| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 5000, | |
| "global_step": 2014, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.009930486593843098, | |
| "grad_norm": Infinity, | |
| "learning_rate": 0.0, | |
| "loss": 14.711, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.019860973187686197, | |
| "grad_norm": 4983.64990234375, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 14.5984, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.029791459781529295, | |
| "grad_norm": 19263.828125, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 14.0403, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.039721946375372394, | |
| "grad_norm": 33094.21484375, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 12.7662, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04965243296921549, | |
| "grad_norm": 92821.109375, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 11.4273, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05958291956305859, | |
| "grad_norm": 8011.658203125, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 10.0555, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06951340615690169, | |
| "grad_norm": 6365.61962890625, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 9.0921, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07944389275074479, | |
| "grad_norm": 398.2843933105469, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 7.661, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.08937437934458789, | |
| "grad_norm": 158.12139892578125, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 6.3558, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.09930486593843098, | |
| "grad_norm": 103.36075592041016, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 5.9047, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.10923535253227408, | |
| "grad_norm": 565.8269653320312, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 5.6256, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.11916583912611718, | |
| "grad_norm": 414.0867919921875, | |
| "learning_rate": 2.2e-06, | |
| "loss": 5.3668, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.12909632571996027, | |
| "grad_norm": 153.89100646972656, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 5.0483, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.13902681231380337, | |
| "grad_norm": 149.46444702148438, | |
| "learning_rate": 2.6e-06, | |
| "loss": 4.8008, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.14895729890764647, | |
| "grad_norm": 259.7363586425781, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 4.6652, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.15888778550148958, | |
| "grad_norm": 575.7271728515625, | |
| "learning_rate": 3e-06, | |
| "loss": 4.4912, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.16881827209533268, | |
| "grad_norm": 316.88525390625, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 4.2479, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.17874875868917578, | |
| "grad_norm": 310.1224670410156, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 4.0592, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.18867924528301888, | |
| "grad_norm": 366.346435546875, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 3.8585, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.19860973187686196, | |
| "grad_norm": 242.16476440429688, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 3.6879, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.20854021847070506, | |
| "grad_norm": 445.31646728515625, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 3.5381, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.21847070506454816, | |
| "grad_norm": 345.4029235839844, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 3.272, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.22840119165839126, | |
| "grad_norm": 243.06161499023438, | |
| "learning_rate": 4.4e-06, | |
| "loss": 3.0486, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.23833167825223436, | |
| "grad_norm": 237.27239990234375, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 2.7942, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.24826216484607747, | |
| "grad_norm": 209.28985595703125, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 2.5742, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.25819265143992054, | |
| "grad_norm": 458.2777099609375, | |
| "learning_rate": 5e-06, | |
| "loss": 2.5509, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.26812313803376364, | |
| "grad_norm": 431.66522216796875, | |
| "learning_rate": 5.2e-06, | |
| "loss": 2.4819, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.27805362462760674, | |
| "grad_norm": 314.18792724609375, | |
| "learning_rate": 5.400000000000001e-06, | |
| "loss": 2.2044, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.28798411122144985, | |
| "grad_norm": 366.9264831542969, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 1.9759, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.29791459781529295, | |
| "grad_norm": 148.99159240722656, | |
| "learning_rate": 5.8e-06, | |
| "loss": 1.7, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.30784508440913605, | |
| "grad_norm": 324.4754638671875, | |
| "learning_rate": 6e-06, | |
| "loss": 1.5349, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.31777557100297915, | |
| "grad_norm": 205.2728729248047, | |
| "learning_rate": 6.200000000000001e-06, | |
| "loss": 1.4228, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.32770605759682225, | |
| "grad_norm": 312.1114807128906, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 1.3114, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.33763654419066536, | |
| "grad_norm": 274.58221435546875, | |
| "learning_rate": 6.600000000000001e-06, | |
| "loss": 1.2638, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.34756703078450846, | |
| "grad_norm": 239.2190399169922, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 1.031, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.35749751737835156, | |
| "grad_norm": 208.7018585205078, | |
| "learning_rate": 7e-06, | |
| "loss": 0.92, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.36742800397219466, | |
| "grad_norm": 237.0693817138672, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 1.665, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.37735849056603776, | |
| "grad_norm": 208.263671875, | |
| "learning_rate": 7.4e-06, | |
| "loss": 0.7744, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.3872889771598808, | |
| "grad_norm": 184.25924682617188, | |
| "learning_rate": 7.600000000000001e-06, | |
| "loss": 0.809, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.3972194637537239, | |
| "grad_norm": 146.69178771972656, | |
| "learning_rate": 7.800000000000002e-06, | |
| "loss": 0.7246, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.407149950347567, | |
| "grad_norm": 207.02853393554688, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.7616, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.4170804369414101, | |
| "grad_norm": 146.81492614746094, | |
| "learning_rate": 8.2e-06, | |
| "loss": 0.5971, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.4270109235352532, | |
| "grad_norm": 135.36524963378906, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 1.2755, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4369414101290963, | |
| "grad_norm": 298.3485107421875, | |
| "learning_rate": 8.6e-06, | |
| "loss": 1.6886, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4468718967229394, | |
| "grad_norm": 240.9180908203125, | |
| "learning_rate": 8.8e-06, | |
| "loss": 0.8591, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.4568023833167825, | |
| "grad_norm": 152.96194458007812, | |
| "learning_rate": 9e-06, | |
| "loss": 0.7211, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.4667328699106256, | |
| "grad_norm": 5467.83447265625, | |
| "learning_rate": 9.180000000000002e-06, | |
| "loss": 1.4453, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.4766633565044687, | |
| "grad_norm": 249.68312072753906, | |
| "learning_rate": 9.38e-06, | |
| "loss": 0.718, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.48659384309831183, | |
| "grad_norm": 75.28675842285156, | |
| "learning_rate": 9.58e-06, | |
| "loss": 0.5454, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.49652432969215493, | |
| "grad_norm": 98.99024963378906, | |
| "learning_rate": 9.780000000000001e-06, | |
| "loss": 0.5422, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.506454816285998, | |
| "grad_norm": 97.877197265625, | |
| "learning_rate": 9.980000000000001e-06, | |
| "loss": 0.5424, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5163853028798411, | |
| "grad_norm": 58.41091537475586, | |
| "learning_rate": 9.940554821664465e-06, | |
| "loss": 0.4753, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 50.91957092285156, | |
| "learning_rate": 9.87450462351387e-06, | |
| "loss": 0.4657, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.5362462760675273, | |
| "grad_norm": 66.574462890625, | |
| "learning_rate": 9.808454425363277e-06, | |
| "loss": 0.378, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5461767626613704, | |
| "grad_norm": 108.19845581054688, | |
| "learning_rate": 9.742404227212683e-06, | |
| "loss": 0.4171, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5561072492552135, | |
| "grad_norm": 44.33118438720703, | |
| "learning_rate": 9.676354029062089e-06, | |
| "loss": 0.4289, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5660377358490566, | |
| "grad_norm": 61.063758850097656, | |
| "learning_rate": 9.610303830911495e-06, | |
| "loss": 0.3511, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5759682224428997, | |
| "grad_norm": 91.71956634521484, | |
| "learning_rate": 9.544253632760899e-06, | |
| "loss": 0.4377, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5858987090367428, | |
| "grad_norm": 46.6739501953125, | |
| "learning_rate": 9.478203434610305e-06, | |
| "loss": 0.351, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5958291956305859, | |
| "grad_norm": 38.2111701965332, | |
| "learning_rate": 9.41215323645971e-06, | |
| "loss": 0.3556, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.605759682224429, | |
| "grad_norm": 61.8499755859375, | |
| "learning_rate": 9.346103038309115e-06, | |
| "loss": 0.4027, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.6156901688182721, | |
| "grad_norm": 219.7929229736328, | |
| "learning_rate": 9.280052840158521e-06, | |
| "loss": 0.4932, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6256206554121152, | |
| "grad_norm": 52.08991241455078, | |
| "learning_rate": 9.214002642007927e-06, | |
| "loss": 0.4588, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.6355511420059583, | |
| "grad_norm": 28.5477294921875, | |
| "learning_rate": 9.147952443857333e-06, | |
| "loss": 0.3302, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6454816285998014, | |
| "grad_norm": 24.50048065185547, | |
| "learning_rate": 9.081902245706737e-06, | |
| "loss": 0.3144, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.6554121151936445, | |
| "grad_norm": 59.86844253540039, | |
| "learning_rate": 9.015852047556145e-06, | |
| "loss": 0.3919, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.6653426017874876, | |
| "grad_norm": 22.47603988647461, | |
| "learning_rate": 8.949801849405549e-06, | |
| "loss": 0.37, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.6752730883813307, | |
| "grad_norm": 21.400466918945312, | |
| "learning_rate": 8.883751651254955e-06, | |
| "loss": 0.3293, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.6852035749751738, | |
| "grad_norm": 16.669824600219727, | |
| "learning_rate": 8.81770145310436e-06, | |
| "loss": 0.2647, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.6951340615690169, | |
| "grad_norm": 23.852384567260742, | |
| "learning_rate": 8.751651254953765e-06, | |
| "loss": 0.3246, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.70506454816286, | |
| "grad_norm": 38.244258880615234, | |
| "learning_rate": 8.685601056803171e-06, | |
| "loss": 0.385, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.7149950347567031, | |
| "grad_norm": 29.78006362915039, | |
| "learning_rate": 8.619550858652577e-06, | |
| "loss": 0.3071, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.7249255213505462, | |
| "grad_norm": 24.519458770751953, | |
| "learning_rate": 8.553500660501983e-06, | |
| "loss": 0.3429, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.7348560079443893, | |
| "grad_norm": 25.153671264648438, | |
| "learning_rate": 8.487450462351387e-06, | |
| "loss": 0.3843, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.7447864945382324, | |
| "grad_norm": 27.24303436279297, | |
| "learning_rate": 8.421400264200793e-06, | |
| "loss": 0.369, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.7547169811320755, | |
| "grad_norm": 14.806120872497559, | |
| "learning_rate": 8.355350066050199e-06, | |
| "loss": 0.301, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.7646474677259185, | |
| "grad_norm": 22.724933624267578, | |
| "learning_rate": 8.289299867899603e-06, | |
| "loss": 0.3101, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.7745779543197616, | |
| "grad_norm": 15.89284610748291, | |
| "learning_rate": 8.22324966974901e-06, | |
| "loss": 0.3665, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.7845084409136047, | |
| "grad_norm": 87.12956237792969, | |
| "learning_rate": 8.157199471598415e-06, | |
| "loss": 0.4025, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.7944389275074478, | |
| "grad_norm": 27.03380012512207, | |
| "learning_rate": 8.091149273447821e-06, | |
| "loss": 0.2919, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8043694141012909, | |
| "grad_norm": 27.622365951538086, | |
| "learning_rate": 8.025099075297227e-06, | |
| "loss": 0.3441, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.814299900695134, | |
| "grad_norm": 32.42015838623047, | |
| "learning_rate": 7.959048877146633e-06, | |
| "loss": 0.3307, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.8242303872889771, | |
| "grad_norm": 16.96957015991211, | |
| "learning_rate": 7.892998678996037e-06, | |
| "loss": 0.3224, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8341608738828202, | |
| "grad_norm": 12.882108688354492, | |
| "learning_rate": 7.826948480845443e-06, | |
| "loss": 0.2992, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.8440913604766633, | |
| "grad_norm": 17.12813949584961, | |
| "learning_rate": 7.760898282694849e-06, | |
| "loss": 0.2783, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.8540218470705064, | |
| "grad_norm": 57.04343795776367, | |
| "learning_rate": 7.694848084544253e-06, | |
| "loss": 0.4445, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.8639523336643495, | |
| "grad_norm": 17.658185958862305, | |
| "learning_rate": 7.628797886393659e-06, | |
| "loss": 0.322, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.8738828202581926, | |
| "grad_norm": 37.49673080444336, | |
| "learning_rate": 7.562747688243065e-06, | |
| "loss": 0.5781, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.8838133068520357, | |
| "grad_norm": 22.256288528442383, | |
| "learning_rate": 7.49669749009247e-06, | |
| "loss": 0.3146, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.8937437934458788, | |
| "grad_norm": 28.0767765045166, | |
| "learning_rate": 7.430647291941876e-06, | |
| "loss": 0.2795, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9036742800397219, | |
| "grad_norm": 26.734256744384766, | |
| "learning_rate": 7.364597093791282e-06, | |
| "loss": 0.286, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.913604766633565, | |
| "grad_norm": 16.944440841674805, | |
| "learning_rate": 7.298546895640688e-06, | |
| "loss": 0.2396, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.9235352532274081, | |
| "grad_norm": 11.377096176147461, | |
| "learning_rate": 7.232496697490093e-06, | |
| "loss": 0.3234, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.9334657398212513, | |
| "grad_norm": 21.92275047302246, | |
| "learning_rate": 7.166446499339499e-06, | |
| "loss": 0.3054, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.9433962264150944, | |
| "grad_norm": 11.469223976135254, | |
| "learning_rate": 7.100396301188904e-06, | |
| "loss": 0.2805, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.9533267130089375, | |
| "grad_norm": 21.188453674316406, | |
| "learning_rate": 7.034346103038309e-06, | |
| "loss": 0.3169, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.9632571996027806, | |
| "grad_norm": 14.333311080932617, | |
| "learning_rate": 6.968295904887715e-06, | |
| "loss": 0.2381, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.9731876861966237, | |
| "grad_norm": 10.782018661499023, | |
| "learning_rate": 6.90224570673712e-06, | |
| "loss": 0.2914, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.9831181727904668, | |
| "grad_norm": 12.117927551269531, | |
| "learning_rate": 6.836195508586526e-06, | |
| "loss": 0.3186, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.9930486593843099, | |
| "grad_norm": 15.109565734863281, | |
| "learning_rate": 6.7701453104359315e-06, | |
| "loss": 0.2323, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.002979145978153, | |
| "grad_norm": 27.17137336730957, | |
| "learning_rate": 6.704095112285337e-06, | |
| "loss": 0.2664, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.012909632571996, | |
| "grad_norm": 17.559146881103516, | |
| "learning_rate": 6.6380449141347425e-06, | |
| "loss": 0.3117, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.0228401191658392, | |
| "grad_norm": 18.243568420410156, | |
| "learning_rate": 6.571994715984148e-06, | |
| "loss": 0.2746, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.0327706057596822, | |
| "grad_norm": 19.41541862487793, | |
| "learning_rate": 6.505944517833554e-06, | |
| "loss": 0.2757, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.0427010923535254, | |
| "grad_norm": 23.584136962890625, | |
| "learning_rate": 6.43989431968296e-06, | |
| "loss": 0.3311, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 8.158661842346191, | |
| "learning_rate": 6.3738441215323654e-06, | |
| "loss": 0.2664, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.0625620655412116, | |
| "grad_norm": 14.265728950500488, | |
| "learning_rate": 6.3077939233817705e-06, | |
| "loss": 0.3737, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.0724925521350546, | |
| "grad_norm": 45.05762481689453, | |
| "learning_rate": 6.2417437252311765e-06, | |
| "loss": 0.2805, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.0824230387288978, | |
| "grad_norm": 26.889284133911133, | |
| "learning_rate": 6.1756935270805816e-06, | |
| "loss": 0.3286, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.0923535253227408, | |
| "grad_norm": 14.202479362487793, | |
| "learning_rate": 6.1096433289299875e-06, | |
| "loss": 0.2859, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.102284011916584, | |
| "grad_norm": 14.509482383728027, | |
| "learning_rate": 6.043593130779393e-06, | |
| "loss": 0.2255, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.112214498510427, | |
| "grad_norm": 17.728469848632812, | |
| "learning_rate": 5.9775429326287985e-06, | |
| "loss": 0.2592, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.1221449851042702, | |
| "grad_norm": 15.352997779846191, | |
| "learning_rate": 5.911492734478204e-06, | |
| "loss": 0.2538, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.1320754716981132, | |
| "grad_norm": 12.579967498779297, | |
| "learning_rate": 5.845442536327609e-06, | |
| "loss": 0.2803, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.1420059582919564, | |
| "grad_norm": 9.809431076049805, | |
| "learning_rate": 5.779392338177015e-06, | |
| "loss": 0.2874, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.1519364448857994, | |
| "grad_norm": 18.45332145690918, | |
| "learning_rate": 5.7133421400264215e-06, | |
| "loss": 0.3388, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.1618669314796426, | |
| "grad_norm": 15.697205543518066, | |
| "learning_rate": 5.6472919418758266e-06, | |
| "loss": 0.6377, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.1717974180734856, | |
| "grad_norm": 245.89466857910156, | |
| "learning_rate": 5.581241743725232e-06, | |
| "loss": 0.4898, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.1817279046673286, | |
| "grad_norm": 13.679891586303711, | |
| "learning_rate": 5.515191545574638e-06, | |
| "loss": 0.3288, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.1916583912611718, | |
| "grad_norm": 16.556507110595703, | |
| "learning_rate": 5.449141347424043e-06, | |
| "loss": 0.3122, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.201588877855015, | |
| "grad_norm": 14.205398559570312, | |
| "learning_rate": 5.383091149273449e-06, | |
| "loss": 0.2981, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.211519364448858, | |
| "grad_norm": 12.833841323852539, | |
| "learning_rate": 5.317040951122854e-06, | |
| "loss": 0.3071, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.221449851042701, | |
| "grad_norm": 23.980009078979492, | |
| "learning_rate": 5.250990752972259e-06, | |
| "loss": 0.2632, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.2313803376365442, | |
| "grad_norm": 17.329069137573242, | |
| "learning_rate": 5.184940554821665e-06, | |
| "loss": 0.3039, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.2413108242303874, | |
| "grad_norm": 24.93738555908203, | |
| "learning_rate": 5.11889035667107e-06, | |
| "loss": 0.3467, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.2512413108242304, | |
| "grad_norm": 14.432474136352539, | |
| "learning_rate": 5.052840158520476e-06, | |
| "loss": 0.2253, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.2611717974180734, | |
| "grad_norm": 21.262495040893555, | |
| "learning_rate": 4.986789960369882e-06, | |
| "loss": 0.3464, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.2711022840119166, | |
| "grad_norm": 16.39492416381836, | |
| "learning_rate": 4.920739762219287e-06, | |
| "loss": 0.2812, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.2810327706057598, | |
| "grad_norm": 9.893121719360352, | |
| "learning_rate": 4.854689564068693e-06, | |
| "loss": 0.302, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.2909632571996028, | |
| "grad_norm": 7.951263427734375, | |
| "learning_rate": 4.788639365918098e-06, | |
| "loss": 0.4157, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.3008937437934458, | |
| "grad_norm": 8.177956581115723, | |
| "learning_rate": 4.722589167767504e-06, | |
| "loss": 0.206, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.310824230387289, | |
| "grad_norm": 6.12276029586792, | |
| "learning_rate": 4.65653896961691e-06, | |
| "loss": 0.2143, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.320754716981132, | |
| "grad_norm": 16.6575984954834, | |
| "learning_rate": 4.590488771466315e-06, | |
| "loss": 0.2677, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.3306852035749752, | |
| "grad_norm": 8.128823280334473, | |
| "learning_rate": 4.52443857331572e-06, | |
| "loss": 0.2279, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.3406156901688182, | |
| "grad_norm": 21.915355682373047, | |
| "learning_rate": 4.458388375165126e-06, | |
| "loss": 0.2724, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.3505461767626614, | |
| "grad_norm": 18.645946502685547, | |
| "learning_rate": 4.392338177014531e-06, | |
| "loss": 0.2307, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.3604766633565044, | |
| "grad_norm": 12.229573249816895, | |
| "learning_rate": 4.326287978863937e-06, | |
| "loss": 0.3043, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.3704071499503476, | |
| "grad_norm": 21.465803146362305, | |
| "learning_rate": 4.260237780713343e-06, | |
| "loss": 0.3025, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.3803376365441906, | |
| "grad_norm": 12.59046459197998, | |
| "learning_rate": 4.194187582562748e-06, | |
| "loss": 0.3685, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.3902681231380338, | |
| "grad_norm": 20.229293823242188, | |
| "learning_rate": 4.128137384412154e-06, | |
| "loss": 0.2627, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.4001986097318768, | |
| "grad_norm": 5.609831809997559, | |
| "learning_rate": 4.062087186261559e-06, | |
| "loss": 0.3011, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.41012909632572, | |
| "grad_norm": 13.317000389099121, | |
| "learning_rate": 3.996036988110964e-06, | |
| "loss": 0.226, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.420059582919563, | |
| "grad_norm": 10.011391639709473, | |
| "learning_rate": 3.92998678996037e-06, | |
| "loss": 0.2295, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.4299900695134062, | |
| "grad_norm": 14.472879409790039, | |
| "learning_rate": 3.863936591809776e-06, | |
| "loss": 0.2577, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.4399205561072492, | |
| "grad_norm": 7.8662614822387695, | |
| "learning_rate": 3.7978863936591815e-06, | |
| "loss": 0.3256, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.4498510427010922, | |
| "grad_norm": 23.64836883544922, | |
| "learning_rate": 3.731836195508587e-06, | |
| "loss": 0.336, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.4597815292949354, | |
| "grad_norm": 4.431447982788086, | |
| "learning_rate": 3.665785997357992e-06, | |
| "loss": 0.3114, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 1.4697120158887786, | |
| "grad_norm": 37.42151641845703, | |
| "learning_rate": 3.5997357992073977e-06, | |
| "loss": 0.358, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.4796425024826216, | |
| "grad_norm": 9.819945335388184, | |
| "learning_rate": 3.533685601056803e-06, | |
| "loss": 0.2852, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 1.4895729890764646, | |
| "grad_norm": 7.768363952636719, | |
| "learning_rate": 3.467635402906209e-06, | |
| "loss": 0.2703, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.4995034756703078, | |
| "grad_norm": 22.80790901184082, | |
| "learning_rate": 3.4015852047556147e-06, | |
| "loss": 0.2419, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 1.509433962264151, | |
| "grad_norm": 14.61506175994873, | |
| "learning_rate": 3.33553500660502e-06, | |
| "loss": 0.2263, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.519364448857994, | |
| "grad_norm": 17.9895076751709, | |
| "learning_rate": 3.2694848084544257e-06, | |
| "loss": 0.3096, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.529294935451837, | |
| "grad_norm": 17.43570899963379, | |
| "learning_rate": 3.2034346103038312e-06, | |
| "loss": 0.2388, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.5392254220456802, | |
| "grad_norm": 24.771135330200195, | |
| "learning_rate": 3.1373844121532363e-06, | |
| "loss": 0.2859, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.5491559086395235, | |
| "grad_norm": 14.774213790893555, | |
| "learning_rate": 3.0713342140026427e-06, | |
| "loss": 0.2478, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.5590863952333665, | |
| "grad_norm": 9.036184310913086, | |
| "learning_rate": 3.0052840158520478e-06, | |
| "loss": 0.2538, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.5690168818272094, | |
| "grad_norm": 15.238768577575684, | |
| "learning_rate": 2.9392338177014533e-06, | |
| "loss": 0.1995, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 7.666260242462158, | |
| "learning_rate": 2.873183619550859e-06, | |
| "loss": 0.2648, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.5888778550148959, | |
| "grad_norm": 14.299159049987793, | |
| "learning_rate": 2.8071334214002643e-06, | |
| "loss": 0.2817, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.5988083416087389, | |
| "grad_norm": 8.902689933776855, | |
| "learning_rate": 2.74108322324967e-06, | |
| "loss": 0.6261, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 1.6087388282025818, | |
| "grad_norm": 11.452303886413574, | |
| "learning_rate": 2.6750330250990754e-06, | |
| "loss": 0.2743, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.618669314796425, | |
| "grad_norm": 8.431117057800293, | |
| "learning_rate": 2.6089828269484813e-06, | |
| "loss": 0.2472, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 1.6285998013902683, | |
| "grad_norm": 13.621753692626953, | |
| "learning_rate": 2.542932628797887e-06, | |
| "loss": 0.3552, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.6385302879841113, | |
| "grad_norm": 28.211891174316406, | |
| "learning_rate": 2.4768824306472924e-06, | |
| "loss": 0.2211, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.6484607745779543, | |
| "grad_norm": 9.634784698486328, | |
| "learning_rate": 2.4108322324966975e-06, | |
| "loss": 0.7658, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.6583912611717975, | |
| "grad_norm": 10.460068702697754, | |
| "learning_rate": 2.3447820343461034e-06, | |
| "loss": 0.2257, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 1.6683217477656405, | |
| "grad_norm": 8.11988639831543, | |
| "learning_rate": 2.278731836195509e-06, | |
| "loss": 0.2626, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.6782522343594835, | |
| "grad_norm": 12.30396842956543, | |
| "learning_rate": 2.2126816380449144e-06, | |
| "loss": 0.2174, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 1.6881827209533267, | |
| "grad_norm": 45.93663787841797, | |
| "learning_rate": 2.14663143989432e-06, | |
| "loss": 0.3282, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.6981132075471699, | |
| "grad_norm": 15.4673490524292, | |
| "learning_rate": 2.0805812417437255e-06, | |
| "loss": 0.2126, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.7080436941410129, | |
| "grad_norm": 13.173709869384766, | |
| "learning_rate": 2.014531043593131e-06, | |
| "loss": 0.2311, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.7179741807348559, | |
| "grad_norm": 17.14373779296875, | |
| "learning_rate": 1.9484808454425365e-06, | |
| "loss": 0.2538, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.727904667328699, | |
| "grad_norm": 13.05432415008545, | |
| "learning_rate": 1.882430647291942e-06, | |
| "loss": 0.3432, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.7378351539225423, | |
| "grad_norm": 6.768345832824707, | |
| "learning_rate": 1.8163804491413476e-06, | |
| "loss": 0.2487, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.7477656405163853, | |
| "grad_norm": 7.953887939453125, | |
| "learning_rate": 1.750330250990753e-06, | |
| "loss": 0.2478, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.7576961271102283, | |
| "grad_norm": 10.380583763122559, | |
| "learning_rate": 1.6842800528401588e-06, | |
| "loss": 0.2797, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.7676266137040715, | |
| "grad_norm": 18.359220504760742, | |
| "learning_rate": 1.6182298546895641e-06, | |
| "loss": 0.3262, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.7775571002979147, | |
| "grad_norm": 39.89724349975586, | |
| "learning_rate": 1.5521796565389696e-06, | |
| "loss": 0.2176, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.7874875868917577, | |
| "grad_norm": 9.103048324584961, | |
| "learning_rate": 1.4861294583883754e-06, | |
| "loss": 0.2202, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.7974180734856007, | |
| "grad_norm": 22.96454429626465, | |
| "learning_rate": 1.420079260237781e-06, | |
| "loss": 0.2874, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.8073485600794439, | |
| "grad_norm": 16.194610595703125, | |
| "learning_rate": 1.3540290620871862e-06, | |
| "loss": 0.168, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.817279046673287, | |
| "grad_norm": 7.695756435394287, | |
| "learning_rate": 1.287978863936592e-06, | |
| "loss": 0.2298, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.82720953326713, | |
| "grad_norm": 13.192455291748047, | |
| "learning_rate": 1.2219286657859975e-06, | |
| "loss": 0.2309, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.837140019860973, | |
| "grad_norm": 16.100967407226562, | |
| "learning_rate": 1.155878467635403e-06, | |
| "loss": 0.2933, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.8470705064548163, | |
| "grad_norm": 4.693732261657715, | |
| "learning_rate": 1.0898282694848085e-06, | |
| "loss": 0.3162, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.8570009930486595, | |
| "grad_norm": 15.807625770568848, | |
| "learning_rate": 1.023778071334214e-06, | |
| "loss": 0.2593, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.8669314796425025, | |
| "grad_norm": 8.402117729187012, | |
| "learning_rate": 9.577278731836198e-07, | |
| "loss": 0.2331, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.8768619662363455, | |
| "grad_norm": 8.832892417907715, | |
| "learning_rate": 8.916776750330252e-07, | |
| "loss": 0.2713, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.8867924528301887, | |
| "grad_norm": 20.126842498779297, | |
| "learning_rate": 8.256274768824307e-07, | |
| "loss": 0.3002, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.896722939424032, | |
| "grad_norm": 7.1585845947265625, | |
| "learning_rate": 7.595772787318363e-07, | |
| "loss": 0.2557, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.906653426017875, | |
| "grad_norm": 9.247796058654785, | |
| "learning_rate": 6.935270805812417e-07, | |
| "loss": 0.2147, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.916583912611718, | |
| "grad_norm": 19.770483016967773, | |
| "learning_rate": 6.274768824306473e-07, | |
| "loss": 0.3213, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 1.9265143992055611, | |
| "grad_norm": 18.85097885131836, | |
| "learning_rate": 5.614266842800529e-07, | |
| "loss": 0.3408, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.9364448857994043, | |
| "grad_norm": 24.601226806640625, | |
| "learning_rate": 4.953764861294584e-07, | |
| "loss": 0.2798, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.9463753723932473, | |
| "grad_norm": 13.348860740661621, | |
| "learning_rate": 4.29326287978864e-07, | |
| "loss": 0.2631, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.9563058589870903, | |
| "grad_norm": 9.122193336486816, | |
| "learning_rate": 3.6327608982826953e-07, | |
| "loss": 0.2529, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 1.9662363455809335, | |
| "grad_norm": 20.501909255981445, | |
| "learning_rate": 2.9722589167767505e-07, | |
| "loss": 0.2518, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.9761668321747765, | |
| "grad_norm": 9.407881736755371, | |
| "learning_rate": 2.311756935270806e-07, | |
| "loss": 0.2078, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 1.9860973187686195, | |
| "grad_norm": 22.076322555541992, | |
| "learning_rate": 1.6512549537648615e-07, | |
| "loss": 0.2104, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.9960278053624627, | |
| "grad_norm": 18.52741241455078, | |
| "learning_rate": 9.907529722589168e-08, | |
| "loss": 0.2356, | |
| "step": 2010 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2014, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.4898226126848e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |