| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 48.57142857142857, | |
| "eval_steps": 500, | |
| "global_step": 3740, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 4405.81884765625, | |
| "learning_rate": 3.436363636363636e-07, | |
| "loss": 1357.5645, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.4285714285714286, | |
| "grad_norm": 5894.37841796875, | |
| "learning_rate": 6.936363636363637e-07, | |
| "loss": 1290.9688, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.142857142857143, | |
| "grad_norm": 3667.852294921875, | |
| "learning_rate": 1.0436363636363635e-06, | |
| "loss": 1140.9515, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.857142857142857, | |
| "grad_norm": 3578.294677734375, | |
| "learning_rate": 1.3936363636363637e-06, | |
| "loss": 942.618, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 2097.637939453125, | |
| "learning_rate": 1.7436363636363636e-06, | |
| "loss": 721.2856, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 4.285714285714286, | |
| "grad_norm": 1840.612548828125, | |
| "learning_rate": 2.0936363636363636e-06, | |
| "loss": 525.8542, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1230.875, | |
| "learning_rate": 2.4436363636363636e-06, | |
| "loss": 397.7722, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 1280.699951171875, | |
| "learning_rate": 2.7936363636363637e-06, | |
| "loss": 313.8322, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 6.428571428571429, | |
| "grad_norm": 4265.28759765625, | |
| "learning_rate": 3.1436363636363637e-06, | |
| "loss": 254.1449, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "grad_norm": 967.1757202148438, | |
| "learning_rate": 3.4936363636363633e-06, | |
| "loss": 218.4529, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 7.857142857142857, | |
| "grad_norm": 765.2302856445312, | |
| "learning_rate": 3.843636363636364e-06, | |
| "loss": 197.915, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 8.571428571428571, | |
| "grad_norm": 700.331787109375, | |
| "learning_rate": 4.193636363636364e-06, | |
| "loss": 170.9019, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 9.285714285714286, | |
| "grad_norm": 613.8758544921875, | |
| "learning_rate": 4.543636363636363e-06, | |
| "loss": 154.6195, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 3405.25732421875, | |
| "learning_rate": 4.893636363636364e-06, | |
| "loss": 136.8394, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 10.714285714285714, | |
| "grad_norm": 1219.8994140625, | |
| "learning_rate": 5.243636363636364e-06, | |
| "loss": 116.0246, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 11.428571428571429, | |
| "grad_norm": 306.2723083496094, | |
| "learning_rate": 5.593636363636363e-06, | |
| "loss": 104.2055, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 12.142857142857142, | |
| "grad_norm": 288.2547302246094, | |
| "learning_rate": 5.943636363636364e-06, | |
| "loss": 88.9812, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 12.857142857142858, | |
| "grad_norm": 1016.5924682617188, | |
| "learning_rate": 6.293636363636363e-06, | |
| "loss": 79.9017, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 13.571428571428571, | |
| "grad_norm": 239.49191284179688, | |
| "learning_rate": 6.643636363636363e-06, | |
| "loss": 73.6932, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 14.285714285714286, | |
| "grad_norm": 193.92002868652344, | |
| "learning_rate": 6.993636363636364e-06, | |
| "loss": 62.2018, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 179.0568084716797, | |
| "learning_rate": 7.343636363636363e-06, | |
| "loss": 54.0213, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 15.714285714285714, | |
| "grad_norm": 166.02944946289062, | |
| "learning_rate": 7.693636363636364e-06, | |
| "loss": 48.1955, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 16.428571428571427, | |
| "grad_norm": 126.53916931152344, | |
| "learning_rate": 8.043636363636364e-06, | |
| "loss": 43.5431, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 17.142857142857142, | |
| "grad_norm": 145.61166381835938, | |
| "learning_rate": 8.393636363636363e-06, | |
| "loss": 41.4399, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 17.857142857142858, | |
| "grad_norm": 122.2297134399414, | |
| "learning_rate": 8.743636363636363e-06, | |
| "loss": 35.2278, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 18.571428571428573, | |
| "grad_norm": 117.88919067382812, | |
| "learning_rate": 9.093636363636363e-06, | |
| "loss": 31.2827, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 19.285714285714285, | |
| "grad_norm": 88.52986907958984, | |
| "learning_rate": 9.443636363636364e-06, | |
| "loss": 28.8076, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 88.49090576171875, | |
| "learning_rate": 9.793636363636364e-06, | |
| "loss": 28.232, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 20.714285714285715, | |
| "grad_norm": 114.45001983642578, | |
| "learning_rate": 1.0143636363636363e-05, | |
| "loss": 26.0885, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 21.428571428571427, | |
| "grad_norm": 74.16987609863281, | |
| "learning_rate": 1.0493636363636363e-05, | |
| "loss": 24.2689, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 22.142857142857142, | |
| "grad_norm": 82.25133514404297, | |
| "learning_rate": 1.0843636363636363e-05, | |
| "loss": 22.6963, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 22.857142857142858, | |
| "grad_norm": 74.76679229736328, | |
| "learning_rate": 1.1193636363636363e-05, | |
| "loss": 21.3257, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 23.571428571428573, | |
| "grad_norm": 68.91163635253906, | |
| "learning_rate": 1.1543636363636365e-05, | |
| "loss": 20.2199, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 24.285714285714285, | |
| "grad_norm": 59.86214828491211, | |
| "learning_rate": 1.1893636363636363e-05, | |
| "loss": 19.3769, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 76.9339370727539, | |
| "learning_rate": 1.2243636363636363e-05, | |
| "loss": 18.0471, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 25.714285714285715, | |
| "grad_norm": 66.16152954101562, | |
| "learning_rate": 1.2593636363636363e-05, | |
| "loss": 17.496, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 26.428571428571427, | |
| "grad_norm": 58.77778244018555, | |
| "learning_rate": 1.2943636363636363e-05, | |
| "loss": 16.9468, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 27.142857142857142, | |
| "grad_norm": 71.95783233642578, | |
| "learning_rate": 1.3293636363636363e-05, | |
| "loss": 16.3026, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 27.857142857142858, | |
| "grad_norm": 131.7191619873047, | |
| "learning_rate": 1.3643636363636363e-05, | |
| "loss": 16.008, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 28.571428571428573, | |
| "grad_norm": 81.78388977050781, | |
| "learning_rate": 1.3993636363636363e-05, | |
| "loss": 15.4616, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 29.285714285714285, | |
| "grad_norm": 60.10112380981445, | |
| "learning_rate": 1.3628501228501228e-05, | |
| "loss": 15.195, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 77.59001922607422, | |
| "learning_rate": 1.325012285012285e-05, | |
| "loss": 14.7557, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 30.714285714285715, | |
| "grad_norm": 77.3856201171875, | |
| "learning_rate": 1.287174447174447e-05, | |
| "loss": 14.2851, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 31.428571428571427, | |
| "grad_norm": 73.75492095947266, | |
| "learning_rate": 1.2493366093366094e-05, | |
| "loss": 13.867, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 32.142857142857146, | |
| "grad_norm": 78.76831817626953, | |
| "learning_rate": 1.2114987714987713e-05, | |
| "loss": 13.6677, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 32.857142857142854, | |
| "grad_norm": 92.99346160888672, | |
| "learning_rate": 1.1736609336609336e-05, | |
| "loss": 13.1851, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 33.57142857142857, | |
| "grad_norm": 97.57978820800781, | |
| "learning_rate": 1.1358230958230958e-05, | |
| "loss": 13.5215, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 34.285714285714285, | |
| "grad_norm": 63.118141174316406, | |
| "learning_rate": 1.0979852579852579e-05, | |
| "loss": 12.8525, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "grad_norm": 63.68766403198242, | |
| "learning_rate": 1.06014742014742e-05, | |
| "loss": 12.441, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 35.714285714285715, | |
| "grad_norm": 70.33533477783203, | |
| "learning_rate": 1.0223095823095823e-05, | |
| "loss": 12.2331, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 36.42857142857143, | |
| "grad_norm": 64.16197967529297, | |
| "learning_rate": 9.844717444717445e-06, | |
| "loss": 11.7547, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 37.142857142857146, | |
| "grad_norm": 85.83612823486328, | |
| "learning_rate": 9.466339066339066e-06, | |
| "loss": 11.8633, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 37.857142857142854, | |
| "grad_norm": 89.40387725830078, | |
| "learning_rate": 9.087960687960689e-06, | |
| "loss": 11.5795, | |
| "step": 2915 | |
| }, | |
| { | |
| "epoch": 38.57142857142857, | |
| "grad_norm": 76.38651275634766, | |
| "learning_rate": 8.709582309582309e-06, | |
| "loss": 11.334, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 39.285714285714285, | |
| "grad_norm": 110.11483764648438, | |
| "learning_rate": 8.331203931203932e-06, | |
| "loss": 10.805, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 60.31315612792969, | |
| "learning_rate": 7.952825552825553e-06, | |
| "loss": 10.9056, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 40.714285714285715, | |
| "grad_norm": 107.96589660644531, | |
| "learning_rate": 7.574447174447175e-06, | |
| "loss": 10.5828, | |
| "step": 3135 | |
| }, | |
| { | |
| "epoch": 41.42857142857143, | |
| "grad_norm": 74.8004379272461, | |
| "learning_rate": 7.1960687960687955e-06, | |
| "loss": 10.2818, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 42.142857142857146, | |
| "grad_norm": 54.093475341796875, | |
| "learning_rate": 6.817690417690418e-06, | |
| "loss": 10.0785, | |
| "step": 3245 | |
| }, | |
| { | |
| "epoch": 42.857142857142854, | |
| "grad_norm": 115.28564453125, | |
| "learning_rate": 6.439312039312039e-06, | |
| "loss": 9.8794, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 43.57142857142857, | |
| "grad_norm": 52.57551956176758, | |
| "learning_rate": 6.06093366093366e-06, | |
| "loss": 9.7072, | |
| "step": 3355 | |
| }, | |
| { | |
| "epoch": 44.285714285714285, | |
| "grad_norm": 51.994075775146484, | |
| "learning_rate": 5.682555282555282e-06, | |
| "loss": 9.6015, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 45.0, | |
| "grad_norm": 73.08104705810547, | |
| "learning_rate": 5.304176904176904e-06, | |
| "loss": 9.305, | |
| "step": 3465 | |
| }, | |
| { | |
| "epoch": 45.714285714285715, | |
| "grad_norm": 75.5807876586914, | |
| "learning_rate": 4.925798525798525e-06, | |
| "loss": 9.2281, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 46.42857142857143, | |
| "grad_norm": 53.319637298583984, | |
| "learning_rate": 4.547420147420147e-06, | |
| "loss": 8.996, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 47.142857142857146, | |
| "grad_norm": 72.28289794921875, | |
| "learning_rate": 4.1690417690417685e-06, | |
| "loss": 9.0072, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 47.857142857142854, | |
| "grad_norm": 66.79710388183594, | |
| "learning_rate": 3.7906633906633902e-06, | |
| "loss": 8.7311, | |
| "step": 3685 | |
| }, | |
| { | |
| "epoch": 48.57142857142857, | |
| "grad_norm": 83.55674743652344, | |
| "learning_rate": 3.4122850122850124e-06, | |
| "loss": 8.5959, | |
| "step": 3740 | |
| } | |
| ], | |
| "logging_steps": 55, | |
| "max_steps": 4235, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 55, | |
| "save_steps": 220, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.8876427477621146e+19, | |
| "train_batch_size": 28, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |