| { | |
| "best_metric": 0.5421165227890015, | |
| "best_model_checkpoint": "./20230701_models/checkpoint-1099", | |
| "epoch": 19.94540491355778, | |
| "global_step": 5480, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_cer": 0.10819683151455398, | |
| "eval_loss": 0.6883996725082397, | |
| "eval_runtime": 259.3566, | |
| "eval_samples_per_second": 8.143, | |
| "eval_steps_per_second": 0.509, | |
| "eval_wer": 0.26681723843242583, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 3.635766423357665e-05, | |
| "loss": 0.9397, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_cer": 0.08392901158491496, | |
| "eval_loss": 0.5709623694419861, | |
| "eval_runtime": 251.7707, | |
| "eval_samples_per_second": 8.389, | |
| "eval_steps_per_second": 0.524, | |
| "eval_wer": 0.2248809669149066, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_cer": 0.07773321083648912, | |
| "eval_loss": 0.5516401529312134, | |
| "eval_runtime": 253.1328, | |
| "eval_samples_per_second": 8.343, | |
| "eval_steps_per_second": 0.521, | |
| "eval_wer": 0.20919301672567453, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "learning_rate": 3.270802919708029e-05, | |
| "loss": 0.2569, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_cer": 0.07259058417549914, | |
| "eval_loss": 0.5421165227890015, | |
| "eval_runtime": 253.562, | |
| "eval_samples_per_second": 8.329, | |
| "eval_steps_per_second": 0.521, | |
| "eval_wer": 0.20198998901233062, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_cer": 0.07377820601877787, | |
| "eval_loss": 0.5494938492774963, | |
| "eval_runtime": 254.0497, | |
| "eval_samples_per_second": 8.313, | |
| "eval_steps_per_second": 0.52, | |
| "eval_wer": 0.2014406055426688, | |
| "step": 1373 | |
| }, | |
| { | |
| "epoch": 5.46, | |
| "learning_rate": 2.9058394160583945e-05, | |
| "loss": 0.109, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_cer": 0.07154861406771686, | |
| "eval_loss": 0.5665440559387207, | |
| "eval_runtime": 253.9146, | |
| "eval_samples_per_second": 8.318, | |
| "eval_steps_per_second": 0.52, | |
| "eval_wer": 0.1971676230008546, | |
| "step": 1648 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_cer": 0.07226566876554552, | |
| "eval_loss": 0.5868140459060669, | |
| "eval_runtime": 253.2966, | |
| "eval_samples_per_second": 8.338, | |
| "eval_steps_per_second": 0.521, | |
| "eval_wer": 0.1968624099621536, | |
| "step": 1923 | |
| }, | |
| { | |
| "epoch": 7.28, | |
| "learning_rate": 2.5408759124087593e-05, | |
| "loss": 0.0481, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_cer": 0.07029376834651668, | |
| "eval_loss": 0.5920885801315308, | |
| "eval_runtime": 252.718, | |
| "eval_samples_per_second": 8.357, | |
| "eval_steps_per_second": 0.522, | |
| "eval_wer": 0.19570260041508974, | |
| "step": 2198 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_cer": 0.07013691263136666, | |
| "eval_loss": 0.5932603478431702, | |
| "eval_runtime": 254.5349, | |
| "eval_samples_per_second": 8.297, | |
| "eval_steps_per_second": 0.519, | |
| "eval_wer": 0.19374923696740326, | |
| "step": 2472 | |
| }, | |
| { | |
| "epoch": 9.1, | |
| "learning_rate": 2.1759124087591242e-05, | |
| "loss": 0.022, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_cer": 0.06856835547986645, | |
| "eval_loss": 0.6067739725112915, | |
| "eval_runtime": 253.5231, | |
| "eval_samples_per_second": 8.331, | |
| "eval_steps_per_second": 0.521, | |
| "eval_wer": 0.1925894274203394, | |
| "step": 2747 | |
| }, | |
| { | |
| "epoch": 10.92, | |
| "learning_rate": 1.8116788321167883e-05, | |
| "loss": 0.0108, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_cer": 0.06744795751450916, | |
| "eval_loss": 0.6042998433113098, | |
| "eval_runtime": 253.2695, | |
| "eval_samples_per_second": 8.339, | |
| "eval_steps_per_second": 0.521, | |
| "eval_wer": 0.1902087657184715, | |
| "step": 3022 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_cer": 0.06674210679633406, | |
| "eval_loss": 0.6130247116088867, | |
| "eval_runtime": 253.5585, | |
| "eval_samples_per_second": 8.329, | |
| "eval_steps_per_second": 0.521, | |
| "eval_wer": 0.18856061530948603, | |
| "step": 3297 | |
| }, | |
| { | |
| "epoch": 12.74, | |
| "learning_rate": 1.4467153284671533e-05, | |
| "loss": 0.0052, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_cer": 0.0659802361798911, | |
| "eval_loss": 0.6161753535270691, | |
| "eval_runtime": 254.095, | |
| "eval_samples_per_second": 8.312, | |
| "eval_steps_per_second": 0.519, | |
| "eval_wer": 0.1861799536076181, | |
| "step": 3571 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_cer": 0.06466937056042306, | |
| "eval_loss": 0.6252880692481995, | |
| "eval_runtime": 254.0952, | |
| "eval_samples_per_second": 8.312, | |
| "eval_steps_per_second": 0.519, | |
| "eval_wer": 0.18471493102185324, | |
| "step": 3846 | |
| }, | |
| { | |
| "epoch": 14.56, | |
| "learning_rate": 1.0817518248175182e-05, | |
| "loss": 0.0028, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_cer": 0.06428843525220158, | |
| "eval_loss": 0.6240524649620056, | |
| "eval_runtime": 253.6298, | |
| "eval_samples_per_second": 8.327, | |
| "eval_steps_per_second": 0.52, | |
| "eval_wer": 0.18270052496642655, | |
| "step": 4121 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_cer": 0.06460214668250162, | |
| "eval_loss": 0.6327393054962158, | |
| "eval_runtime": 254.0581, | |
| "eval_samples_per_second": 8.313, | |
| "eval_steps_per_second": 0.52, | |
| "eval_wer": 0.18428763276767182, | |
| "step": 4396 | |
| }, | |
| { | |
| "epoch": 16.38, | |
| "learning_rate": 7.167883211678832e-06, | |
| "loss": 0.0016, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_cer": 0.0632464651444193, | |
| "eval_loss": 0.6320570707321167, | |
| "eval_runtime": 253.728, | |
| "eval_samples_per_second": 8.324, | |
| "eval_steps_per_second": 0.52, | |
| "eval_wer": 0.18190697106580392, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_cer": 0.06377305218813722, | |
| "eval_loss": 0.6347479224205017, | |
| "eval_runtime": 254.3801, | |
| "eval_samples_per_second": 8.303, | |
| "eval_steps_per_second": 0.519, | |
| "eval_wer": 0.18239531192772554, | |
| "step": 4945 | |
| }, | |
| { | |
| "epoch": 18.2, | |
| "learning_rate": 3.5182481751824822e-06, | |
| "loss": 0.0011, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_cer": 0.06367221637125507, | |
| "eval_loss": 0.6383982300758362, | |
| "eval_runtime": 254.2837, | |
| "eval_samples_per_second": 8.306, | |
| "eval_steps_per_second": 0.519, | |
| "eval_wer": 0.1815407154193627, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 19.95, | |
| "eval_cer": 0.06354897259506577, | |
| "eval_loss": 0.6386234164237976, | |
| "eval_runtime": 253.9317, | |
| "eval_samples_per_second": 8.317, | |
| "eval_steps_per_second": 0.52, | |
| "eval_wer": 0.18233426931998534, | |
| "step": 5480 | |
| } | |
| ], | |
| "max_steps": 5480, | |
| "num_train_epochs": 20, | |
| "total_flos": 9.42327114459343e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |