| { | |
| "best_metric": 0.6926687955856323, | |
| "best_model_checkpoint": "mobilebert_sa_GLUE_Experiment_rte_256/checkpoint-20", | |
| "epoch": 6.0, | |
| "global_step": 120, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.6937, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5270758122743683, | |
| "eval_loss": 0.6926687955856323, | |
| "eval_runtime": 0.6537, | |
| "eval_samples_per_second": 423.726, | |
| "eval_steps_per_second": 4.589, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.6936, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5306859205776173, | |
| "eval_loss": 0.6928806304931641, | |
| "eval_runtime": 0.6521, | |
| "eval_samples_per_second": 424.806, | |
| "eval_steps_per_second": 4.601, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.693, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5018050541516246, | |
| "eval_loss": 0.693047285079956, | |
| "eval_runtime": 0.6509, | |
| "eval_samples_per_second": 425.563, | |
| "eval_steps_per_second": 4.609, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.693, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.48736462093862815, | |
| "eval_loss": 0.693395733833313, | |
| "eval_runtime": 0.6512, | |
| "eval_samples_per_second": 425.381, | |
| "eval_steps_per_second": 4.607, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.6927, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.4584837545126354, | |
| "eval_loss": 0.6946930885314941, | |
| "eval_runtime": 0.65, | |
| "eval_samples_per_second": 426.135, | |
| "eval_steps_per_second": 4.615, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.6909, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5126353790613718, | |
| "eval_loss": 0.6941628456115723, | |
| "eval_runtime": 0.6493, | |
| "eval_samples_per_second": 426.604, | |
| "eval_steps_per_second": 4.62, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "step": 120, | |
| "total_flos": 393536788758528.0, | |
| "train_loss": 0.6928122917811076, | |
| "train_runtime": 181.0273, | |
| "train_samples_per_second": 687.741, | |
| "train_steps_per_second": 5.524 | |
| } | |
| ], | |
| "max_steps": 1000, | |
| "num_train_epochs": 50, | |
| "total_flos": 393536788758528.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |