| { | |
| "best_metric": 0.4843260943889618, | |
| "best_model_checkpoint": "hBERTv1_new_pretrain_48_sst2/checkpoint-527", | |
| "epoch": 6.0, | |
| "global_step": 3162, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 3.9200000000000004e-05, | |
| "loss": 0.3894, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8084862385321101, | |
| "eval_loss": 0.4843260943889618, | |
| "eval_runtime": 1.5514, | |
| "eval_samples_per_second": 562.079, | |
| "eval_steps_per_second": 4.512, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 3.8400000000000005e-05, | |
| "loss": 0.2353, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7947247706422018, | |
| "eval_loss": 0.6085422039031982, | |
| "eval_runtime": 1.5494, | |
| "eval_samples_per_second": 562.793, | |
| "eval_steps_per_second": 4.518, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 3.76e-05, | |
| "loss": 0.1917, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7901376146788991, | |
| "eval_loss": 0.5546697378158569, | |
| "eval_runtime": 1.56, | |
| "eval_samples_per_second": 558.99, | |
| "eval_steps_per_second": 4.487, | |
| "step": 1581 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 3.680000000000001e-05, | |
| "loss": 0.1653, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.805045871559633, | |
| "eval_loss": 0.6457037925720215, | |
| "eval_runtime": 1.5511, | |
| "eval_samples_per_second": 562.167, | |
| "eval_steps_per_second": 4.513, | |
| "step": 2108 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 3.6e-05, | |
| "loss": 0.1453, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7935779816513762, | |
| "eval_loss": 0.5609800815582275, | |
| "eval_runtime": 1.5514, | |
| "eval_samples_per_second": 562.072, | |
| "eval_steps_per_second": 4.512, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 3.52e-05, | |
| "loss": 0.1296, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.786697247706422, | |
| "eval_loss": 0.645728349685669, | |
| "eval_runtime": 1.5574, | |
| "eval_samples_per_second": 559.892, | |
| "eval_steps_per_second": 4.495, | |
| "step": 3162 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "step": 3162, | |
| "total_flos": 5.984501559302554e+16, | |
| "train_loss": 0.20944157648056388, | |
| "train_runtime": 2148.5153, | |
| "train_samples_per_second": 1567.338, | |
| "train_steps_per_second": 12.264 | |
| } | |
| ], | |
| "max_steps": 26350, | |
| "num_train_epochs": 50, | |
| "total_flos": 5.984501559302554e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |