666250200232e-05, | |
| "loss": 4.3568, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 4.547667026519775, | |
| "learning_rate": 9.680223689071364e-05, | |
| "loss": 4.4106, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.32571428571428573, | |
| "grad_norm": 3.7820076942443848, | |
| "learning_rate": 9.663365805777814e-05, | |
| "loss": 4.3657, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.3314285714285714, | |
| "grad_norm": 4.624260902404785, | |
| "learning_rate": 9.646090361587827e-05, | |
| "loss": 4.267, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.33714285714285713, | |
| "grad_norm": 3.133650064468384, | |
| "learning_rate": 9.628398903356239e-05, | |
| "loss": 4.1364, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.34285714285714286, | |
| "grad_norm": 2.8589680194854736, | |
| "learning_rate": 9.610293015188067e-05, | |
| "loss": 3.8508, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3485714285714286, | |
| "grad_norm": 3.252520799636841, | |
| "learning_rate": 9.591774318296661e-05, | |
| "loss": 4.2999, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.35428571428571426, | |
| "grad_norm": 3.090085744857788, | |
| "learning_rate": 9.572844470858537e-05, | |
| "loss": 4.178, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 3.4801862239837646, | |
| "learning_rate": 9.553505167864908e-05, | |
| "loss": 4.2238, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.3657142857142857, | |
| "grad_norm": 2.6730804443359375, | |
| "learning_rate": 9.533758140969912e-05, | |
| "loss": 3.6732, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.37142857142857144, | |
| "grad_norm": 3.389982223510742, | |
| "learning_rate": 9.513605158335562e-05, | |
| "loss": 3.8594, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.37714285714285717, | |
| "grad_norm": 2.627145767211914, | |
| "learning_rate": 9.493048024473412e-05, | |
| "loss": 3.6253, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.38285714285714284, | |
| "grad_norm": 4.195974826812744, | |
| "learning_rate": 9.47208858008299e-05, | |
| "loss": 4.2083, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.38857142857142857, | |
| "grad_norm": 3.1231813430786133, | |
| "learning_rate": 9.450728701886983e-05, | |
| "loss": 3.9886, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.3942857142857143, | |
| "grad_norm": 3.270745038986206, | |
| "learning_rate": 9.428970302463185e-05, | |
| "loss": 3.914, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 3.25655460357666, | |
| "learning_rate": 9.406815330073244e-05, | |
| "loss": 4.0926, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4057142857142857, | |
| "grad_norm": 3.268716335296631, | |
| "learning_rate": 9.384265768488225e-05, | |
| "loss": 4.19, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4114285714285714, | |
| "grad_norm": 2.743410348892212, | |
| "learning_rate": 9.36132363681097e-05, | |
| "loss": 4.0209, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.41714285714285715, | |
| "grad_norm": 3.016958236694336, | |
| "learning_rate": 9.337990989295306e-05, | |
| "loss": 4.1743, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4228571428571429, | |
| "grad_norm": 2.9224915504455566, | |
| "learning_rate": 9.314269915162114e-05, | |
| "loss": 4.1256, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.42857142857142855, | |
| "grad_norm": 3.1441285610198975, | |
| "learning_rate": 9.290162538412256e-05, | |
| "loss": 4.0118, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4342857142857143, | |
| "grad_norm": 2.9432973861694336, | |
| "learning_rate": 9.265671017636383e-05, | |
| "loss": 4.1406, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 3.142632484436035, | |
| "learning_rate": 9.240797545821667e-05, | |
| "loss": 4.1144, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.44571428571428573, | |
| "grad_norm": 2.4169223308563232, | |
| "learning_rate": 9.215544350155422e-05, | |
| "loss": 3.931, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.4514285714285714, | |
| "grad_norm": 3.0668325424194336, | |
| "learning_rate": 9.1899136918257e-05, | |
| "loss": 4.1284, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 3.300251007080078, | |
| "learning_rate": 9.163907865818806e-05, | |
| "loss": 4.1006, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.46285714285714286, | |
| "grad_norm": 2.1830291748046875, | |
| "learning_rate": 9.13752920071381e-05, | |
| "loss": 3.9359, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.4685714285714286, | |
| "grad_norm": 2.779247522354126, | |
| "learning_rate": 9.110780058474052e-05, | |
| "loss": 3.8829, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.4742857142857143, | |
| "grad_norm": 2.6296870708465576, | |
| "learning_rate": 9.08366283423563e-05, | |
| "loss": 3.9985, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 2.5750529766082764, | |
| "learning_rate": 9.056179956092962e-05, | |
| "loss": 4.0632, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.4857142857142857, | |
| "grad_norm": 2.8682682514190674, | |
| "learning_rate": 9.028333884881357e-05, | |
| "loss": 3.9295, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.49142857142857144, | |
| "grad_norm": 1.976205825805664, | |
| "learning_rate": 9.000127113956674e-05, | |
| "loss": 3.9259, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.49714285714285716, | |
| "grad_norm": 2.873842716217041, | |
| "learning_rate": 8.971562168972064e-05, | |
| "loss": 4.052, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5028571428571429, | |
| "grad_norm": 2.570995569229126, | |
| "learning_rate": 8.94264160765183e-05, | |
| "loss": 3.8477, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5085714285714286, | |
| "grad_norm": 2.031536340713501, | |
| "learning_rate": 8.913368019562391e-05, | |
| "loss": 3.7269, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5142857142857142, | |
| "grad_norm": 2.7151906490325928, | |
| "learning_rate": 8.883744025880428e-05, | |
| "loss": 3.9988, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 2.0973048210144043, | |
| "learning_rate": 8.853772279158166e-05, | |
| "loss": 3.7282, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5257142857142857, | |
| "grad_norm": 2.9533870220184326, | |
| "learning_rate": 8.823455463085873e-05, | |
| "loss": 3.7731, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5314285714285715, | |
| "grad_norm": 2.6501846313476562, | |
| "learning_rate": 8.79279629225156e-05, | |
| "loss": 3.9616, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.5371428571428571, | |
| "grad_norm": 3.004484176635742, | |
| "learning_rate": 8.761797511897906e-05, | |
| "loss": 3.8008, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.5428571428571428, | |
| "grad_norm": 2.8424887657165527, | |
| "learning_rate": 8.730461897676464e-05, | |
| "loss": 4.0264, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.5485714285714286, | |
| "grad_norm": 2.5832021236419678, | |
| "learning_rate": 8.698792255399104e-05, | |
| "loss": 4.0264, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.5542857142857143, | |
| "grad_norm": 2.3830084800720215, | |
| "learning_rate": 8.666791420786803e-05, | |
| "loss": 3.5794, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 2.384873628616333, | |
| "learning_rate": 8.634462259215719e-05, | |
| "loss": 3.9785, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.5657142857142857, | |
| "grad_norm": 2.895443916320801, | |
| "learning_rate": 8.60180766546062e-05, | |
| "loss": 3.9547, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.5714285714285714, | |
| "grad_norm": 11.3631010055542, | |
| "learning_rate": 8.568830563435694e-05, | |
| "loss": 3.943, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5771428571428572, | |
| "grad_norm": 3.431328058242798, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 3.9428, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.5828571428571429, | |
| "grad_norm": 2.937948703765869, | |
| "learning_rate": 8.501920674356754e-05, | |
| "loss": 3.8091, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.5885714285714285, | |
| "grad_norm": 2.8436367511749268, | |
| "learning_rate": 8.467993878459004e-05, | |
| "loss": 3.9259, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.5942857142857143, | |
| "grad_norm": 2.3942790031433105, | |
| "learning_rate": 8.433756556067506e-05, | |
| "loss": 3.73, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 2.446234703063965, | |
| "learning_rate": 8.39921177281503e-05, | |
| "loss": 3.7169, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6057142857142858, | |
| "grad_norm": 3.4125404357910156, | |
| "learning_rate": 8.364362621864595e-05, | |
| "loss": 3.9326, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6114285714285714, | |
| "grad_norm": 3.143772602081299, | |
| "learning_rate": 8.329212223632511e-05, | |
| "loss": 3.8299, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6171428571428571, | |
| "grad_norm": 4.136781215667725, | |
| "learning_rate": 8.293763725508969e-05, | |
| "loss": 3.884, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6228571428571429, | |
| "grad_norm": 2.4363884925842285, | |
| "learning_rate": 8.258020301576224e-05, | |
| "loss": 3.8588, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.6285714285714286, | |
| "grad_norm": 2.848367214202881, | |
| "learning_rate": 8.221985152324385e-05, | |
| "loss": 3.8175, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6342857142857142, | |
| "grad_norm": 2.929558515548706, | |
| "learning_rate": 8.185661504364844e-05, | |
| "loss": 3.8789, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.6439623832702637, | |
| "learning_rate": 8.149052610141357e-05, | |
| "loss": 3.4318, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.6457142857142857, | |
| "grad_norm": 3.097602605819702, | |
| "learning_rate": 8.112161747638823e-05, | |
| "loss": 3.8056, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.6514285714285715, | |
| "grad_norm": 2.1639912128448486, | |
| "learning_rate": 8.074992220089769e-05, | |
| "loss": 3.2157, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.6571428571428571, | |
| "grad_norm": 2.78663969039917, | |
| "learning_rate": 8.037547355678577e-05, | |
| "loss": 3.8593, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.6628571428571428, | |
| "grad_norm": 2.2711899280548096, | |
| "learning_rate": 7.999830507243478e-05, | |
| "loss": 3.6799, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.6685714285714286, | |
| "grad_norm": 2.797144651412964, | |
| "learning_rate": 7.961845051976334e-05, | |
| "loss": 3.7965, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.6742857142857143, | |
| "grad_norm": 2.9775500297546387, | |
| "learning_rate": 7.923594391120236e-05, | |
| "loss": 3.7851, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 3.076859951019287, | |
| "learning_rate": 7.88508194966497e-05, | |
| "loss": 3.8162, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.6857142857142857, | |
| "grad_norm": 2.914283514022827, | |
| "learning_rate": 7.846311176040331e-05, | |
| "loss": 3.7705, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6914285714285714, | |
| "grad_norm": 3.632291078567505, | |
| "learning_rate": 7.80728554180734e-05, | |
| "loss": 3.5655, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.6971428571428572, | |
| "grad_norm": 3.6447882652282715, | |
| "learning_rate": 7.768008541347423e-05, | |
| "loss": 3.8604, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7028571428571428, | |
| "grad_norm": 3.2929651737213135, | |
| "learning_rate": 7.728483691549491e-05, | |
| "loss": 3.7626, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.7085714285714285, | |
| "grad_norm": 3.310927629470825, | |
| "learning_rate": 7.688714531495061e-05, | |
| "loss": 3.6779, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 3.0188148021698, | |
| "learning_rate": 7.648704622141347e-05, | |
| "loss": 3.8624, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 3.025285243988037, | |
| "learning_rate": 7.608457546002424e-05, | |
| "loss": 3.7654, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.7257142857142858, | |
| "grad_norm": 3.460146903991699, | |
| "learning_rate": 7.567976906828431e-05, | |
| "loss": 3.5852, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.7314285714285714, | |
| "grad_norm": 2.4294722080230713, | |
| "learning_rate": 7.527266329282905e-05, | |
| "loss": 3.5941, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.7371428571428571, | |
| "grad_norm": 2.584930419921875, | |
| "learning_rate": 7.486329458618215e-05, | |
| "loss": 3.6992, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.7428571428571429, | |
| "grad_norm": 4.177308082580566, | |
| "learning_rate": 7.445169960349167e-05, | |
| "loss": 3.7905, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7485714285714286, | |
| "grad_norm": 3.4889941215515137, | |
| "learning_rate": 7.403791519924794e-05, | |
| "loss": 3.7045, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.7542857142857143, | |
| "grad_norm": 2.7580018043518066, | |
| "learning_rate": 7.362197842398355e-05, | |
| "loss": 3.8002, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 3.120452404022217, | |
| "learning_rate": 7.320392652095585e-05, | |
| "loss": 3.6098, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.7657142857142857, | |
| "grad_norm": 3.56699538230896, | |
| "learning_rate": 7.278379692281208e-05, | |
| "loss": 3.6904, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.7714285714285715, | |
| "grad_norm": 3.2482192516326904, | |
| "learning_rate": 7.23616272482378e-05, | |
| "loss": 3.761, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.7771428571428571, | |
| "grad_norm": 2.721845865249634, | |
| "learning_rate": 7.193745529858826e-05, | |
| "loss": 3.6866, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.7828571428571428, | |
| "grad_norm": 3.651869773864746, | |
| "learning_rate": 7.151131905450386e-05, | |
| "loss": 3.7425, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.7885714285714286, | |
| "grad_norm": 3.018843412399292, | |
| "learning_rate": 7.10832566725092e-05, | |
| "loss": 3.7759, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.7942857142857143, | |
| "grad_norm": 3.1514625549316406, | |
| "learning_rate": 7.065330648159656e-05, | |
| "loss": 3.6555, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.248689889907837, | |
| "learning_rate": 7.022150697979384e-05, | |
| "loss": 3.5785, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.8057142857142857, | |
| "grad_norm": 3.0343570709228516, | |
| "learning_rate": 6.97878968307176e-05, | |
| "loss": 3.5526, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.8114285714285714, | |
| "grad_norm": 2.7514946460723877, | |
| "learning_rate": 6.935251486011087e-05, | |
| "loss": 3.5474, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.8171428571428572, | |
| "grad_norm": 2.9358835220336914, | |
| "learning_rate": 6.891540005236675e-05, | |
| "loss": 3.6117, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.8228571428571428, | |
| "grad_norm": 2.2576956748962402, | |
| "learning_rate": 6.847659154703785e-05, | |
| "loss": 3.6269, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.8285714285714286, | |
| "grad_norm": 2.26851224899292, | |
| "learning_rate": 6.803612863533148e-05, | |
| "loss": 3.4223, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.8342857142857143, | |
| "grad_norm": 2.2012686729431152, | |
| "learning_rate": 6.759405075659166e-05, | |
| "loss": 3.3818, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 2.869168281555176, | |
| "learning_rate": 6.715039749476763e-05, | |
| "loss": 3.7192, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.8457142857142858, | |
| "grad_norm": 2.460110902786255, | |
| "learning_rate": 6.67052085748695e-05, | |
| "loss": 3.5411, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.8514285714285714, | |
| "grad_norm": 2.6905150413513184, | |
| "learning_rate": 6.625852385941119e-05, | |
| "loss": 3.2541, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.8571428571428571, | |
| "grad_norm": 2.948406934738159, | |
| "learning_rate": 6.58103833448412e-05, | |
| "loss": 3.6832, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8628571428571429, | |
| "grad_norm": 2.650905132293701, | |
| "learning_rate": 6.536082715796125e-05, | |
| "loss": 3.5974, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.8685714285714285, | |
| "grad_norm": 2.497577667236328, | |
| "learning_rate": 6.490989555233327e-05, | |
| "loss": 3.4876, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.8742857142857143, | |
| "grad_norm": 2.563850164413452, | |
| "learning_rate": 6.445762890467517e-05, | |
| "loss": 3.5673, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 2.6091361045837402, | |
| "learning_rate": 6.400406771124536e-05, | |
| "loss": 3.5386, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.8857142857142857, | |
| "grad_norm": 2.8260977268218994, | |
| "learning_rate": 6.354925258421675e-05, | |
| "loss": 3.5121, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.8914285714285715, | |
| "grad_norm": 2.714653253555298, | |
| "learning_rate": 6.309322424804034e-05, | |
| "loss": 3.6508, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.8971428571428571, | |
| "grad_norm": 2.722688913345337, | |
| "learning_rate": 6.263602353579868e-05, | |
| "loss": 3.4951, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.9028571428571428, | |
| "grad_norm": 2.320486068725586, | |
| "learning_rate": 6.21776913855496e-05, | |
| "loss": 3.3986, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.9085714285714286, | |
| "grad_norm": 2.469144582748413, | |
| "learning_rate": 6.171826883666074e-05, | |
| "loss": 3.6311, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.9142857142857143, | |
| "grad_norm": 2.785983085632324, | |
| "learning_rate": 6.125779702613471e-05, | |
| "loss": 3.5662, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 2.77034068107605, | |
| "learning_rate": 6.079631718492569e-05, | |
| "loss": 3.6006, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.9257142857142857, | |
| "grad_norm": 2.363115072250366, | |
| "learning_rate": 6.0333870634247645e-05, | |
| "loss": 3.6385, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.9314285714285714, | |
| "grad_norm": 3.1901586055755615, | |
| "learning_rate": 5.9870498781874365e-05, | |
| "loss": 3.5302, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.9371428571428572, | |
| "grad_norm": 2.394575357437134, | |
| "learning_rate": 5.940624311843169e-05, | |
| "loss": 3.6074, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.9428571428571428, | |
| "grad_norm": 2.175870895385742, | |
| "learning_rate": 5.8941145213682594e-05, | |
| "loss": 3.4362, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.9485714285714286, | |
| "grad_norm": 2.1898438930511475, | |
| "learning_rate": 5.847524671280484e-05, | |
| "loss": 3.5994, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.9542857142857143, | |
| "grad_norm": 2.4758195877075195, | |
| "learning_rate": 5.8008589332662136e-05, | |
| "loss": 3.5983, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 2.353057384490967, | |
| "learning_rate": 5.7541214858068705e-05, | |
| "loss": 3.4419, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.9657142857142857, | |
| "grad_norm": 2.291353940963745, | |
| "learning_rate": 5.7073165138047924e-05, | |
| "loss": 3.5085, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.9714285714285714, | |
| "grad_norm": 2.6800222396850586, | |
| "learning_rate": 5.660448208208513e-05, | |
| "loss": 3.5026, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.9771428571428571, | |
| "grad_norm": 2.412360191345215, | |
| "learning_rate": 5.613520765637489e-05, | |
| "loss": 3.5663, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.9828571428571429, | |
| "grad_norm": 3.5210206508636475, | |
| "learning_rate": 5.56653838800635e-05, | |
| "loss": 3.5395, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.9885714285714285, | |
| "grad_norm": 2.7100698947906494, | |
| "learning_rate": 5.519505282148644e-05, | |
| "loss": 3.5653, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.9942857142857143, | |
| "grad_norm": 2.5808451175689697, | |
| "learning_rate": 5.472425659440157e-05, | |
| "loss": 3.5048, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.92059588432312, | |
| "learning_rate": 5.425303735421828e-05, | |
| "loss": 3.4002, | |
| "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 | |
| } | |