| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 285, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.017543859649122806, | |
| "grad_norm": 2.17390962259075, | |
| "learning_rate": 0.0, | |
| "loss": 0.8567, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03508771929824561, | |
| "grad_norm": 2.0656455049606737, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 0.8846, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 2.2708121056670953, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 1.0291, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.07017543859649122, | |
| "grad_norm": 2.365595492243331, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.0705, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08771929824561403, | |
| "grad_norm": 1.8384054758679238, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 0.8647, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 2.036117612986614, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.0162, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.12280701754385964, | |
| "grad_norm": 2.325763682934943, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.0055, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.14035087719298245, | |
| "grad_norm": 1.7398457854985685, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 0.9548, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 1.6439720372202178, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.9699, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.17543859649122806, | |
| "grad_norm": 1.374454745766853, | |
| "learning_rate": 6e-06, | |
| "loss": 0.9953, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.19298245614035087, | |
| "grad_norm": 1.3979480248099998, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.0385, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 1.3821620513994028, | |
| "learning_rate": 7.333333333333333e-06, | |
| "loss": 0.907, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.22807017543859648, | |
| "grad_norm": 1.4676374577990563, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.9428, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.24561403508771928, | |
| "grad_norm": 1.4070454710605471, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 0.9202, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 1.1923480023349917, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 0.9786, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2807017543859649, | |
| "grad_norm": 1.261121398600595, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9293, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2982456140350877, | |
| "grad_norm": 1.3814288026612527, | |
| "learning_rate": 9.999661540018812e-06, | |
| "loss": 1.0259, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 1.436314840212135, | |
| "learning_rate": 9.99864620589731e-06, | |
| "loss": 0.9098, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 1.1220881828694826, | |
| "learning_rate": 9.99695413509548e-06, | |
| "loss": 0.9303, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.3508771929824561, | |
| "grad_norm": 0.9534985786023815, | |
| "learning_rate": 9.994585556692624e-06, | |
| "loss": 0.9658, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3684210526315789, | |
| "grad_norm": 0.8453013121687268, | |
| "learning_rate": 9.991540791356342e-06, | |
| "loss": 0.904, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.38596491228070173, | |
| "grad_norm": 0.82784292842882, | |
| "learning_rate": 9.987820251299121e-06, | |
| "loss": 0.7715, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.40350877192982454, | |
| "grad_norm": 0.9205287128744017, | |
| "learning_rate": 9.98342444022253e-06, | |
| "loss": 0.9057, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 0.775173804826771, | |
| "learning_rate": 9.978353953249023e-06, | |
| "loss": 0.8032, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.43859649122807015, | |
| "grad_norm": 0.8335288829302911, | |
| "learning_rate": 9.972609476841368e-06, | |
| "loss": 0.9325, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.45614035087719296, | |
| "grad_norm": 0.888478333312054, | |
| "learning_rate": 9.966191788709716e-06, | |
| "loss": 0.8962, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.47368421052631576, | |
| "grad_norm": 0.8225468788134248, | |
| "learning_rate": 9.959101757706308e-06, | |
| "loss": 0.8168, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.49122807017543857, | |
| "grad_norm": 0.7149445296020257, | |
| "learning_rate": 9.951340343707852e-06, | |
| "loss": 0.8853, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5087719298245614, | |
| "grad_norm": 0.6010876915017637, | |
| "learning_rate": 9.942908597485558e-06, | |
| "loss": 0.9066, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 0.5720183594087992, | |
| "learning_rate": 9.933807660562898e-06, | |
| "loss": 0.7922, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.543859649122807, | |
| "grad_norm": 0.5481778830987162, | |
| "learning_rate": 9.924038765061042e-06, | |
| "loss": 0.8479, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.5614035087719298, | |
| "grad_norm": 0.5880660111787017, | |
| "learning_rate": 9.913603233532067e-06, | |
| "loss": 0.7787, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.5789473684210527, | |
| "grad_norm": 0.7553250242786413, | |
| "learning_rate": 9.902502478779897e-06, | |
| "loss": 0.877, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5964912280701754, | |
| "grad_norm": 0.707908642247859, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 0.9764, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.6140350877192983, | |
| "grad_norm": 0.6322147241224138, | |
| "learning_rate": 9.878311400921072e-06, | |
| "loss": 0.821, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 0.6043957998914848, | |
| "learning_rate": 9.86522435289912e-06, | |
| "loss": 0.8361, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6491228070175439, | |
| "grad_norm": 0.5786996358697318, | |
| "learning_rate": 9.851478631379982e-06, | |
| "loss": 0.8706, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.554932463436758, | |
| "learning_rate": 9.83707609731432e-06, | |
| "loss": 0.8839, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.6842105263157895, | |
| "grad_norm": 0.5551379586898564, | |
| "learning_rate": 9.822018700574696e-06, | |
| "loss": 0.9064, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.7017543859649122, | |
| "grad_norm": 0.48594696065414417, | |
| "learning_rate": 9.806308479691595e-06, | |
| "loss": 0.8728, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7192982456140351, | |
| "grad_norm": 0.5388656565361478, | |
| "learning_rate": 9.789947561577445e-06, | |
| "loss": 0.877, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 0.5567627975551519, | |
| "learning_rate": 9.77293816123866e-06, | |
| "loss": 0.8122, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7543859649122807, | |
| "grad_norm": 0.5194666382388214, | |
| "learning_rate": 9.755282581475769e-06, | |
| "loss": 0.845, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.7719298245614035, | |
| "grad_norm": 0.5091089613539003, | |
| "learning_rate": 9.736983212571646e-06, | |
| "loss": 0.7616, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 0.5072634456763613, | |
| "learning_rate": 9.718042531967918e-06, | |
| "loss": 0.8778, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8070175438596491, | |
| "grad_norm": 0.5399374251736943, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.8548, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8245614035087719, | |
| "grad_norm": 0.7945467224319946, | |
| "learning_rate": 9.678247579197658e-06, | |
| "loss": 0.8591, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.43527176017346525, | |
| "learning_rate": 9.657398694630713e-06, | |
| "loss": 0.8433, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.8596491228070176, | |
| "grad_norm": 0.5111674285583461, | |
| "learning_rate": 9.635919272833938e-06, | |
| "loss": 0.9353, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8771929824561403, | |
| "grad_norm": 0.45287978154695185, | |
| "learning_rate": 9.613812221777212e-06, | |
| "loss": 0.6752, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8947368421052632, | |
| "grad_norm": 0.483149833089933, | |
| "learning_rate": 9.591080534401371e-06, | |
| "loss": 0.9043, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.9122807017543859, | |
| "grad_norm": 0.5566166400784528, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 0.8724, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.9298245614035088, | |
| "grad_norm": 0.44710547324238276, | |
| "learning_rate": 9.543755644867823e-06, | |
| "loss": 0.8242, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 0.49749712723352163, | |
| "learning_rate": 9.519168849742603e-06, | |
| "loss": 0.7795, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.9649122807017544, | |
| "grad_norm": 0.4946270677335486, | |
| "learning_rate": 9.493970231495836e-06, | |
| "loss": 0.8189, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.9824561403508771, | |
| "grad_norm": 0.48109640594683084, | |
| "learning_rate": 9.468163201617063e-06, | |
| "loss": 0.8304, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.0373379068156559, | |
| "learning_rate": 9.441751253965022e-06, | |
| "loss": 0.7409, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.0175438596491229, | |
| "grad_norm": 0.5529568613040887, | |
| "learning_rate": 9.414737964294636e-06, | |
| "loss": 0.8502, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.0350877192982457, | |
| "grad_norm": 0.5460306870950874, | |
| "learning_rate": 9.38712698977291e-06, | |
| "loss": 0.7386, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 0.5319797898052587, | |
| "learning_rate": 9.358922068483813e-06, | |
| "loss": 0.8253, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0701754385964912, | |
| "grad_norm": 0.510779472814898, | |
| "learning_rate": 9.330127018922195e-06, | |
| "loss": 0.701, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.087719298245614, | |
| "grad_norm": 0.5048143244056522, | |
| "learning_rate": 9.30074573947683e-06, | |
| "loss": 0.784, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.1052631578947367, | |
| "grad_norm": 0.5900743386050855, | |
| "learning_rate": 9.27078220790263e-06, | |
| "loss": 0.811, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.1228070175438596, | |
| "grad_norm": 0.5505743346977228, | |
| "learning_rate": 9.24024048078213e-06, | |
| "loss": 0.8042, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.1403508771929824, | |
| "grad_norm": 0.44978217130587644, | |
| "learning_rate": 9.209124692976287e-06, | |
| "loss": 0.7262, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.1578947368421053, | |
| "grad_norm": 0.5145292288951332, | |
| "learning_rate": 9.177439057064684e-06, | |
| "loss": 0.728, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.1754385964912282, | |
| "grad_norm": 0.5214390694817843, | |
| "learning_rate": 9.145187862775208e-06, | |
| "loss": 0.7264, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.1929824561403508, | |
| "grad_norm": 0.5066506620345953, | |
| "learning_rate": 9.112375476403313e-06, | |
| "loss": 0.7718, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 0.49737078722569933, | |
| "learning_rate": 9.079006340220862e-06, | |
| "loss": 0.8193, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.2280701754385965, | |
| "grad_norm": 0.5067668860627292, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.7688, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2456140350877192, | |
| "grad_norm": 0.4852108009224975, | |
| "learning_rate": 9.01061596377522e-06, | |
| "loss": 0.8195, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 0.5616257895314729, | |
| "learning_rate": 8.97560398247424e-06, | |
| "loss": 0.7836, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.280701754385965, | |
| "grad_norm": 0.5132935934270176, | |
| "learning_rate": 8.94005376803361e-06, | |
| "loss": 0.7839, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.2982456140350878, | |
| "grad_norm": 0.45624188624339923, | |
| "learning_rate": 8.903970133383297e-06, | |
| "loss": 0.7043, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 0.4581965675617688, | |
| "learning_rate": 8.867357963669821e-06, | |
| "loss": 0.8323, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.429138693152652, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.7273, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.3508771929824561, | |
| "grad_norm": 0.49060408547786327, | |
| "learning_rate": 8.792567916744346e-06, | |
| "loss": 0.699, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.368421052631579, | |
| "grad_norm": 0.5859513305916572, | |
| "learning_rate": 8.754400164907496e-06, | |
| "loss": 0.8183, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.3859649122807016, | |
| "grad_norm": 0.5513151429997558, | |
| "learning_rate": 8.715724127386971e-06, | |
| "loss": 0.8008, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.4035087719298245, | |
| "grad_norm": 0.4737439492665639, | |
| "learning_rate": 8.676545040299145e-06, | |
| "loss": 0.7514, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.4210526315789473, | |
| "grad_norm": 0.4430384961971829, | |
| "learning_rate": 8.636868207865244e-06, | |
| "loss": 0.7744, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.4385964912280702, | |
| "grad_norm": 0.5190092302827058, | |
| "learning_rate": 8.596699001693257e-06, | |
| "loss": 0.7597, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.456140350877193, | |
| "grad_norm": 0.5046348927774563, | |
| "learning_rate": 8.556042860050686e-06, | |
| "loss": 0.6947, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 0.4635738845874236, | |
| "learning_rate": 8.51490528712831e-06, | |
| "loss": 0.6223, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.4912280701754386, | |
| "grad_norm": 0.4831983648615089, | |
| "learning_rate": 8.473291852294986e-06, | |
| "loss": 0.7768, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.5087719298245614, | |
| "grad_norm": 0.5332933107143463, | |
| "learning_rate": 8.43120818934367e-06, | |
| "loss": 0.7609, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "grad_norm": 0.4572083597753319, | |
| "learning_rate": 8.388659995728662e-06, | |
| "loss": 0.8929, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.543859649122807, | |
| "grad_norm": 0.4307686377642636, | |
| "learning_rate": 8.345653031794292e-06, | |
| "loss": 0.7789, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.5614035087719298, | |
| "grad_norm": 0.4386926697969288, | |
| "learning_rate": 8.302193119995038e-06, | |
| "loss": 0.8187, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 0.4679489325146897, | |
| "learning_rate": 8.258286144107277e-06, | |
| "loss": 0.7199, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5964912280701755, | |
| "grad_norm": 0.5056871328815058, | |
| "learning_rate": 8.213938048432697e-06, | |
| "loss": 0.7548, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.6140350877192984, | |
| "grad_norm": 0.43887641793071935, | |
| "learning_rate": 8.16915483699355e-06, | |
| "loss": 0.765, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.631578947368421, | |
| "grad_norm": 0.3941631697826748, | |
| "learning_rate": 8.123942572719801e-06, | |
| "loss": 0.7036, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.6491228070175439, | |
| "grad_norm": 0.4054073543636392, | |
| "learning_rate": 8.078307376628292e-06, | |
| "loss": 0.7105, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.5217356950954174, | |
| "learning_rate": 8.032255426994069e-06, | |
| "loss": 0.8078, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 0.46552564314897565, | |
| "learning_rate": 7.985792958513932e-06, | |
| "loss": 0.7144, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.7017543859649122, | |
| "grad_norm": 0.4602499932945155, | |
| "learning_rate": 7.938926261462366e-06, | |
| "loss": 0.8785, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.719298245614035, | |
| "grad_norm": 0.4583598887371004, | |
| "learning_rate": 7.891661680839932e-06, | |
| "loss": 0.7595, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.736842105263158, | |
| "grad_norm": 0.49088309689877446, | |
| "learning_rate": 7.84400561551426e-06, | |
| "loss": 0.7932, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.7543859649122808, | |
| "grad_norm": 0.4458906523300551, | |
| "learning_rate": 7.795964517353734e-06, | |
| "loss": 0.7369, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.7719298245614035, | |
| "grad_norm": 0.5403061121819602, | |
| "learning_rate": 7.747544890354031e-06, | |
| "loss": 0.807, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.7894736842105263, | |
| "grad_norm": 0.46467906874111714, | |
| "learning_rate": 7.698753289757565e-06, | |
| "loss": 0.8054, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.807017543859649, | |
| "grad_norm": 0.5217810211362015, | |
| "learning_rate": 7.649596321166024e-06, | |
| "loss": 0.8234, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.8245614035087718, | |
| "grad_norm": 0.4599315445356595, | |
| "learning_rate": 7.600080639646077e-06, | |
| "loss": 0.6772, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 0.6035507673354507, | |
| "learning_rate": 7.550212948828377e-06, | |
| "loss": 0.6509, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.8596491228070176, | |
| "grad_norm": 0.583339597386936, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.8411, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.8771929824561404, | |
| "grad_norm": 0.5209554672483673, | |
| "learning_rate": 7.449448591190436e-06, | |
| "loss": 0.7711, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 0.48901130794212655, | |
| "learning_rate": 7.398565566251232e-06, | |
| "loss": 0.7934, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.912280701754386, | |
| "grad_norm": 0.4539228102042579, | |
| "learning_rate": 7.347357813929455e-06, | |
| "loss": 0.709, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.9298245614035088, | |
| "grad_norm": 0.5201074143823162, | |
| "learning_rate": 7.295832266935059e-06, | |
| "loss": 0.8134, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.9473684210526314, | |
| "grad_norm": 0.4715467247650074, | |
| "learning_rate": 7.243995901002312e-06, | |
| "loss": 0.8123, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.9649122807017543, | |
| "grad_norm": 0.507075923476382, | |
| "learning_rate": 7.191855733945388e-06, | |
| "loss": 0.7815, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.9824561403508771, | |
| "grad_norm": 0.49768921308850306, | |
| "learning_rate": 7.1394188247082715e-06, | |
| "loss": 0.8565, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.6096102537231267, | |
| "learning_rate": 7.08669227240909e-06, | |
| "loss": 0.7625, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.017543859649123, | |
| "grad_norm": 0.7043872225087232, | |
| "learning_rate": 7.033683215379002e-06, | |
| "loss": 0.639, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.0350877192982457, | |
| "grad_norm": 0.4650758122271315, | |
| "learning_rate": 6.980398830195785e-06, | |
| "loss": 0.679, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.0526315789473686, | |
| "grad_norm": 0.6088715032360652, | |
| "learning_rate": 6.9268463307122425e-06, | |
| "loss": 0.856, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.0701754385964914, | |
| "grad_norm": 1.1886435575279408, | |
| "learning_rate": 6.873032967079562e-06, | |
| "loss": 0.7205, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.087719298245614, | |
| "grad_norm": 0.7299450594567016, | |
| "learning_rate": 6.818966024765758e-06, | |
| "loss": 0.699, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 0.5218546566800123, | |
| "learning_rate": 6.7646528235693445e-06, | |
| "loss": 0.6804, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.1228070175438596, | |
| "grad_norm": 0.6837638183370279, | |
| "learning_rate": 6.710100716628345e-06, | |
| "loss": 0.6719, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.1403508771929824, | |
| "grad_norm": 0.8011052935154169, | |
| "learning_rate": 6.655317089424791e-06, | |
| "loss": 0.6274, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.1578947368421053, | |
| "grad_norm": 0.5179957788560543, | |
| "learning_rate": 6.600309358784858e-06, | |
| "loss": 0.7775, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.175438596491228, | |
| "grad_norm": 0.5284478498795763, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.6762, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.192982456140351, | |
| "grad_norm": 0.8914879101655206, | |
| "learning_rate": 6.48965140519241e-06, | |
| "loss": 0.7761, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.2105263157894735, | |
| "grad_norm": 0.615326049306575, | |
| "learning_rate": 6.434016163555452e-06, | |
| "loss": 0.683, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.2280701754385963, | |
| "grad_norm": 0.5357982726280757, | |
| "learning_rate": 6.378186779084996e-06, | |
| "loss": 0.631, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.245614035087719, | |
| "grad_norm": 0.4972443171346764, | |
| "learning_rate": 6.322170810186013e-06, | |
| "loss": 0.6556, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.263157894736842, | |
| "grad_norm": 0.5925486478706566, | |
| "learning_rate": 6.26597584052401e-06, | |
| "loss": 0.702, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.280701754385965, | |
| "grad_norm": 0.6045387842509244, | |
| "learning_rate": 6.209609477998339e-06, | |
| "loss": 0.7129, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.2982456140350878, | |
| "grad_norm": 0.5068352790941699, | |
| "learning_rate": 6.153079353712201e-06, | |
| "loss": 0.6936, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.3157894736842106, | |
| "grad_norm": 0.4752699531116613, | |
| "learning_rate": 6.0963931209395165e-06, | |
| "loss": 0.6548, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 0.7356749771666816, | |
| "learning_rate": 6.039558454088796e-06, | |
| "loss": 0.7499, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.3508771929824563, | |
| "grad_norm": 0.6299561523301039, | |
| "learning_rate": 5.982583047664151e-06, | |
| "loss": 0.6646, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.3684210526315788, | |
| "grad_norm": 0.5537196627441248, | |
| "learning_rate": 5.925474615223573e-06, | |
| "loss": 0.6813, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.3859649122807016, | |
| "grad_norm": 0.48007719841912333, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.6733, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.4035087719298245, | |
| "grad_norm": 0.580380599597088, | |
| "learning_rate": 5.810889615527839e-06, | |
| "loss": 0.7844, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.4210526315789473, | |
| "grad_norm": 0.5471309082295681, | |
| "learning_rate": 5.753428561247416e-06, | |
| "loss": 0.7678, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.43859649122807, | |
| "grad_norm": 0.5424411686426356, | |
| "learning_rate": 5.695865504800328e-06, | |
| "loss": 0.7252, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.456140350877193, | |
| "grad_norm": 0.43983913055239504, | |
| "learning_rate": 5.638208239302975e-06, | |
| "loss": 0.6829, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.473684210526316, | |
| "grad_norm": 0.5268292458729069, | |
| "learning_rate": 5.5804645706261515e-06, | |
| "loss": 0.7266, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.4912280701754383, | |
| "grad_norm": 0.49884015815218946, | |
| "learning_rate": 5.522642316338268e-06, | |
| "loss": 0.7302, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.5087719298245617, | |
| "grad_norm": 0.5162775281794123, | |
| "learning_rate": 5.464749304646963e-06, | |
| "loss": 0.5789, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.526315789473684, | |
| "grad_norm": 0.45596749191400715, | |
| "learning_rate": 5.406793373339292e-06, | |
| "loss": 0.667, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.543859649122807, | |
| "grad_norm": 0.4322555929721634, | |
| "learning_rate": 5.348782368720627e-06, | |
| "loss": 0.6827, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.56140350877193, | |
| "grad_norm": 0.629335899313763, | |
| "learning_rate": 5.290724144552379e-06, | |
| "loss": 0.7443, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.5789473684210527, | |
| "grad_norm": 0.4630093139376179, | |
| "learning_rate": 5.232626560988735e-06, | |
| "loss": 0.6827, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.5964912280701755, | |
| "grad_norm": 0.4926669165411533, | |
| "learning_rate": 5.174497483512506e-06, | |
| "loss": 0.7015, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.6140350877192984, | |
| "grad_norm": 0.5691316055268869, | |
| "learning_rate": 5.116344781870282e-06, | |
| "loss": 0.6736, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 1.6440448280494468, | |
| "learning_rate": 5.0581763290069865e-06, | |
| "loss": 0.7655, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.6491228070175437, | |
| "grad_norm": 0.5874310401233204, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7407, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.5051664053486694, | |
| "learning_rate": 4.941823670993016e-06, | |
| "loss": 0.7383, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.6842105263157894, | |
| "grad_norm": 0.7059935160622425, | |
| "learning_rate": 4.883655218129719e-06, | |
| "loss": 0.7345, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.7017543859649122, | |
| "grad_norm": 0.5425365685968302, | |
| "learning_rate": 4.825502516487497e-06, | |
| "loss": 0.731, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.719298245614035, | |
| "grad_norm": 0.5367101949736602, | |
| "learning_rate": 4.767373439011267e-06, | |
| "loss": 0.7431, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.736842105263158, | |
| "grad_norm": 0.4298043489183557, | |
| "learning_rate": 4.7092758554476215e-06, | |
| "loss": 0.6835, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.754385964912281, | |
| "grad_norm": 0.4435540143227753, | |
| "learning_rate": 4.651217631279374e-06, | |
| "loss": 0.6996, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.7719298245614032, | |
| "grad_norm": 0.47605481943738515, | |
| "learning_rate": 4.59320662666071e-06, | |
| "loss": 0.6125, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.7894736842105265, | |
| "grad_norm": 0.448337343959995, | |
| "learning_rate": 4.53525069535304e-06, | |
| "loss": 0.6746, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.807017543859649, | |
| "grad_norm": 0.5232939294129769, | |
| "learning_rate": 4.477357683661734e-06, | |
| "loss": 0.5788, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.824561403508772, | |
| "grad_norm": 0.4845439919711512, | |
| "learning_rate": 4.4195354293738484e-06, | |
| "loss": 0.6832, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.8421052631578947, | |
| "grad_norm": 0.5880835459380688, | |
| "learning_rate": 4.361791760697027e-06, | |
| "loss": 0.7134, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.8596491228070176, | |
| "grad_norm": 0.5195455020831384, | |
| "learning_rate": 4.304134495199675e-06, | |
| "loss": 0.7226, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.8771929824561404, | |
| "grad_norm": 0.4277108869557638, | |
| "learning_rate": 4.246571438752585e-06, | |
| "loss": 0.6966, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.8947368421052633, | |
| "grad_norm": 0.4759690887257714, | |
| "learning_rate": 4.189110384472164e-06, | |
| "loss": 0.7588, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.912280701754386, | |
| "grad_norm": 0.4252744794262439, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.6955, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.9298245614035086, | |
| "grad_norm": 0.5631695994904372, | |
| "learning_rate": 4.074525384776428e-06, | |
| "loss": 0.6349, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "grad_norm": 0.4687066932903656, | |
| "learning_rate": 4.017416952335849e-06, | |
| "loss": 0.6758, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.9649122807017543, | |
| "grad_norm": 0.46018431361348, | |
| "learning_rate": 3.960441545911205e-06, | |
| "loss": 0.7019, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.982456140350877, | |
| "grad_norm": 0.4676344690200961, | |
| "learning_rate": 3.903606879060483e-06, | |
| "loss": 0.7158, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.5371477993119267, | |
| "learning_rate": 3.8469206462878e-06, | |
| "loss": 0.541, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 3.017543859649123, | |
| "grad_norm": 0.930470607125682, | |
| "learning_rate": 3.790390522001662e-06, | |
| "loss": 0.5662, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 3.0350877192982457, | |
| "grad_norm": 0.4334793420680317, | |
| "learning_rate": 3.7340241594759917e-06, | |
| "loss": 0.6868, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 3.0526315789473686, | |
| "grad_norm": 0.6566115040957992, | |
| "learning_rate": 3.6778291898139907e-06, | |
| "loss": 0.7169, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 3.0701754385964914, | |
| "grad_norm": 0.7572917331474158, | |
| "learning_rate": 3.6218132209150047e-06, | |
| "loss": 0.7015, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.087719298245614, | |
| "grad_norm": 0.8289193497415033, | |
| "learning_rate": 3.5659838364445505e-06, | |
| "loss": 0.6275, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 3.1052631578947367, | |
| "grad_norm": 0.6401669904301059, | |
| "learning_rate": 3.51034859480759e-06, | |
| "loss": 0.7362, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 3.1228070175438596, | |
| "grad_norm": 0.5597907899606493, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.6992, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 3.1403508771929824, | |
| "grad_norm": 0.6248106529300018, | |
| "learning_rate": 3.399690641215142e-06, | |
| "loss": 0.6029, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 3.1578947368421053, | |
| "grad_norm": 0.8542419050639002, | |
| "learning_rate": 3.3446829105752103e-06, | |
| "loss": 0.6078, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.175438596491228, | |
| "grad_norm": 0.6039080068492244, | |
| "learning_rate": 3.289899283371657e-06, | |
| "loss": 0.6479, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 3.192982456140351, | |
| "grad_norm": 0.4900082801480502, | |
| "learning_rate": 3.2353471764306567e-06, | |
| "loss": 0.783, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 3.2105263157894735, | |
| "grad_norm": 0.5465493977847152, | |
| "learning_rate": 3.1810339752342446e-06, | |
| "loss": 0.691, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 3.2280701754385963, | |
| "grad_norm": 0.7038789296633885, | |
| "learning_rate": 3.12696703292044e-06, | |
| "loss": 0.6297, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 3.245614035087719, | |
| "grad_norm": 0.45512339054222933, | |
| "learning_rate": 3.0731536692877596e-06, | |
| "loss": 0.6178, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.263157894736842, | |
| "grad_norm": 0.4726443434592247, | |
| "learning_rate": 3.019601169804216e-06, | |
| "loss": 0.5521, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 3.280701754385965, | |
| "grad_norm": 0.4685542303670508, | |
| "learning_rate": 2.966316784621e-06, | |
| "loss": 0.6422, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 3.2982456140350878, | |
| "grad_norm": 0.4631944982140366, | |
| "learning_rate": 2.9133077275909112e-06, | |
| "loss": 0.7025, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 3.3157894736842106, | |
| "grad_norm": 0.5924576043505926, | |
| "learning_rate": 2.86058117529173e-06, | |
| "loss": 0.6118, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 0.5491181276770968, | |
| "learning_rate": 2.8081442660546126e-06, | |
| "loss": 0.6246, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.3508771929824563, | |
| "grad_norm": 0.48553116690644144, | |
| "learning_rate": 2.7560040989976894e-06, | |
| "loss": 0.6553, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 3.3684210526315788, | |
| "grad_norm": 0.42771474305951634, | |
| "learning_rate": 2.7041677330649408e-06, | |
| "loss": 0.5547, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.3859649122807016, | |
| "grad_norm": 0.42573934654431783, | |
| "learning_rate": 2.6526421860705474e-06, | |
| "loss": 0.5732, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 3.4035087719298245, | |
| "grad_norm": 0.49907488920065995, | |
| "learning_rate": 2.601434433748771e-06, | |
| "loss": 0.6067, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 3.4210526315789473, | |
| "grad_norm": 0.9592779665804725, | |
| "learning_rate": 2.550551408809566e-06, | |
| "loss": 0.5338, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 3.43859649122807, | |
| "grad_norm": 0.5162021459372709, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.6548, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 3.456140350877193, | |
| "grad_norm": 0.6059073128510198, | |
| "learning_rate": 2.4497870511716237e-06, | |
| "loss": 0.4876, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 3.473684210526316, | |
| "grad_norm": 0.5408928651808667, | |
| "learning_rate": 2.3999193603539234e-06, | |
| "loss": 0.6035, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 3.4912280701754383, | |
| "grad_norm": 0.5014176227298074, | |
| "learning_rate": 2.3504036788339763e-06, | |
| "loss": 0.6407, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 3.5087719298245617, | |
| "grad_norm": 0.42383422877087884, | |
| "learning_rate": 2.3012467102424373e-06, | |
| "loss": 0.5928, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.526315789473684, | |
| "grad_norm": 0.5288112882769218, | |
| "learning_rate": 2.2524551096459703e-06, | |
| "loss": 0.6449, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 3.543859649122807, | |
| "grad_norm": 0.5915282534619672, | |
| "learning_rate": 2.204035482646267e-06, | |
| "loss": 0.5212, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 3.56140350877193, | |
| "grad_norm": 0.4379025127931876, | |
| "learning_rate": 2.155994384485742e-06, | |
| "loss": 0.7214, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 3.5789473684210527, | |
| "grad_norm": 0.6021207135490363, | |
| "learning_rate": 2.1083383191600676e-06, | |
| "loss": 0.5894, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 3.5964912280701755, | |
| "grad_norm": 0.6294414164869144, | |
| "learning_rate": 2.061073738537635e-06, | |
| "loss": 0.6079, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 3.6140350877192984, | |
| "grad_norm": 0.41822179650316876, | |
| "learning_rate": 2.0142070414860704e-06, | |
| "loss": 0.5321, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 3.6315789473684212, | |
| "grad_norm": 0.630440938984064, | |
| "learning_rate": 1.9677445730059348e-06, | |
| "loss": 0.5494, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 3.6491228070175437, | |
| "grad_norm": 0.5862793381144009, | |
| "learning_rate": 1.9216926233717087e-06, | |
| "loss": 0.5808, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 3.6666666666666665, | |
| "grad_norm": 0.5268752099196548, | |
| "learning_rate": 1.8760574272802002e-06, | |
| "loss": 0.6434, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 3.6842105263157894, | |
| "grad_norm": 0.5007793054169828, | |
| "learning_rate": 1.8308451630064484e-06, | |
| "loss": 0.5732, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.7017543859649122, | |
| "grad_norm": 0.5739463035570035, | |
| "learning_rate": 1.7860619515673034e-06, | |
| "loss": 0.6594, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 3.719298245614035, | |
| "grad_norm": 0.46715504630709853, | |
| "learning_rate": 1.7417138558927244e-06, | |
| "loss": 0.6267, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 3.736842105263158, | |
| "grad_norm": 0.488393121378992, | |
| "learning_rate": 1.6978068800049624e-06, | |
| "loss": 0.6851, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 3.754385964912281, | |
| "grad_norm": 0.5140374668442462, | |
| "learning_rate": 1.6543469682057105e-06, | |
| "loss": 0.689, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 3.7719298245614032, | |
| "grad_norm": 0.4525928435963649, | |
| "learning_rate": 1.611340004271339e-06, | |
| "loss": 0.7266, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 3.7894736842105265, | |
| "grad_norm": 0.5071799443297559, | |
| "learning_rate": 1.5687918106563326e-06, | |
| "loss": 0.6257, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 3.807017543859649, | |
| "grad_norm": 0.49537060760832996, | |
| "learning_rate": 1.5267081477050132e-06, | |
| "loss": 0.6738, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 3.824561403508772, | |
| "grad_norm": 0.5190429330062747, | |
| "learning_rate": 1.4850947128716914e-06, | |
| "loss": 0.6904, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 3.8421052631578947, | |
| "grad_norm": 0.5576619065435625, | |
| "learning_rate": 1.4439571399493146e-06, | |
| "loss": 0.6532, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 3.8596491228070176, | |
| "grad_norm": 0.45071014511639373, | |
| "learning_rate": 1.4033009983067454e-06, | |
| "loss": 0.5567, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.8771929824561404, | |
| "grad_norm": 0.4546145568524574, | |
| "learning_rate": 1.3631317921347564e-06, | |
| "loss": 0.6472, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 3.8947368421052633, | |
| "grad_norm": 0.5098743553412578, | |
| "learning_rate": 1.3234549597008572e-06, | |
| "loss": 0.6379, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 3.912280701754386, | |
| "grad_norm": 0.41377944529383687, | |
| "learning_rate": 1.2842758726130283e-06, | |
| "loss": 0.5682, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 3.9298245614035086, | |
| "grad_norm": 0.5475393565937495, | |
| "learning_rate": 1.2455998350925042e-06, | |
| "loss": 0.5845, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 3.9473684210526314, | |
| "grad_norm": 0.5757723134815058, | |
| "learning_rate": 1.2074320832556558e-06, | |
| "loss": 0.6864, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 3.9649122807017543, | |
| "grad_norm": 0.6574082849648699, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.659, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 3.982456140350877, | |
| "grad_norm": 0.5445003115149143, | |
| "learning_rate": 1.132642036330181e-06, | |
| "loss": 0.6818, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.5083655780487909, | |
| "learning_rate": 1.096029866616704e-06, | |
| "loss": 0.5752, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 4.017543859649122, | |
| "grad_norm": 0.6893475891673069, | |
| "learning_rate": 1.0599462319663906e-06, | |
| "loss": 0.642, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 4.035087719298246, | |
| "grad_norm": 0.9993929249585877, | |
| "learning_rate": 1.0243960175257605e-06, | |
| "loss": 0.5493, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 4.052631578947368, | |
| "grad_norm": 0.747527250717234, | |
| "learning_rate": 9.893840362247809e-07, | |
| "loss": 0.5833, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 4.0701754385964914, | |
| "grad_norm": 0.7078497451232868, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.5685, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 4.087719298245614, | |
| "grad_norm": 0.4353674677764059, | |
| "learning_rate": 9.209936597791407e-07, | |
| "loss": 0.6209, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 4.105263157894737, | |
| "grad_norm": 0.48384928069683086, | |
| "learning_rate": 8.876245235966884e-07, | |
| "loss": 0.6159, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 4.12280701754386, | |
| "grad_norm": 0.43587471733687433, | |
| "learning_rate": 8.54812137224792e-07, | |
| "loss": 0.5494, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 4.140350877192983, | |
| "grad_norm": 0.5222256648879957, | |
| "learning_rate": 8.225609429353187e-07, | |
| "loss": 0.6468, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 4.157894736842105, | |
| "grad_norm": 0.807514003918214, | |
| "learning_rate": 7.908753070237124e-07, | |
| "loss": 0.5068, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 4.175438596491228, | |
| "grad_norm": 0.9582546420430038, | |
| "learning_rate": 7.597595192178702e-07, | |
| "loss": 0.5983, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 4.192982456140351, | |
| "grad_norm": 0.6487220861426606, | |
| "learning_rate": 7.292177920973726e-07, | |
| "loss": 0.5825, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 4.2105263157894735, | |
| "grad_norm": 0.6467921576978426, | |
| "learning_rate": 6.992542605231739e-07, | |
| "loss": 0.5056, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.228070175438597, | |
| "grad_norm": 0.468812894263145, | |
| "learning_rate": 6.698729810778065e-07, | |
| "loss": 0.5795, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 4.245614035087719, | |
| "grad_norm": 0.5460441388318322, | |
| "learning_rate": 6.410779315161885e-07, | |
| "loss": 0.6202, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 4.2631578947368425, | |
| "grad_norm": 0.543629153168606, | |
| "learning_rate": 6.128730102270897e-07, | |
| "loss": 0.5269, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 4.280701754385965, | |
| "grad_norm": 0.4994724523422941, | |
| "learning_rate": 5.852620357053651e-07, | |
| "loss": 0.6431, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 4.298245614035087, | |
| "grad_norm": 0.43914958911605123, | |
| "learning_rate": 5.582487460349806e-07, | |
| "loss": 0.6377, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 4.315789473684211, | |
| "grad_norm": 0.5121836516984376, | |
| "learning_rate": 5.318367983829393e-07, | |
| "loss": 0.6905, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 4.333333333333333, | |
| "grad_norm": 0.46643763904680813, | |
| "learning_rate": 5.06029768504166e-07, | |
| "loss": 0.638, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 4.350877192982456, | |
| "grad_norm": 0.4440677188665689, | |
| "learning_rate": 4.808311502573976e-07, | |
| "loss": 0.5915, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 4.368421052631579, | |
| "grad_norm": 0.46091229577561793, | |
| "learning_rate": 4.562443551321788e-07, | |
| "loss": 0.4897, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 4.385964912280702, | |
| "grad_norm": 0.47583839753482565, | |
| "learning_rate": 4.322727117869951e-07, | |
| "loss": 0.6752, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.4035087719298245, | |
| "grad_norm": 0.5055411035758013, | |
| "learning_rate": 4.089194655986306e-07, | |
| "loss": 0.6585, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 4.421052631578947, | |
| "grad_norm": 0.4100105705691473, | |
| "learning_rate": 3.8618777822278854e-07, | |
| "loss": 0.543, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 4.43859649122807, | |
| "grad_norm": 0.4439560766500642, | |
| "learning_rate": 3.6408072716606346e-07, | |
| "loss": 0.595, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 4.456140350877193, | |
| "grad_norm": 0.48231258644010433, | |
| "learning_rate": 3.426013053692878e-07, | |
| "loss": 0.6319, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 4.473684210526316, | |
| "grad_norm": 0.4931208352229961, | |
| "learning_rate": 3.2175242080234314e-07, | |
| "loss": 0.7107, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 4.491228070175438, | |
| "grad_norm": 0.4648087648876863, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.6523, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 4.508771929824562, | |
| "grad_norm": 0.4411066618159384, | |
| "learning_rate": 2.819574680320825e-07, | |
| "loss": 0.6142, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 4.526315789473684, | |
| "grad_norm": 0.4558209313500185, | |
| "learning_rate": 2.63016787428354e-07, | |
| "loss": 0.5524, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 4.543859649122807, | |
| "grad_norm": 0.42882981661411557, | |
| "learning_rate": 2.447174185242324e-07, | |
| "loss": 0.6856, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 4.56140350877193, | |
| "grad_norm": 0.6487060400752643, | |
| "learning_rate": 2.2706183876134047e-07, | |
| "loss": 0.474, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.578947368421053, | |
| "grad_norm": 0.44249006389467477, | |
| "learning_rate": 2.1005243842255552e-07, | |
| "loss": 0.539, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 4.5964912280701755, | |
| "grad_norm": 0.43632496512575486, | |
| "learning_rate": 1.9369152030840553e-07, | |
| "loss": 0.5375, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 4.614035087719298, | |
| "grad_norm": 0.4523152174134078, | |
| "learning_rate": 1.779812994253055e-07, | |
| "loss": 0.4791, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 4.631578947368421, | |
| "grad_norm": 0.605994596379473, | |
| "learning_rate": 1.6292390268568103e-07, | |
| "loss": 0.5545, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 4.649122807017544, | |
| "grad_norm": 0.43759323919736876, | |
| "learning_rate": 1.4852136862001766e-07, | |
| "loss": 0.6225, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 4.666666666666667, | |
| "grad_norm": 0.5104977080281076, | |
| "learning_rate": 1.3477564710088097e-07, | |
| "loss": 0.5659, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 4.684210526315789, | |
| "grad_norm": 0.4507004983448691, | |
| "learning_rate": 1.2168859907892904e-07, | |
| "loss": 0.5877, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 4.701754385964913, | |
| "grad_norm": 0.5136244631976102, | |
| "learning_rate": 1.0926199633097156e-07, | |
| "loss": 0.4858, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 4.719298245614035, | |
| "grad_norm": 0.5125812479285721, | |
| "learning_rate": 9.749752122010347e-08, | |
| "loss": 0.6235, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 4.7368421052631575, | |
| "grad_norm": 0.44090762594602095, | |
| "learning_rate": 8.639676646793382e-08, | |
| "loss": 0.5883, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 4.754385964912281, | |
| "grad_norm": 0.4182348618019405, | |
| "learning_rate": 7.59612349389599e-08, | |
| "loss": 0.6013, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 4.771929824561403, | |
| "grad_norm": 0.6164574573314598, | |
| "learning_rate": 6.61923394371039e-08, | |
| "loss": 0.7147, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 4.7894736842105265, | |
| "grad_norm": 0.4505107085404214, | |
| "learning_rate": 5.709140251444201e-08, | |
| "loss": 0.6087, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 4.807017543859649, | |
| "grad_norm": 0.5622232442022067, | |
| "learning_rate": 4.865965629214819e-08, | |
| "loss": 0.5801, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 4.824561403508772, | |
| "grad_norm": 0.43936812719582996, | |
| "learning_rate": 4.0898242293691546e-08, | |
| "loss": 0.5448, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 4.842105263157895, | |
| "grad_norm": 0.45067076226329594, | |
| "learning_rate": 3.3808211290284886e-08, | |
| "loss": 0.5924, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 4.859649122807017, | |
| "grad_norm": 0.5142701735880818, | |
| "learning_rate": 2.7390523158633552e-08, | |
| "loss": 0.6054, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 4.87719298245614, | |
| "grad_norm": 0.3765895024210577, | |
| "learning_rate": 2.1646046750978255e-08, | |
| "loss": 0.6177, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 4.894736842105263, | |
| "grad_norm": 0.4061442588365641, | |
| "learning_rate": 1.657555977746972e-08, | |
| "loss": 0.5788, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 4.912280701754386, | |
| "grad_norm": 0.5225224582633996, | |
| "learning_rate": 1.2179748700879013e-08, | |
| "loss": 0.6126, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 4.9298245614035086, | |
| "grad_norm": 0.47246871997912304, | |
| "learning_rate": 8.459208643659122e-09, | |
| "loss": 0.6049, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 4.947368421052632, | |
| "grad_norm": 0.646430004066623, | |
| "learning_rate": 5.414443307377171e-09, | |
| "loss": 0.7031, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 4.964912280701754, | |
| "grad_norm": 0.5683102682642917, | |
| "learning_rate": 3.0458649045211897e-09, | |
| "loss": 0.5542, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 4.982456140350877, | |
| "grad_norm": 0.4451463968495303, | |
| "learning_rate": 1.3537941026914302e-09, | |
| "loss": 0.6292, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.4370227235552666, | |
| "learning_rate": 3.384599811889766e-10, | |
| "loss": 0.6342, | |
| "step": 285 | |
| } | |
| ], | |
| "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": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 65402569555968.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |