| { | |
| "best_metric": 0.9461009174311926, | |
| "best_model_checkpoint": "./electra-base-discriminator/fine_tuned_models/checkpoint-8420", | |
| "epoch": 10.0, | |
| "global_step": 21050, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 1.8398562688161604e-05, | |
| "loss": 0.2346, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9426605504587156, | |
| "eval_loss": 0.21175895631313324, | |
| "eval_runtime": 2.2043, | |
| "eval_samples_per_second": 395.593, | |
| "eval_steps_per_second": 49.449, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 1.6354277945032536e-05, | |
| "loss": 0.1204, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9438073394495413, | |
| "eval_loss": 0.19780370593070984, | |
| "eval_runtime": 2.1909, | |
| "eval_samples_per_second": 398.006, | |
| "eval_steps_per_second": 49.751, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 1.4309993201903469e-05, | |
| "loss": 0.0893, | |
| "step": 6315 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9415137614678899, | |
| "eval_loss": 0.2090323567390442, | |
| "eval_runtime": 2.2098, | |
| "eval_samples_per_second": 394.601, | |
| "eval_steps_per_second": 49.325, | |
| "step": 6315 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 1.2265708458774402e-05, | |
| "loss": 0.0708, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9461009174311926, | |
| "eval_loss": 0.22068844735622406, | |
| "eval_runtime": 2.1815, | |
| "eval_samples_per_second": 399.725, | |
| "eval_steps_per_second": 49.966, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 1.0221423715645335e-05, | |
| "loss": 0.0539, | |
| "step": 10525 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9461009174311926, | |
| "eval_loss": 0.21388141810894012, | |
| "eval_runtime": 2.1881, | |
| "eval_samples_per_second": 398.521, | |
| "eval_steps_per_second": 49.815, | |
| "step": 10525 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 8.177138972516268e-06, | |
| "loss": 0.0424, | |
| "step": 12630 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9438073394495413, | |
| "eval_loss": 0.23279055953025818, | |
| "eval_runtime": 2.1898, | |
| "eval_samples_per_second": 398.209, | |
| "eval_steps_per_second": 49.776, | |
| "step": 12630 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 6.132854229387201e-06, | |
| "loss": 0.0316, | |
| "step": 14735 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9392201834862385, | |
| "eval_loss": 0.27414366602897644, | |
| "eval_runtime": 2.1831, | |
| "eval_samples_per_second": 399.423, | |
| "eval_steps_per_second": 49.928, | |
| "step": 14735 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 4.088569486258134e-06, | |
| "loss": 0.0259, | |
| "step": 16840 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9369266055045872, | |
| "eval_loss": 0.3532903790473938, | |
| "eval_runtime": 2.1863, | |
| "eval_samples_per_second": 398.847, | |
| "eval_steps_per_second": 49.856, | |
| "step": 16840 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 2.044284743129067e-06, | |
| "loss": 0.0184, | |
| "step": 18945 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9369266055045872, | |
| "eval_loss": 0.3735915720462799, | |
| "eval_runtime": 2.1998, | |
| "eval_samples_per_second": 396.399, | |
| "eval_steps_per_second": 49.55, | |
| "step": 18945 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.0137, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9357798165137615, | |
| "eval_loss": 0.39657318592071533, | |
| "eval_runtime": 2.1994, | |
| "eval_samples_per_second": 396.467, | |
| "eval_steps_per_second": 49.558, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 21050, | |
| "total_flos": 2.831471218376448e+16, | |
| "train_loss": 0.07010283381808682, | |
| "train_runtime": 4415.7756, | |
| "train_samples_per_second": 152.519, | |
| "train_steps_per_second": 4.767 | |
| } | |
| ], | |
| "max_steps": 21050, | |
| "num_train_epochs": 10, | |
| "total_flos": 2.831471218376448e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |