| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 27.586206896551722, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.3793103448275863, | |
| "grad_norm": 0.300048828125, | |
| "learning_rate": 0.00019876883405951377, | |
| "loss": 1.0738, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.7586206896551726, | |
| "grad_norm": 0.272705078125, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 0.8853, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.137931034482759, | |
| "grad_norm": 0.213134765625, | |
| "learning_rate": 0.0001891006524188368, | |
| "loss": 0.7883, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.517241379310345, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 0.7353, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 6.896551724137931, | |
| "grad_norm": 0.132568359375, | |
| "learning_rate": 0.00017071067811865476, | |
| "loss": 0.7072, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 8.275862068965518, | |
| "grad_norm": 0.1220703125, | |
| "learning_rate": 0.00015877852522924732, | |
| "loss": 0.6986, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 9.655172413793103, | |
| "grad_norm": 0.1285400390625, | |
| "learning_rate": 0.00014539904997395468, | |
| "loss": 0.6873, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 11.03448275862069, | |
| "grad_norm": 0.1348876953125, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 0.6762, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 12.413793103448276, | |
| "grad_norm": 0.1441650390625, | |
| "learning_rate": 0.0001156434465040231, | |
| "loss": 0.6668, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 13.793103448275861, | |
| "grad_norm": 0.1851806640625, | |
| "learning_rate": 0.0001, | |
| "loss": 0.6648, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 15.172413793103448, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 8.435655349597689e-05, | |
| "loss": 0.6564, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 16.551724137931036, | |
| "grad_norm": 0.1644287109375, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 0.6511, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 17.93103448275862, | |
| "grad_norm": 0.1719970703125, | |
| "learning_rate": 5.4600950026045326e-05, | |
| "loss": 0.6476, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 19.310344827586206, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 0.6387, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 20.689655172413794, | |
| "grad_norm": 0.170166015625, | |
| "learning_rate": 2.9289321881345254e-05, | |
| "loss": 0.6438, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 22.06896551724138, | |
| "grad_norm": 0.185302734375, | |
| "learning_rate": 1.9098300562505266e-05, | |
| "loss": 0.6341, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 23.448275862068964, | |
| "grad_norm": 0.1927490234375, | |
| "learning_rate": 1.0899347581163221e-05, | |
| "loss": 0.638, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 24.82758620689655, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 4.8943483704846475e-06, | |
| "loss": 0.6323, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 26.20689655172414, | |
| "grad_norm": 0.183837890625, | |
| "learning_rate": 1.231165940486234e-06, | |
| "loss": 0.6338, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 27.586206896551722, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 0.0, | |
| "loss": 0.6344, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 27.586206896551722, | |
| "step": 200, | |
| "total_flos": 4.07672658591744e+16, | |
| "train_loss": 0.6996826219558716, | |
| "train_runtime": 575.4411, | |
| "train_samples_per_second": 11.122, | |
| "train_steps_per_second": 0.348 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 29, | |
| "save_steps": 500, | |
| "total_flos": 4.07672658591744e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |