| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 6.651884700665189, | |
| "eval_steps": 1000, | |
| "global_step": 6000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 1.994e-05, | |
| "loss": 1.0646, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "eval_loss": 0.1494615077972412, | |
| "eval_runtime": 282.0461, | |
| "eval_samples_per_second": 32.062, | |
| "eval_steps_per_second": 1.003, | |
| "eval_wer": 0.10898120702366802, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 2.965802752293578e-05, | |
| "loss": 0.0739, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "eval_loss": 0.13595812022686005, | |
| "eval_runtime": 283.0515, | |
| "eval_samples_per_second": 31.948, | |
| "eval_steps_per_second": 1.0, | |
| "eval_wer": 0.09173801176629903, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "learning_rate": 2.8969954128440366e-05, | |
| "loss": 0.0523, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "eval_loss": 0.12313445657491684, | |
| "eval_runtime": 284.3097, | |
| "eval_samples_per_second": 31.807, | |
| "eval_steps_per_second": 0.995, | |
| "eval_wer": 0.08295090353741291, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "learning_rate": 2.8281880733944955e-05, | |
| "loss": 0.0403, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "eval_loss": 0.12026045471429825, | |
| "eval_runtime": 284.5505, | |
| "eval_samples_per_second": 31.78, | |
| "eval_steps_per_second": 0.995, | |
| "eval_wer": 0.07766961074916116, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 5.54, | |
| "learning_rate": 2.7594495412844036e-05, | |
| "loss": 0.0316, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 5.54, | |
| "eval_loss": 0.12377411127090454, | |
| "eval_runtime": 284.9055, | |
| "eval_samples_per_second": 31.74, | |
| "eval_steps_per_second": 0.993, | |
| "eval_wer": 0.07822632822256662, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 6.65, | |
| "learning_rate": 2.6906422018348625e-05, | |
| "loss": 0.0291, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 6.65, | |
| "eval_loss": 0.13142403960227966, | |
| "eval_runtime": 283.6473, | |
| "eval_samples_per_second": 31.881, | |
| "eval_steps_per_second": 0.998, | |
| "eval_wer": 0.07968583078798092, | |
| "step": 6000 | |
| } | |
| ], | |
| "logging_steps": 1000, | |
| "max_steps": 45100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 1000, | |
| "total_flos": 2.956600942328697e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |