| { | |
| "best_metric": 0.9426605504587156, | |
| "best_model_checkpoint": "./electra-base-discriminator/fine_tuned_models/checkpoint-16840", | |
| "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": 0.2682, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9369266055045872, | |
| "eval_loss": 0.1812526434659958, | |
| "eval_runtime": 3.0097, | |
| "eval_samples_per_second": 289.727, | |
| "eval_steps_per_second": 36.216, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 1.6354277945032536e-05, | |
| "loss": 0.1453, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9380733944954128, | |
| "eval_loss": 0.20405027270317078, | |
| "eval_runtime": 3.1176, | |
| "eval_samples_per_second": 279.707, | |
| "eval_steps_per_second": 34.963, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 1.4309993201903469e-05, | |
| "loss": 0.1111, | |
| "step": 6315 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9415137614678899, | |
| "eval_loss": 0.2648782730102539, | |
| "eval_runtime": 3.2046, | |
| "eval_samples_per_second": 272.105, | |
| "eval_steps_per_second": 34.013, | |
| "step": 6315 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 1.2265708458774402e-05, | |
| "loss": 0.0896, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9334862385321101, | |
| "eval_loss": 0.2367054671049118, | |
| "eval_runtime": 2.8594, | |
| "eval_samples_per_second": 304.96, | |
| "eval_steps_per_second": 38.12, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 1.0221423715645335e-05, | |
| "loss": 0.0728, | |
| "step": 10525 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9334862385321101, | |
| "eval_loss": 0.21712276339530945, | |
| "eval_runtime": 3.0232, | |
| "eval_samples_per_second": 288.434, | |
| "eval_steps_per_second": 36.054, | |
| "step": 10525 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 8.177138972516268e-06, | |
| "loss": 0.0591, | |
| "step": 12630 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9323394495412844, | |
| "eval_loss": 0.2933523654937744, | |
| "eval_runtime": 3.0069, | |
| "eval_samples_per_second": 290.0, | |
| "eval_steps_per_second": 36.25, | |
| "step": 12630 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 6.132854229387201e-06, | |
| "loss": 0.0477, | |
| "step": 14735 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9346330275229358, | |
| "eval_loss": 0.2837688624858856, | |
| "eval_runtime": 3.0561, | |
| "eval_samples_per_second": 285.331, | |
| "eval_steps_per_second": 35.666, | |
| "step": 14735 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 4.088569486258134e-06, | |
| "loss": 0.0392, | |
| "step": 16840 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9426605504587156, | |
| "eval_loss": 0.3375774323940277, | |
| "eval_runtime": 3.0202, | |
| "eval_samples_per_second": 288.719, | |
| "eval_steps_per_second": 36.09, | |
| "step": 16840 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 2.044284743129067e-06, | |
| "loss": 0.0317, | |
| "step": 18945 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9369266055045872, | |
| "eval_loss": 0.36306461691856384, | |
| "eval_runtime": 2.9672, | |
| "eval_samples_per_second": 293.879, | |
| "eval_steps_per_second": 36.735, | |
| "step": 18945 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.0252, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9323394495412844, | |
| "eval_loss": 0.40919286012649536, | |
| "eval_runtime": 2.918, | |
| "eval_samples_per_second": 298.839, | |
| "eval_steps_per_second": 37.355, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 21050, | |
| "total_flos": 3.033885148659456e+16, | |
| "train_loss": 0.08898545235749378, | |
| "train_runtime": 5505.5884, | |
| "train_samples_per_second": 122.328, | |
| "train_steps_per_second": 3.823 | |
| } | |
| ], | |
| "max_steps": 21050, | |
| "num_train_epochs": 10, | |
| "total_flos": 3.033885148659456e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |