| { | |
| "best_metric": 0.8027194738388062, | |
| "best_model_checkpoint": "mobilebert_add_GLUE_Experiment_logit_kd_qqp_256/checkpoint-14215", | |
| "epoch": 10.0, | |
| "global_step": 28430, | |
| "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": 1.2838, | |
| "step": 2843 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6318327974276527, | |
| "eval_combined_score": 0.3159163987138264, | |
| "eval_f1": 0.0, | |
| "eval_loss": 1.2200075387954712, | |
| "eval_runtime": 88.7654, | |
| "eval_samples_per_second": 455.47, | |
| "eval_steps_per_second": 3.56, | |
| "step": 2843 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.0184, | |
| "step": 5686 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7472668810289389, | |
| "eval_combined_score": 0.6698277918787585, | |
| "eval_f1": 0.5923887027285782, | |
| "eval_loss": 0.842215895652771, | |
| "eval_runtime": 88.4928, | |
| "eval_samples_per_second": 456.873, | |
| "eval_steps_per_second": 3.571, | |
| "step": 5686 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.8633, | |
| "step": 8529 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7520158298293347, | |
| "eval_combined_score": 0.6741638176365227, | |
| "eval_f1": 0.5963118054437107, | |
| "eval_loss": 0.8231735229492188, | |
| "eval_runtime": 88.3481, | |
| "eval_samples_per_second": 457.622, | |
| "eval_steps_per_second": 3.577, | |
| "step": 8529 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.834, | |
| "step": 11372 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7562700964630225, | |
| "eval_combined_score": 0.6917039902626902, | |
| "eval_f1": 0.627137884062358, | |
| "eval_loss": 0.8192626237869263, | |
| "eval_runtime": 88.4479, | |
| "eval_samples_per_second": 457.105, | |
| "eval_steps_per_second": 3.573, | |
| "step": 11372 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.812, | |
| "step": 14215 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7595597328716299, | |
| "eval_combined_score": 0.6980041829771966, | |
| "eval_f1": 0.6364486330827631, | |
| "eval_loss": 0.8027194738388062, | |
| "eval_runtime": 88.4938, | |
| "eval_samples_per_second": 456.868, | |
| "eval_steps_per_second": 3.571, | |
| "step": 14215 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.7871, | |
| "step": 17058 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6318327974276527, | |
| "eval_combined_score": 0.3159163987138264, | |
| "eval_f1": 0.0, | |
| "eval_loss": NaN, | |
| "eval_runtime": 88.4592, | |
| "eval_samples_per_second": 457.047, | |
| "eval_steps_per_second": 3.572, | |
| "step": 17058 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.0, | |
| "step": 19901 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6318327974276527, | |
| "eval_combined_score": 0.3159163987138264, | |
| "eval_f1": 0.0, | |
| "eval_loss": NaN, | |
| "eval_runtime": 88.3936, | |
| "eval_samples_per_second": 457.386, | |
| "eval_steps_per_second": 3.575, | |
| "step": 19901 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.0, | |
| "step": 22744 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.6318327974276527, | |
| "eval_combined_score": 0.3159163987138264, | |
| "eval_f1": 0.0, | |
| "eval_loss": NaN, | |
| "eval_runtime": 88.1736, | |
| "eval_samples_per_second": 458.527, | |
| "eval_steps_per_second": 3.584, | |
| "step": 22744 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 4.1e-05, | |
| "loss": 0.0, | |
| "step": 25587 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.6318327974276527, | |
| "eval_combined_score": 0.3159163987138264, | |
| "eval_f1": 0.0, | |
| "eval_loss": NaN, | |
| "eval_runtime": 88.1069, | |
| "eval_samples_per_second": 458.875, | |
| "eval_steps_per_second": 3.587, | |
| "step": 25587 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 4e-05, | |
| "loss": 0.0, | |
| "step": 28430 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.6318327974276527, | |
| "eval_combined_score": 0.3159163987138264, | |
| "eval_f1": 0.0, | |
| "eval_loss": NaN, | |
| "eval_runtime": 88.084, | |
| "eval_samples_per_second": 458.994, | |
| "eval_steps_per_second": 3.587, | |
| "step": 28430 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 28430, | |
| "total_flos": 9.378208193445888e+16, | |
| "train_loss": 0.559854058350554, | |
| "train_runtime": 21499.6946, | |
| "train_samples_per_second": 846.166, | |
| "train_steps_per_second": 6.612 | |
| } | |
| ], | |
| "max_steps": 142150, | |
| "num_train_epochs": 50, | |
| "total_flos": 9.378208193445888e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |