| { | |
| "best_metric": 0.9306122448979591, | |
| "best_model_checkpoint": "rubert-tiny2-finetuned-classification\\run-35\\checkpoint-2200", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 2200, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 5.2795502129364494e-05, | |
| "loss": 1.1083, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8979591836734694, | |
| "eval_loss": 0.4396211504936218, | |
| "eval_runtime": 1.736, | |
| "eval_samples_per_second": 141.129, | |
| "eval_steps_per_second": 9.217, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 3.72674132677867e-05, | |
| "loss": 0.3815, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9183673469387755, | |
| "eval_loss": 0.37450405955314636, | |
| "eval_runtime": 1.562, | |
| "eval_samples_per_second": 156.85, | |
| "eval_steps_per_second": 10.243, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 2.173932440620891e-05, | |
| "loss": 0.2772, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9224489795918367, | |
| "eval_loss": 0.3433334529399872, | |
| "eval_runtime": 1.386, | |
| "eval_samples_per_second": 176.767, | |
| "eval_steps_per_second": 11.544, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "learning_rate": 6.211235544631118e-06, | |
| "loss": 0.2167, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9306122448979591, | |
| "eval_loss": 0.3520706593990326, | |
| "eval_runtime": 1.168, | |
| "eval_samples_per_second": 209.761, | |
| "eval_steps_per_second": 13.699, | |
| "step": 2200 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "total_flos": 4026586324950.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 6.832359099094229e-05, | |
| "num_train_epochs": 4, | |
| "per_device_train_batch_size": 4, | |
| "seed": 37 | |
| } | |
| } | |