| { | |
| "best_metric": 35.49900739938639, | |
| "best_model_checkpoint": "./checkpoint-400", | |
| "epoch": 26.025, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 5.2500000000000006e-06, | |
| "loss": 1.9832, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "learning_rate": 9.833333333333333e-06, | |
| "loss": 1.0676, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "eval_loss": 0.6486711502075195, | |
| "eval_runtime": 75.7517, | |
| "eval_samples_per_second": 4.066, | |
| "eval_steps_per_second": 0.132, | |
| "eval_wer": 62.53383865728208, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "learning_rate": 9.13888888888889e-06, | |
| "loss": 0.4189, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "learning_rate": 8.444444444444446e-06, | |
| "loss": 0.2252, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "eval_loss": 0.3487340211868286, | |
| "eval_runtime": 68.3563, | |
| "eval_samples_per_second": 4.506, | |
| "eval_steps_per_second": 0.146, | |
| "eval_wer": 36.49160801299404, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "learning_rate": 7.75e-06, | |
| "loss": 0.1387, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 9.04, | |
| "learning_rate": 7.055555555555557e-06, | |
| "loss": 0.0787, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 9.04, | |
| "eval_loss": 0.39337122440338135, | |
| "eval_runtime": 65.9218, | |
| "eval_samples_per_second": 4.672, | |
| "eval_steps_per_second": 0.152, | |
| "eval_wer": 35.643385670456595, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 11.03, | |
| "learning_rate": 6.361111111111111e-06, | |
| "loss": 0.037, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "learning_rate": 5.666666666666667e-06, | |
| "loss": 0.0178, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "eval_loss": 0.5057036280632019, | |
| "eval_runtime": 66.1768, | |
| "eval_samples_per_second": 4.654, | |
| "eval_steps_per_second": 0.151, | |
| "eval_wer": 36.004331348132105, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 14.04, | |
| "learning_rate": 4.9722222222222224e-06, | |
| "loss": 0.0092, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "learning_rate": 4.277777777777778e-06, | |
| "loss": 0.0048, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "eval_loss": 0.5589306950569153, | |
| "eval_runtime": 76.2478, | |
| "eval_samples_per_second": 4.039, | |
| "eval_steps_per_second": 0.131, | |
| "eval_wer": 35.82385850929435, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 18.01, | |
| "learning_rate": 3.5833333333333335e-06, | |
| "loss": 0.0029, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 19.04, | |
| "learning_rate": 2.888888888888889e-06, | |
| "loss": 0.0022, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 19.04, | |
| "eval_loss": 0.5882430672645569, | |
| "eval_runtime": 74.2044, | |
| "eval_samples_per_second": 4.151, | |
| "eval_steps_per_second": 0.135, | |
| "eval_wer": 35.733622089875475, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 21.02, | |
| "learning_rate": 2.1944444444444445e-06, | |
| "loss": 0.0018, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.0015, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 23.01, | |
| "eval_loss": 0.5984734296798706, | |
| "eval_runtime": 65.8458, | |
| "eval_samples_per_second": 4.678, | |
| "eval_steps_per_second": 0.152, | |
| "eval_wer": 35.5711965349215, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 24.04, | |
| "learning_rate": 8.055555555555557e-07, | |
| "loss": 0.0013, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "learning_rate": 1.1111111111111112e-07, | |
| "loss": 0.0013, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "eval_loss": 0.6032947301864624, | |
| "eval_runtime": 71.6045, | |
| "eval_samples_per_second": 4.301, | |
| "eval_steps_per_second": 0.14, | |
| "eval_wer": 35.49900739938639, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 26.02, | |
| "step": 400, | |
| "total_flos": 7.22271539920896e+18, | |
| "train_loss": 0.24950471434742214, | |
| "train_runtime": 3885.9915, | |
| "train_samples_per_second": 6.588, | |
| "train_steps_per_second": 0.103 | |
| } | |
| ], | |
| "max_steps": 400, | |
| "num_train_epochs": 9223372036854775807, | |
| "total_flos": 7.22271539920896e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |