| { | |
| "best_metric": 0.9380733944954128, | |
| "best_model_checkpoint": "./electra-base-discriminator/fine_tuned_models/checkpoint-14735", | |
| "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": 0.2696, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9334862385321101, | |
| "eval_loss": 0.20129133760929108, | |
| "eval_runtime": 2.0684, | |
| "eval_samples_per_second": 421.585, | |
| "eval_steps_per_second": 52.698, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 1.6354277945032536e-05, | |
| "loss": 0.1453, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9369266055045872, | |
| "eval_loss": 0.23204709589481354, | |
| "eval_runtime": 2.4051, | |
| "eval_samples_per_second": 362.559, | |
| "eval_steps_per_second": 45.32, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 1.4309993201903469e-05, | |
| "loss": 0.1135, | |
| "step": 6315 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.930045871559633, | |
| "eval_loss": 0.2923499047756195, | |
| "eval_runtime": 2.4212, | |
| "eval_samples_per_second": 360.148, | |
| "eval_steps_per_second": 45.018, | |
| "step": 6315 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 1.2265708458774402e-05, | |
| "loss": 0.0906, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9346330275229358, | |
| "eval_loss": 0.32609012722969055, | |
| "eval_runtime": 1.9468, | |
| "eval_samples_per_second": 447.926, | |
| "eval_steps_per_second": 55.991, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 1.0221423715645335e-05, | |
| "loss": 0.072, | |
| "step": 10525 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9288990825688074, | |
| "eval_loss": 0.3121407628059387, | |
| "eval_runtime": 2.1923, | |
| "eval_samples_per_second": 397.751, | |
| "eval_steps_per_second": 49.719, | |
| "step": 10525 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 8.177138972516268e-06, | |
| "loss": 0.0583, | |
| "step": 12630 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9311926605504587, | |
| "eval_loss": 0.28844115138053894, | |
| "eval_runtime": 2.3738, | |
| "eval_samples_per_second": 367.343, | |
| "eval_steps_per_second": 45.918, | |
| "step": 12630 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 6.132854229387201e-06, | |
| "loss": 0.0484, | |
| "step": 14735 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9380733944954128, | |
| "eval_loss": 0.32522284984588623, | |
| "eval_runtime": 2.2471, | |
| "eval_samples_per_second": 388.052, | |
| "eval_steps_per_second": 48.506, | |
| "step": 14735 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 4.088569486258134e-06, | |
| "loss": 0.0385, | |
| "step": 16840 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9334862385321101, | |
| "eval_loss": 0.38597530126571655, | |
| "eval_runtime": 2.1627, | |
| "eval_samples_per_second": 403.202, | |
| "eval_steps_per_second": 50.4, | |
| "step": 16840 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 2.044284743129067e-06, | |
| "loss": 0.0308, | |
| "step": 18945 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9288990825688074, | |
| "eval_loss": 0.4163713753223419, | |
| "eval_runtime": 2.2597, | |
| "eval_samples_per_second": 385.896, | |
| "eval_steps_per_second": 48.237, | |
| "step": 18945 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.0245, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.926605504587156, | |
| "eval_loss": 0.4771242141723633, | |
| "eval_runtime": 2.1728, | |
| "eval_samples_per_second": 401.329, | |
| "eval_steps_per_second": 50.166, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 21050, | |
| "total_flos": 3.033885148659456e+16, | |
| "train_loss": 0.08915798015096126, | |
| "train_runtime": 3440.211, | |
| "train_samples_per_second": 195.77, | |
| "train_steps_per_second": 6.119 | |
| } | |
| ], | |
| "max_steps": 21050, | |
| "num_train_epochs": 10, | |
| "total_flos": 3.033885148659456e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |