| { | |
| "best_metric": 240.85956416464893, | |
| "best_model_checkpoint": "./checkpoint-100", | |
| "epoch": 57.142857142857146, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 2.9076923076923076e-07, | |
| "loss": 2.8018, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 7.14, | |
| "learning_rate": 2.7153846153846153e-07, | |
| "loss": 2.4337, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 10.71, | |
| "learning_rate": 2.523076923076923e-07, | |
| "loss": 2.1631, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 14.29, | |
| "learning_rate": 2.3307692307692306e-07, | |
| "loss": 1.9754, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 14.29, | |
| "eval_loss": 1.9261106252670288, | |
| "eval_runtime": 721.539, | |
| "eval_samples_per_second": 0.71, | |
| "eval_steps_per_second": 0.044, | |
| "eval_wer": 240.85956416464893, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 17.86, | |
| "learning_rate": 2.1384615384615385e-07, | |
| "loss": 1.8067, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 21.43, | |
| "learning_rate": 1.9461538461538462e-07, | |
| "loss": 1.7049, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 1.7538461538461539e-07, | |
| "loss": 1.606, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 28.57, | |
| "learning_rate": 1.5615384615384615e-07, | |
| "loss": 1.5323, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 28.57, | |
| "eval_loss": 1.5718045234680176, | |
| "eval_runtime": 653.16, | |
| "eval_samples_per_second": 0.784, | |
| "eval_steps_per_second": 0.049, | |
| "eval_wer": 168.5608353510896, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 32.14, | |
| "learning_rate": 1.3692307692307692e-07, | |
| "loss": 1.4644, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 35.71, | |
| "learning_rate": 1.1769230769230768e-07, | |
| "loss": 1.4134, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 39.29, | |
| "learning_rate": 9.846153846153846e-08, | |
| "loss": 1.3706, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 42.86, | |
| "learning_rate": 7.923076923076923e-08, | |
| "loss": 1.338, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 42.86, | |
| "eval_loss": 1.4249473810195923, | |
| "eval_runtime": 548.9427, | |
| "eval_samples_per_second": 0.933, | |
| "eval_steps_per_second": 0.058, | |
| "eval_wer": 96.64800242130751, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 46.43, | |
| "learning_rate": 6e-08, | |
| "loss": 1.3112, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 4.076923076923077e-08, | |
| "loss": 1.3008, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 53.57, | |
| "learning_rate": 2.1538461538461537e-08, | |
| "loss": 1.2859, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 57.14, | |
| "learning_rate": 2.3076923076923076e-09, | |
| "loss": 1.282, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 57.14, | |
| "eval_loss": 1.3860349655151367, | |
| "eval_runtime": 506.783, | |
| "eval_samples_per_second": 1.01, | |
| "eval_steps_per_second": 0.063, | |
| "eval_wer": 89.14951573849879, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 57.14, | |
| "step": 400, | |
| "total_flos": 7.10814699528192e+18, | |
| "train_loss": 1.6743954944610595, | |
| "train_runtime": 3832.5455, | |
| "train_samples_per_second": 6.68, | |
| "train_steps_per_second": 0.104 | |
| } | |
| ], | |
| "max_steps": 400, | |
| "num_train_epochs": 58, | |
| "total_flos": 7.10814699528192e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |