| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "global_step": 2220, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.6e-06, | |
| "loss": 38.5727, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 7.35e-06, | |
| "loss": 24.0751, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 1.1099999999999999e-05, | |
| "loss": 15.8855, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 1.485e-05, | |
| "loss": 12.8886, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "learning_rate": 1.8599999999999998e-05, | |
| "loss": 9.9218, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 2.2349999999999998e-05, | |
| "loss": 7.1493, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "learning_rate": 2.6099999999999997e-05, | |
| "loss": 5.3703, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 2.985e-05, | |
| "loss": 4.6384, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "learning_rate": 3.36e-05, | |
| "loss": 4.3898, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "learning_rate": 3.735e-05, | |
| "loss": 4.2279, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "eval_loss": 4.11016845703125, | |
| "eval_runtime": 236.4257, | |
| "eval_samples_per_second": 18.962, | |
| "eval_steps_per_second": 2.373, | |
| "eval_wer": 2.082517793594306, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "learning_rate": 4.11e-05, | |
| "loss": 4.1565, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 5.41, | |
| "learning_rate": 4.484999999999999e-05, | |
| "loss": 4.1112, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.86, | |
| "learning_rate": 4.8599999999999995e-05, | |
| "loss": 4.0744, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 6.31, | |
| "learning_rate": 5.234999999999999e-05, | |
| "loss": 4.0272, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.76, | |
| "learning_rate": 5.6099999999999995e-05, | |
| "loss": 3.9966, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 7.21, | |
| "learning_rate": 5.985e-05, | |
| "loss": 3.8571, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 7.66, | |
| "learning_rate": 6.359999999999999e-05, | |
| "loss": 3.4211, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 8.11, | |
| "learning_rate": 6.735e-05, | |
| "loss": 2.833, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 8.56, | |
| "learning_rate": 7.11e-05, | |
| "loss": 2.5621, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "learning_rate": 7.484999999999999e-05, | |
| "loss": 2.4119, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "eval_loss": 1.173694133758545, | |
| "eval_runtime": 237.1778, | |
| "eval_samples_per_second": 18.901, | |
| "eval_steps_per_second": 2.365, | |
| "eval_wer": 2.3580960854092528, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 9.46, | |
| "learning_rate": 4.227272727272727e-05, | |
| "loss": 2.2872, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 9.91, | |
| "learning_rate": 8.181818181818181e-06, | |
| "loss": 2.186, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 2220, | |
| "total_flos": 1.6498328689753115e+19, | |
| "train_loss": 7.544306694924295, | |
| "train_runtime": 6173.2652, | |
| "train_samples_per_second": 17.208, | |
| "train_steps_per_second": 0.36 | |
| } | |
| ], | |
| "max_steps": 2220, | |
| "num_train_epochs": 10, | |
| "total_flos": 1.6498328689753115e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |