| { | |
| "best_metric": 61.90141246696956, | |
| "best_model_checkpoint": "./models/rootflo/fauna-v0.8/checkpoint-2000", | |
| "epoch": 12.834224598930481, | |
| "eval_steps": 1000, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 2.1390374331550803, | |
| "grad_norm": 0.44384175539016724, | |
| "learning_rate": 1.2475000000000001e-06, | |
| "loss": 0.0576, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.278074866310161, | |
| "grad_norm": 0.428988516330719, | |
| "learning_rate": 2.4975000000000002e-06, | |
| "loss": 0.0492, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.278074866310161, | |
| "eval_loss": 0.05711566284298897, | |
| "eval_runtime": 1266.3226, | |
| "eval_samples_per_second": 2.053, | |
| "eval_steps_per_second": 0.006, | |
| "eval_wer": 62.958386403817265, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.4171122994652405, | |
| "grad_norm": 0.5753424167633057, | |
| "learning_rate": 3.745e-06, | |
| "loss": 0.0391, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 8.556149732620321, | |
| "grad_norm": 0.6431854963302612, | |
| "learning_rate": 4.9950000000000005e-06, | |
| "loss": 0.0287, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.556149732620321, | |
| "eval_loss": 0.07155641168355942, | |
| "eval_runtime": 1251.3824, | |
| "eval_samples_per_second": 2.078, | |
| "eval_steps_per_second": 0.006, | |
| "eval_wer": 61.90141246696956, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 10.695187165775401, | |
| "grad_norm": 0.6504905819892883, | |
| "learning_rate": 4.952345503673166e-06, | |
| "loss": 0.0183, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 12.834224598930481, | |
| "grad_norm": 0.8102747201919556, | |
| "learning_rate": 4.810824323930563e-06, | |
| "loss": 0.0115, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.834224598930481, | |
| "eval_loss": 0.08655478060245514, | |
| "eval_runtime": 1264.8967, | |
| "eval_samples_per_second": 2.056, | |
| "eval_steps_per_second": 0.006, | |
| "eval_wer": 62.38047259661996, | |
| "step": 3000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 10000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 43, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.25 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 1 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.337679407027684e+21, | |
| "train_batch_size": 72, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |