| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 59.991596638655466, | |
| "global_step": 3540, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 5.08, | |
| "learning_rate": 0.0003546, | |
| "loss": 6.338, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.08, | |
| "eval_loss": 3.1361563205718994, | |
| "eval_runtime": 51.9297, | |
| "eval_samples_per_second": 12.632, | |
| "eval_wer": 1.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 10.17, | |
| "learning_rate": 0.0005715592105263158, | |
| "loss": 3.1258, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 10.17, | |
| "eval_loss": 3.0660765171051025, | |
| "eval_runtime": 50.6635, | |
| "eval_samples_per_second": 12.948, | |
| "eval_wer": 1.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 15.25, | |
| "learning_rate": 0.0005132368421052633, | |
| "loss": 1.4079, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 15.25, | |
| "eval_loss": 0.5029116272926331, | |
| "eval_runtime": 51.4993, | |
| "eval_samples_per_second": 12.738, | |
| "eval_wer": 0.6116052520118593, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 20.34, | |
| "learning_rate": 0.00045491447368421056, | |
| "loss": 0.1446, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 20.34, | |
| "eval_loss": 0.5113032460212708, | |
| "eval_runtime": 51.276, | |
| "eval_samples_per_second": 12.794, | |
| "eval_wer": 0.5307073274036426, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 25.42, | |
| "learning_rate": 0.0003965921052631579, | |
| "loss": 0.0714, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 25.42, | |
| "eval_loss": 0.5406283140182495, | |
| "eval_runtime": 51.4595, | |
| "eval_samples_per_second": 12.748, | |
| "eval_wer": 0.5006353240152478, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 30.5, | |
| "learning_rate": 0.00033826973684210527, | |
| "loss": 0.0459, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 30.5, | |
| "eval_loss": 0.5775919556617737, | |
| "eval_runtime": 51.4074, | |
| "eval_samples_per_second": 12.761, | |
| "eval_wer": 0.49767047861075814, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 35.59, | |
| "learning_rate": 0.0002799473684210526, | |
| "loss": 0.0372, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 35.59, | |
| "eval_loss": 0.5846429467201233, | |
| "eval_runtime": 51.8427, | |
| "eval_samples_per_second": 12.654, | |
| "eval_wer": 0.4805167301990682, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 40.67, | |
| "learning_rate": 0.00022162500000000003, | |
| "loss": 0.0297, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 40.67, | |
| "eval_loss": 0.5748910903930664, | |
| "eval_runtime": 51.453, | |
| "eval_samples_per_second": 12.75, | |
| "eval_wer": 0.4841168996188056, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 45.76, | |
| "learning_rate": 0.0001633026315789474, | |
| "loss": 0.0254, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 45.76, | |
| "eval_loss": 0.592321515083313, | |
| "eval_runtime": 51.3943, | |
| "eval_samples_per_second": 12.764, | |
| "eval_wer": 0.4614570097416349, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 50.84, | |
| "learning_rate": 0.00010498026315789474, | |
| "loss": 0.0198, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 50.84, | |
| "eval_loss": 0.5621945261955261, | |
| "eval_runtime": 51.5882, | |
| "eval_samples_per_second": 12.716, | |
| "eval_wer": 0.4404913172384583, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 55.92, | |
| "learning_rate": 4.665789473684211e-05, | |
| "loss": 0.0193, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 55.92, | |
| "eval_loss": 0.5657602548599243, | |
| "eval_runtime": 71.8217, | |
| "eval_samples_per_second": 9.134, | |
| "eval_wer": 0.4470563320626853, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 59.99, | |
| "step": 3540, | |
| "total_flos": 3.4970554953107223e+19, | |
| "train_runtime": 23737.912, | |
| "train_samples_per_second": 0.149 | |
| } | |
| ], | |
| "max_steps": 3540, | |
| "num_train_epochs": 60, | |
| "total_flos": 3.4970554953107223e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |