| { | |
| "best_metric": 0.9426605504587156, | |
| "best_model_checkpoint": "./electra-base-discriminator/fine_tuned_models/checkpoint-2105", | |
| "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": 0.2722, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9426605504587156, | |
| "eval_loss": 0.18346692621707916, | |
| "eval_runtime": 1.9622, | |
| "eval_samples_per_second": 444.393, | |
| "eval_steps_per_second": 55.549, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 1.6354277945032536e-05, | |
| "loss": 0.147, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9369266055045872, | |
| "eval_loss": 0.21615296602249146, | |
| "eval_runtime": 2.1784, | |
| "eval_samples_per_second": 400.29, | |
| "eval_steps_per_second": 50.036, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 1.4309993201903469e-05, | |
| "loss": 0.1107, | |
| "step": 6315 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9380733944954128, | |
| "eval_loss": 0.19852299988269806, | |
| "eval_runtime": 2.3414, | |
| "eval_samples_per_second": 372.432, | |
| "eval_steps_per_second": 46.554, | |
| "step": 6315 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 1.2265708458774402e-05, | |
| "loss": 0.0895, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9369266055045872, | |
| "eval_loss": 0.22530296444892883, | |
| "eval_runtime": 1.8651, | |
| "eval_samples_per_second": 467.524, | |
| "eval_steps_per_second": 58.441, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 1.0221423715645335e-05, | |
| "loss": 0.0742, | |
| "step": 10525 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.930045871559633, | |
| "eval_loss": 0.2544686794281006, | |
| "eval_runtime": 1.9656, | |
| "eval_samples_per_second": 443.633, | |
| "eval_steps_per_second": 55.454, | |
| "step": 10525 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 8.177138972516268e-06, | |
| "loss": 0.0589, | |
| "step": 12630 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9369266055045872, | |
| "eval_loss": 0.2618718445301056, | |
| "eval_runtime": 2.1931, | |
| "eval_samples_per_second": 397.616, | |
| "eval_steps_per_second": 49.702, | |
| "step": 12630 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 6.132854229387201e-06, | |
| "loss": 0.0476, | |
| "step": 14735 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9311926605504587, | |
| "eval_loss": 0.3039261996746063, | |
| "eval_runtime": 2.0002, | |
| "eval_samples_per_second": 435.966, | |
| "eval_steps_per_second": 54.496, | |
| "step": 14735 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 4.088569486258134e-06, | |
| "loss": 0.0397, | |
| "step": 16840 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9334862385321101, | |
| "eval_loss": 0.30574584007263184, | |
| "eval_runtime": 2.0703, | |
| "eval_samples_per_second": 421.19, | |
| "eval_steps_per_second": 52.649, | |
| "step": 16840 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 2.044284743129067e-06, | |
| "loss": 0.0316, | |
| "step": 18945 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9380733944954128, | |
| "eval_loss": 0.38161835074424744, | |
| "eval_runtime": 2.2112, | |
| "eval_samples_per_second": 394.362, | |
| "eval_steps_per_second": 49.295, | |
| "step": 18945 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.0261, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9346330275229358, | |
| "eval_loss": 0.3954842984676361, | |
| "eval_runtime": 2.1873, | |
| "eval_samples_per_second": 398.661, | |
| "eval_steps_per_second": 49.833, | |
| "step": 21050 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 21050, | |
| "total_flos": 3.033885148659456e+16, | |
| "train_loss": 0.08974739491514809, | |
| "train_runtime": 3412.6761, | |
| "train_samples_per_second": 197.35, | |
| "train_steps_per_second": 6.168 | |
| } | |
| ], | |
| "max_steps": 21050, | |
| "num_train_epochs": 10, | |
| "total_flos": 3.033885148659456e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |