| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.13458950201884254, | |
| "eval_steps": 4, | |
| "global_step": 50, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0026917900403768506, | |
| "grad_norm": 2.0982859134674072, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7376, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0026917900403768506, | |
| "eval_loss": 0.7193137407302856, | |
| "eval_runtime": 10.1138, | |
| "eval_samples_per_second": 7.811, | |
| "eval_steps_per_second": 7.811, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005383580080753701, | |
| "grad_norm": 2.031587839126587, | |
| "learning_rate": 4e-05, | |
| "loss": 0.7275, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.008075370121130552, | |
| "grad_norm": 2.1216890811920166, | |
| "learning_rate": 6e-05, | |
| "loss": 0.6912, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.010767160161507403, | |
| "grad_norm": 1.755471110343933, | |
| "learning_rate": 8e-05, | |
| "loss": 0.6835, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.010767160161507403, | |
| "eval_loss": 0.6272252202033997, | |
| "eval_runtime": 10.0256, | |
| "eval_samples_per_second": 7.88, | |
| "eval_steps_per_second": 7.88, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.013458950201884253, | |
| "grad_norm": 1.9561946392059326, | |
| "learning_rate": 0.0001, | |
| "loss": 0.7364, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.016150740242261104, | |
| "grad_norm": 1.8313077688217163, | |
| "learning_rate": 0.00012, | |
| "loss": 0.6385, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.018842530282637954, | |
| "grad_norm": 1.4374486207962036, | |
| "learning_rate": 0.00014, | |
| "loss": 0.4741, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.021534320323014805, | |
| "grad_norm": 1.279863715171814, | |
| "learning_rate": 0.00016, | |
| "loss": 0.3742, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.021534320323014805, | |
| "eval_loss": 0.3805864155292511, | |
| "eval_runtime": 10.1128, | |
| "eval_samples_per_second": 7.812, | |
| "eval_steps_per_second": 7.812, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.024226110363391656, | |
| "grad_norm": 1.7297841310501099, | |
| "learning_rate": 0.00018, | |
| "loss": 0.3654, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.026917900403768506, | |
| "grad_norm": 1.6552913188934326, | |
| "learning_rate": 0.0002, | |
| "loss": 0.3738, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.029609690444145357, | |
| "grad_norm": 1.50761079788208, | |
| "learning_rate": 0.0001996917333733128, | |
| "loss": 0.4244, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.03230148048452221, | |
| "grad_norm": 1.431064248085022, | |
| "learning_rate": 0.00019876883405951377, | |
| "loss": 0.2788, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03230148048452221, | |
| "eval_loss": 0.32530486583709717, | |
| "eval_runtime": 10.0708, | |
| "eval_samples_per_second": 7.844, | |
| "eval_steps_per_second": 7.844, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.034993270524899055, | |
| "grad_norm": 1.4582064151763916, | |
| "learning_rate": 0.00019723699203976766, | |
| "loss": 0.3575, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03768506056527591, | |
| "grad_norm": 1.6408145427703857, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 0.4609, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.040376850605652756, | |
| "grad_norm": 1.7364938259124756, | |
| "learning_rate": 0.0001923879532511287, | |
| "loss": 0.4607, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04306864064602961, | |
| "grad_norm": 2.5827343463897705, | |
| "learning_rate": 0.0001891006524188368, | |
| "loss": 0.5915, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04306864064602961, | |
| "eval_loss": 0.29014351963996887, | |
| "eval_runtime": 10.1162, | |
| "eval_samples_per_second": 7.809, | |
| "eval_steps_per_second": 7.809, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04576043068640646, | |
| "grad_norm": 1.0929858684539795, | |
| "learning_rate": 0.00018526401643540922, | |
| "loss": 0.2242, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04845222072678331, | |
| "grad_norm": 1.4121438264846802, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 0.2983, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.05114401076716016, | |
| "grad_norm": 0.9734262824058533, | |
| "learning_rate": 0.0001760405965600031, | |
| "loss": 0.3084, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.05383580080753701, | |
| "grad_norm": 0.9649080038070679, | |
| "learning_rate": 0.00017071067811865476, | |
| "loss": 0.2583, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05383580080753701, | |
| "eval_loss": 0.28261926770210266, | |
| "eval_runtime": 10.3045, | |
| "eval_samples_per_second": 7.667, | |
| "eval_steps_per_second": 7.667, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05652759084791386, | |
| "grad_norm": 1.8083579540252686, | |
| "learning_rate": 0.00016494480483301836, | |
| "loss": 0.3447, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.059219380888290714, | |
| "grad_norm": 1.2060580253601074, | |
| "learning_rate": 0.00015877852522924732, | |
| "loss": 0.1974, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.06191117092866756, | |
| "grad_norm": 0.7741371989250183, | |
| "learning_rate": 0.0001522498564715949, | |
| "loss": 0.1914, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.06460296096904442, | |
| "grad_norm": 1.2901431322097778, | |
| "learning_rate": 0.00014539904997395468, | |
| "loss": 0.2147, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06460296096904442, | |
| "eval_loss": 0.28900590538978577, | |
| "eval_runtime": 10.1131, | |
| "eval_samples_per_second": 7.812, | |
| "eval_steps_per_second": 7.812, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06729475100942127, | |
| "grad_norm": 9.104133605957031, | |
| "learning_rate": 0.000138268343236509, | |
| "loss": 0.6002, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06998654104979811, | |
| "grad_norm": 1.2012132406234741, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 0.2677, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.07267833109017496, | |
| "grad_norm": 0.8289370536804199, | |
| "learning_rate": 0.00012334453638559057, | |
| "loss": 0.1694, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.07537012113055182, | |
| "grad_norm": 1.3279316425323486, | |
| "learning_rate": 0.0001156434465040231, | |
| "loss": 0.2077, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.07537012113055182, | |
| "eval_loss": 0.2843972444534302, | |
| "eval_runtime": 10.1018, | |
| "eval_samples_per_second": 7.82, | |
| "eval_steps_per_second": 7.82, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.07806191117092867, | |
| "grad_norm": 0.8260064721107483, | |
| "learning_rate": 0.0001078459095727845, | |
| "loss": 0.2503, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.08075370121130551, | |
| "grad_norm": 1.2920253276824951, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3839, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08344549125168237, | |
| "grad_norm": 1.1818400621414185, | |
| "learning_rate": 9.215409042721552e-05, | |
| "loss": 0.2586, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.08613728129205922, | |
| "grad_norm": 1.3729400634765625, | |
| "learning_rate": 8.435655349597689e-05, | |
| "loss": 0.4248, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.08613728129205922, | |
| "eval_loss": 0.27306219935417175, | |
| "eval_runtime": 10.1319, | |
| "eval_samples_per_second": 7.797, | |
| "eval_steps_per_second": 7.797, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.08882907133243607, | |
| "grad_norm": 0.9806796312332153, | |
| "learning_rate": 7.66554636144095e-05, | |
| "loss": 0.3062, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.09152086137281291, | |
| "grad_norm": 1.1644279956817627, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 0.3511, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.09421265141318977, | |
| "grad_norm": 0.9300962090492249, | |
| "learning_rate": 6.173165676349103e-05, | |
| "loss": 0.2276, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.09690444145356662, | |
| "grad_norm": 0.9256501793861389, | |
| "learning_rate": 5.4600950026045326e-05, | |
| "loss": 0.3099, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.09690444145356662, | |
| "eval_loss": 0.27339035272598267, | |
| "eval_runtime": 10.1945, | |
| "eval_samples_per_second": 7.749, | |
| "eval_steps_per_second": 7.749, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.09959623149394348, | |
| "grad_norm": 1.368626356124878, | |
| "learning_rate": 4.7750143528405126e-05, | |
| "loss": 0.4373, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.10228802153432032, | |
| "grad_norm": 0.7500585913658142, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 0.2919, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.10497981157469717, | |
| "grad_norm": 0.903423011302948, | |
| "learning_rate": 3.5055195166981645e-05, | |
| "loss": 0.3673, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.10767160161507403, | |
| "grad_norm": 0.7334548830986023, | |
| "learning_rate": 2.9289321881345254e-05, | |
| "loss": 0.2465, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10767160161507403, | |
| "eval_loss": 0.27042919397354126, | |
| "eval_runtime": 10.1181, | |
| "eval_samples_per_second": 7.808, | |
| "eval_steps_per_second": 7.808, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.11036339165545088, | |
| "grad_norm": 0.9937801957130432, | |
| "learning_rate": 2.3959403439996907e-05, | |
| "loss": 0.2871, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.11305518169582772, | |
| "grad_norm": 1.202613115310669, | |
| "learning_rate": 1.9098300562505266e-05, | |
| "loss": 0.3097, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.11574697173620457, | |
| "grad_norm": 0.7323685884475708, | |
| "learning_rate": 1.4735983564590783e-05, | |
| "loss": 0.2534, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.11843876177658143, | |
| "grad_norm": 1.1680642366409302, | |
| "learning_rate": 1.0899347581163221e-05, | |
| "loss": 0.2693, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.11843876177658143, | |
| "eval_loss": 0.26846635341644287, | |
| "eval_runtime": 10.1072, | |
| "eval_samples_per_second": 7.816, | |
| "eval_steps_per_second": 7.816, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.12113055181695828, | |
| "grad_norm": 0.6449961066246033, | |
| "learning_rate": 7.612046748871327e-06, | |
| "loss": 0.1167, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.12382234185733512, | |
| "grad_norm": 1.3696563243865967, | |
| "learning_rate": 4.8943483704846475e-06, | |
| "loss": 0.3693, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.126514131897712, | |
| "grad_norm": 1.968651294708252, | |
| "learning_rate": 2.7630079602323442e-06, | |
| "loss": 0.4617, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.12920592193808883, | |
| "grad_norm": 0.8639482259750366, | |
| "learning_rate": 1.231165940486234e-06, | |
| "loss": 0.231, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.12920592193808883, | |
| "eval_loss": 0.26753100752830505, | |
| "eval_runtime": 10.1196, | |
| "eval_samples_per_second": 7.807, | |
| "eval_steps_per_second": 7.807, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.13189771197846567, | |
| "grad_norm": 1.1134506464004517, | |
| "learning_rate": 3.0826662668720364e-07, | |
| "loss": 0.2964, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.13458950201884254, | |
| "grad_norm": 0.8477309346199036, | |
| "learning_rate": 0.0, | |
| "loss": 0.2486, | |
| "step": 50 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 50, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 4, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.021642847702221e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |