| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 171, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.017543859649122806, | |
| "grad_norm": 2.16679561289133, | |
| "learning_rate": 0.0, | |
| "loss": 0.8483, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03508771929824561, | |
| "grad_norm": 2.0661611701125664, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 0.8763, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 2.2703699566521167, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 1.0209, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.07017543859649122, | |
| "grad_norm": 2.3467447169533786, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.0622, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08771929824561403, | |
| "grad_norm": 1.843987567276686, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 0.8564, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 2.025183429740274, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.0078, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.12280701754385964, | |
| "grad_norm": 1.715914844595336, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.9971, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.14035087719298245, | |
| "grad_norm": 1.7586304736238754, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 0.9461, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 2.0895766846468975, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.9608, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.17543859649122806, | |
| "grad_norm": 1.3266913831957696, | |
| "learning_rate": 6e-06, | |
| "loss": 0.9865, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.19298245614035087, | |
| "grad_norm": 1.3505904361209118, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.0305, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 1.406349569576858, | |
| "learning_rate": 7.333333333333333e-06, | |
| "loss": 0.8987, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.22807017543859648, | |
| "grad_norm": 1.5607966014683508, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.9343, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.24561403508771928, | |
| "grad_norm": 1.6420771105271474, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 0.9125, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 1.2411935656694923, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 0.9709, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2807017543859649, | |
| "grad_norm": 1.285099856857185, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9217, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2982456140350877, | |
| "grad_norm": 1.4175055233514002, | |
| "learning_rate": 9.999661540018812e-06, | |
| "loss": 1.0184, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 1.4530633248486682, | |
| "learning_rate": 9.99864620589731e-06, | |
| "loss": 0.9018, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 1.0989691782403928, | |
| "learning_rate": 9.99695413509548e-06, | |
| "loss": 0.9223, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.3508771929824561, | |
| "grad_norm": 1.5827873971167827, | |
| "learning_rate": 9.994585556692624e-06, | |
| "loss": 0.9576, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3684210526315789, | |
| "grad_norm": 0.8592190455694358, | |
| "learning_rate": 9.991540791356342e-06, | |
| "loss": 0.8955, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.38596491228070173, | |
| "grad_norm": 0.8299857961232732, | |
| "learning_rate": 9.987820251299121e-06, | |
| "loss": 0.764, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.40350877192982454, | |
| "grad_norm": 0.9171035252108198, | |
| "learning_rate": 9.98342444022253e-06, | |
| "loss": 0.8973, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 0.740990179490497, | |
| "learning_rate": 9.978353953249023e-06, | |
| "loss": 0.7952, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.43859649122807015, | |
| "grad_norm": 0.816867836066284, | |
| "learning_rate": 9.972609476841368e-06, | |
| "loss": 0.9235, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.45614035087719296, | |
| "grad_norm": 0.9174781904591494, | |
| "learning_rate": 9.966191788709716e-06, | |
| "loss": 0.8875, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.47368421052631576, | |
| "grad_norm": 0.8146943719573134, | |
| "learning_rate": 9.959101757706308e-06, | |
| "loss": 0.8082, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.49122807017543857, | |
| "grad_norm": 0.7068924098220349, | |
| "learning_rate": 9.951340343707852e-06, | |
| "loss": 0.8764, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5087719298245614, | |
| "grad_norm": 0.5772155779629496, | |
| "learning_rate": 9.942908597485558e-06, | |
| "loss": 0.8979, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 0.5957140303254264, | |
| "learning_rate": 9.933807660562898e-06, | |
| "loss": 0.7838, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.543859649122807, | |
| "grad_norm": 0.5373347072638098, | |
| "learning_rate": 9.924038765061042e-06, | |
| "loss": 0.8394, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5614035087719298, | |
| "grad_norm": 0.5489672984945708, | |
| "learning_rate": 9.913603233532067e-06, | |
| "loss": 0.7705, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5789473684210527, | |
| "grad_norm": 0.7080158661684792, | |
| "learning_rate": 9.902502478779897e-06, | |
| "loss": 0.8684, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5964912280701754, | |
| "grad_norm": 0.6637923153891901, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 0.9679, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.6140350877192983, | |
| "grad_norm": 0.614625362769404, | |
| "learning_rate": 9.878311400921072e-06, | |
| "loss": 0.8121, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 0.6692936903098997, | |
| "learning_rate": 9.86522435289912e-06, | |
| "loss": 0.8273, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6491228070175439, | |
| "grad_norm": 0.5563500120725223, | |
| "learning_rate": 9.851478631379982e-06, | |
| "loss": 0.8624, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.5291416665431522, | |
| "learning_rate": 9.83707609731432e-06, | |
| "loss": 0.8753, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6842105263157895, | |
| "grad_norm": 0.5476616236709854, | |
| "learning_rate": 9.822018700574696e-06, | |
| "loss": 0.8981, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.7017543859649122, | |
| "grad_norm": 0.48226927788687945, | |
| "learning_rate": 9.806308479691595e-06, | |
| "loss": 0.8637, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7192982456140351, | |
| "grad_norm": 0.5462565995554229, | |
| "learning_rate": 9.789947561577445e-06, | |
| "loss": 0.8686, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 0.5505131342762903, | |
| "learning_rate": 9.77293816123866e-06, | |
| "loss": 0.8037, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7543859649122807, | |
| "grad_norm": 0.5184970156439064, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.8363, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7719298245614035, | |
| "grad_norm": 0.5089766005853604, | |
| "learning_rate": 9.736983212571646e-06, | |
| "loss": 0.7526, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 0.48754905624702916, | |
| "learning_rate": 9.718042531967918e-06, | |
| "loss": 0.8691, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8070175438596491, | |
| "grad_norm": 0.5413618026254184, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.8463, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8245614035087719, | |
| "grad_norm": 0.5572304500075312, | |
| "learning_rate": 9.678247579197658e-06, | |
| "loss": 0.8501, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.4244378287340113, | |
| "learning_rate": 9.657398694630713e-06, | |
| "loss": 0.8341, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.8596491228070176, | |
| "grad_norm": 0.5105742345404872, | |
| "learning_rate": 9.635919272833938e-06, | |
| "loss": 0.9266, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8771929824561403, | |
| "grad_norm": 0.4564062461733754, | |
| "learning_rate": 9.613812221777212e-06, | |
| "loss": 0.6665, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8947368421052632, | |
| "grad_norm": 0.49700992641846237, | |
| "learning_rate": 9.591080534401371e-06, | |
| "loss": 0.8952, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.9122807017543859, | |
| "grad_norm": 0.5028869634727683, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 0.8637, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.9298245614035088, | |
| "grad_norm": 0.47195418650900134, | |
| "learning_rate": 9.543755644867823e-06, | |
| "loss": 0.8156, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 0.49745215491197114, | |
| "learning_rate": 9.519168849742603e-06, | |
| "loss": 0.7709, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.9649122807017544, | |
| "grad_norm": 0.5022824826217431, | |
| "learning_rate": 9.493970231495836e-06, | |
| "loss": 0.8101, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.9824561403508771, | |
| "grad_norm": 0.47037159516710164, | |
| "learning_rate": 9.468163201617063e-06, | |
| "loss": 0.8214, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.5073480413670769, | |
| "learning_rate": 9.441751253965022e-06, | |
| "loss": 0.7322, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.0175438596491229, | |
| "grad_norm": 0.558679189801696, | |
| "learning_rate": 9.414737964294636e-06, | |
| "loss": 0.8407, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.0350877192982457, | |
| "grad_norm": 0.5248854391000436, | |
| "learning_rate": 9.38712698977291e-06, | |
| "loss": 0.7291, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 0.526466393071228, | |
| "learning_rate": 9.358922068483813e-06, | |
| "loss": 0.8166, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0701754385964912, | |
| "grad_norm": 0.5132233085494246, | |
| "learning_rate": 9.330127018922195e-06, | |
| "loss": 0.6905, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.087719298245614, | |
| "grad_norm": 0.49612442378563076, | |
| "learning_rate": 9.30074573947683e-06, | |
| "loss": 0.7738, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.1052631578947367, | |
| "grad_norm": 0.6785920047255632, | |
| "learning_rate": 9.27078220790263e-06, | |
| "loss": 0.8021, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.1228070175438596, | |
| "grad_norm": 0.5531838463016018, | |
| "learning_rate": 9.24024048078213e-06, | |
| "loss": 0.7934, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.1403508771929824, | |
| "grad_norm": 0.4526261590072452, | |
| "learning_rate": 9.209124692976287e-06, | |
| "loss": 0.717, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.1578947368421053, | |
| "grad_norm": 0.5292749617520837, | |
| "learning_rate": 9.177439057064684e-06, | |
| "loss": 0.7182, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.1754385964912282, | |
| "grad_norm": 0.6367696683412207, | |
| "learning_rate": 9.145187862775208e-06, | |
| "loss": 0.716, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.1929824561403508, | |
| "grad_norm": 0.4945932965291515, | |
| "learning_rate": 9.112375476403313e-06, | |
| "loss": 0.7615, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 0.5322300581906048, | |
| "learning_rate": 9.079006340220862e-06, | |
| "loss": 0.8096, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.2280701754385965, | |
| "grad_norm": 0.5094405988611275, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.7595, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2456140350877192, | |
| "grad_norm": 0.4734122799419434, | |
| "learning_rate": 9.01061596377522e-06, | |
| "loss": 0.8098, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 0.5740614471352885, | |
| "learning_rate": 8.97560398247424e-06, | |
| "loss": 0.773, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.280701754385965, | |
| "grad_norm": 0.5016714162861189, | |
| "learning_rate": 8.94005376803361e-06, | |
| "loss": 0.7777, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.2982456140350878, | |
| "grad_norm": 0.45812911376162835, | |
| "learning_rate": 8.903970133383297e-06, | |
| "loss": 0.6953, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 0.46763716191667287, | |
| "learning_rate": 8.867357963669821e-06, | |
| "loss": 0.8275, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.43596863128161417, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.7183, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.3508771929824561, | |
| "grad_norm": 0.4945275868037314, | |
| "learning_rate": 8.792567916744346e-06, | |
| "loss": 0.6897, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.368421052631579, | |
| "grad_norm": 0.5612215033828726, | |
| "learning_rate": 8.754400164907496e-06, | |
| "loss": 0.8075, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.3859649122807016, | |
| "grad_norm": 0.5483185703803954, | |
| "learning_rate": 8.715724127386971e-06, | |
| "loss": 0.7905, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.4035087719298245, | |
| "grad_norm": 0.46102392119204544, | |
| "learning_rate": 8.676545040299145e-06, | |
| "loss": 0.7406, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.4210526315789473, | |
| "grad_norm": 0.46209506924815424, | |
| "learning_rate": 8.636868207865244e-06, | |
| "loss": 0.7648, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.4385964912280702, | |
| "grad_norm": 0.4802676336967102, | |
| "learning_rate": 8.596699001693257e-06, | |
| "loss": 0.7503, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.456140350877193, | |
| "grad_norm": 0.4962673086311627, | |
| "learning_rate": 8.556042860050686e-06, | |
| "loss": 0.6859, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 0.4513099203379895, | |
| "learning_rate": 8.51490528712831e-06, | |
| "loss": 0.6124, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.4912280701754386, | |
| "grad_norm": 0.5096059517945377, | |
| "learning_rate": 8.473291852294986e-06, | |
| "loss": 0.7689, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.5087719298245614, | |
| "grad_norm": 0.4939119212097664, | |
| "learning_rate": 8.43120818934367e-06, | |
| "loss": 0.752, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "grad_norm": 0.5068556336691591, | |
| "learning_rate": 8.388659995728662e-06, | |
| "loss": 0.8851, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.543859649122807, | |
| "grad_norm": 0.7173504244274543, | |
| "learning_rate": 8.345653031794292e-06, | |
| "loss": 0.7702, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.5614035087719298, | |
| "grad_norm": 0.4461296063334958, | |
| "learning_rate": 8.302193119995038e-06, | |
| "loss": 0.8074, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 0.4656230678628809, | |
| "learning_rate": 8.258286144107277e-06, | |
| "loss": 0.709, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5964912280701755, | |
| "grad_norm": 0.505174912402358, | |
| "learning_rate": 8.213938048432697e-06, | |
| "loss": 0.7449, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.6140350877192984, | |
| "grad_norm": 0.44734144858468905, | |
| "learning_rate": 8.16915483699355e-06, | |
| "loss": 0.7561, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.631578947368421, | |
| "grad_norm": 0.4043431187007708, | |
| "learning_rate": 8.123942572719801e-06, | |
| "loss": 0.6941, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.6491228070175439, | |
| "grad_norm": 0.4114852464980933, | |
| "learning_rate": 8.078307376628292e-06, | |
| "loss": 0.7015, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.5151338846051445, | |
| "learning_rate": 8.032255426994069e-06, | |
| "loss": 0.7998, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 0.44205634304224184, | |
| "learning_rate": 7.985792958513932e-06, | |
| "loss": 0.7046, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.7017543859649122, | |
| "grad_norm": 0.4632079866563968, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.8707, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.719298245614035, | |
| "grad_norm": 0.44637907441962105, | |
| "learning_rate": 7.891661680839932e-06, | |
| "loss": 0.7484, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.736842105263158, | |
| "grad_norm": 0.4685858659647265, | |
| "learning_rate": 7.84400561551426e-06, | |
| "loss": 0.7848, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.7543859649122808, | |
| "grad_norm": 0.4568222027797856, | |
| "learning_rate": 7.795964517353734e-06, | |
| "loss": 0.7284, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.7719298245614035, | |
| "grad_norm": 0.5476306461412136, | |
| "learning_rate": 7.747544890354031e-06, | |
| "loss": 0.8004, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.7894736842105263, | |
| "grad_norm": 0.42492487518146715, | |
| "learning_rate": 7.698753289757565e-06, | |
| "loss": 0.7961, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.807017543859649, | |
| "grad_norm": 0.545769480072505, | |
| "learning_rate": 7.649596321166024e-06, | |
| "loss": 0.8133, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.8245614035087718, | |
| "grad_norm": 0.451892173394677, | |
| "learning_rate": 7.600080639646077e-06, | |
| "loss": 0.6685, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 0.4574149168680395, | |
| "learning_rate": 7.550212948828377e-06, | |
| "loss": 0.6425, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.8596491228070176, | |
| "grad_norm": 0.6386131675227494, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.8332, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.8771929824561404, | |
| "grad_norm": 0.52554380144195, | |
| "learning_rate": 7.449448591190436e-06, | |
| "loss": 0.7636, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 0.4784856775642456, | |
| "learning_rate": 7.398565566251232e-06, | |
| "loss": 0.7843, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.912280701754386, | |
| "grad_norm": 0.4484950014881073, | |
| "learning_rate": 7.347357813929455e-06, | |
| "loss": 0.6982, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.9298245614035088, | |
| "grad_norm": 0.5118269573463115, | |
| "learning_rate": 7.295832266935059e-06, | |
| "loss": 0.8068, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.9473684210526314, | |
| "grad_norm": 0.4450031299580753, | |
| "learning_rate": 7.243995901002312e-06, | |
| "loss": 0.8023, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.9649122807017543, | |
| "grad_norm": 0.5068416928084, | |
| "learning_rate": 7.191855733945388e-06, | |
| "loss": 0.772, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.9824561403508771, | |
| "grad_norm": 0.4892383894676538, | |
| "learning_rate": 7.1394188247082715e-06, | |
| "loss": 0.8481, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.6085246217123668, | |
| "learning_rate": 7.08669227240909e-06, | |
| "loss": 0.7516, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.017543859649123, | |
| "grad_norm": 0.6772580738450293, | |
| "learning_rate": 7.033683215379002e-06, | |
| "loss": 0.6293, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.0350877192982457, | |
| "grad_norm": 0.450882413053603, | |
| "learning_rate": 6.980398830195785e-06, | |
| "loss": 0.6693, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.0526315789473686, | |
| "grad_norm": 0.5904252312622317, | |
| "learning_rate": 6.9268463307122425e-06, | |
| "loss": 0.8456, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.0701754385964914, | |
| "grad_norm": 1.1536488969024838, | |
| "learning_rate": 6.873032967079562e-06, | |
| "loss": 0.711, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.087719298245614, | |
| "grad_norm": 0.6026710867167002, | |
| "learning_rate": 6.818966024765758e-06, | |
| "loss": 0.69, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 0.5519349206113559, | |
| "learning_rate": 6.7646528235693445e-06, | |
| "loss": 0.6711, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.1228070175438596, | |
| "grad_norm": 0.7092889454443306, | |
| "learning_rate": 6.710100716628345e-06, | |
| "loss": 0.6621, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.1403508771929824, | |
| "grad_norm": 0.7562555187908985, | |
| "learning_rate": 6.655317089424791e-06, | |
| "loss": 0.6179, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.1578947368421053, | |
| "grad_norm": 0.5185446425019185, | |
| "learning_rate": 6.600309358784858e-06, | |
| "loss": 0.7665, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.175438596491228, | |
| "grad_norm": 0.6715654786999794, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.6678, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.192982456140351, | |
| "grad_norm": 0.8688326949247829, | |
| "learning_rate": 6.48965140519241e-06, | |
| "loss": 0.7695, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.2105263157894735, | |
| "grad_norm": 0.5932308012835962, | |
| "learning_rate": 6.434016163555452e-06, | |
| "loss": 0.673, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.2280701754385963, | |
| "grad_norm": 0.5969608442734933, | |
| "learning_rate": 6.378186779084996e-06, | |
| "loss": 0.6205, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.245614035087719, | |
| "grad_norm": 0.5467650711645329, | |
| "learning_rate": 6.322170810186013e-06, | |
| "loss": 0.6458, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.263157894736842, | |
| "grad_norm": 0.5975174551567328, | |
| "learning_rate": 6.26597584052401e-06, | |
| "loss": 0.6918, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.280701754385965, | |
| "grad_norm": 0.5653314043661133, | |
| "learning_rate": 6.209609477998339e-06, | |
| "loss": 0.7028, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.2982456140350878, | |
| "grad_norm": 0.4960077976946443, | |
| "learning_rate": 6.153079353712201e-06, | |
| "loss": 0.6824, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.3157894736842106, | |
| "grad_norm": 0.4983550517376789, | |
| "learning_rate": 6.0963931209395165e-06, | |
| "loss": 0.6458, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 0.7749550480666753, | |
| "learning_rate": 6.039558454088796e-06, | |
| "loss": 0.7408, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.3508771929824563, | |
| "grad_norm": 0.5931221599787317, | |
| "learning_rate": 5.982583047664151e-06, | |
| "loss": 0.6533, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.3684210526315788, | |
| "grad_norm": 0.5024240270484248, | |
| "learning_rate": 5.925474615223573e-06, | |
| "loss": 0.6721, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.3859649122807016, | |
| "grad_norm": 0.47570042970709975, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.6642, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.4035087719298245, | |
| "grad_norm": 0.5947487689879833, | |
| "learning_rate": 5.810889615527839e-06, | |
| "loss": 0.7744, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.4210526315789473, | |
| "grad_norm": 0.5346541161405614, | |
| "learning_rate": 5.753428561247416e-06, | |
| "loss": 0.7585, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.43859649122807, | |
| "grad_norm": 0.5180671185567975, | |
| "learning_rate": 5.695865504800328e-06, | |
| "loss": 0.7204, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.456140350877193, | |
| "grad_norm": 0.43182983522466606, | |
| "learning_rate": 5.638208239302975e-06, | |
| "loss": 0.6737, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.473684210526316, | |
| "grad_norm": 0.5484865360730465, | |
| "learning_rate": 5.5804645706261515e-06, | |
| "loss": 0.7157, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.4912280701754383, | |
| "grad_norm": 0.5199783786417116, | |
| "learning_rate": 5.522642316338268e-06, | |
| "loss": 0.7212, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.5087719298245617, | |
| "grad_norm": 0.5007118536470155, | |
| "learning_rate": 5.464749304646963e-06, | |
| "loss": 0.5684, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "grad_norm": 0.46333226257806054, | |
| "learning_rate": 5.406793373339292e-06, | |
| "loss": 0.6569, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.543859649122807, | |
| "grad_norm": 0.421788277957023, | |
| "learning_rate": 5.348782368720627e-06, | |
| "loss": 0.671, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.56140350877193, | |
| "grad_norm": 0.6035891274019219, | |
| "learning_rate": 5.290724144552379e-06, | |
| "loss": 0.7342, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.5789473684210527, | |
| "grad_norm": 0.5026577286411762, | |
| "learning_rate": 5.232626560988735e-06, | |
| "loss": 0.6719, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.5964912280701755, | |
| "grad_norm": 0.4919422409949355, | |
| "learning_rate": 5.174497483512506e-06, | |
| "loss": 0.6919, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.6140350877192984, | |
| "grad_norm": 0.5254028230139784, | |
| "learning_rate": 5.116344781870282e-06, | |
| "loss": 0.667, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 1.553583156537818, | |
| "learning_rate": 5.0581763290069865e-06, | |
| "loss": 0.7562, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.6491228070175437, | |
| "grad_norm": 1.2338257655099236, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7293, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.49675373409102974, | |
| "learning_rate": 4.941823670993016e-06, | |
| "loss": 0.7299, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.6842105263157894, | |
| "grad_norm": 0.6695686964076406, | |
| "learning_rate": 4.883655218129719e-06, | |
| "loss": 0.7256, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.7017543859649122, | |
| "grad_norm": 0.5325751205321517, | |
| "learning_rate": 4.825502516487497e-06, | |
| "loss": 0.7213, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.719298245614035, | |
| "grad_norm": 0.48003824273117945, | |
| "learning_rate": 4.767373439011267e-06, | |
| "loss": 0.7345, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.736842105263158, | |
| "grad_norm": 0.42462804611415217, | |
| "learning_rate": 4.7092758554476215e-06, | |
| "loss": 0.6735, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.754385964912281, | |
| "grad_norm": 0.43618923437191776, | |
| "learning_rate": 4.651217631279374e-06, | |
| "loss": 0.6899, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.7719298245614032, | |
| "grad_norm": 0.45538207304086625, | |
| "learning_rate": 4.59320662666071e-06, | |
| "loss": 0.6031, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.7894736842105265, | |
| "grad_norm": 0.4370465622626561, | |
| "learning_rate": 4.53525069535304e-06, | |
| "loss": 0.6665, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.807017543859649, | |
| "grad_norm": 0.4923278629527568, | |
| "learning_rate": 4.477357683661734e-06, | |
| "loss": 0.5707, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.824561403508772, | |
| "grad_norm": 0.4969701434753241, | |
| "learning_rate": 4.4195354293738484e-06, | |
| "loss": 0.679, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.8421052631578947, | |
| "grad_norm": 0.5546933259848088, | |
| "learning_rate": 4.361791760697027e-06, | |
| "loss": 0.7043, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.8596491228070176, | |
| "grad_norm": 0.48264831967683647, | |
| "learning_rate": 4.304134495199675e-06, | |
| "loss": 0.7143, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.8771929824561404, | |
| "grad_norm": 0.4110210125093407, | |
| "learning_rate": 4.246571438752585e-06, | |
| "loss": 0.686, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.8947368421052633, | |
| "grad_norm": 0.49525919522727535, | |
| "learning_rate": 4.189110384472164e-06, | |
| "loss": 0.7517, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.912280701754386, | |
| "grad_norm": 0.4131613566268219, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.6856, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.9298245614035086, | |
| "grad_norm": 0.5175270937490849, | |
| "learning_rate": 4.074525384776428e-06, | |
| "loss": 0.6252, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 0.45596345968043617, | |
| "learning_rate": 4.017416952335849e-06, | |
| "loss": 0.6667, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.9649122807017543, | |
| "grad_norm": 0.4415633061813911, | |
| "learning_rate": 3.960441545911205e-06, | |
| "loss": 0.692, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.982456140350877, | |
| "grad_norm": 0.4727376414506676, | |
| "learning_rate": 3.903606879060483e-06, | |
| "loss": 0.7066, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.5594331001253366, | |
| "learning_rate": 3.8469206462878e-06, | |
| "loss": 0.531, | |
| "step": 171 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 285, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 39317700280320.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |