| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 2.9999236061950283, |
| "global_step": 36813, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.2, |
| "learning_rate": 4.92818774595697e-05, |
| "loss": 0.6685, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.41, |
| "learning_rate": 4.5691264757418205e-05, |
| "loss": 0.5333, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.61, |
| "learning_rate": 4.2100652055266715e-05, |
| "loss": 0.4987, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.81, |
| "learning_rate": 3.851003935311522e-05, |
| "loss": 0.4736, |
| "step": 10000 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_accuracy": 0.8371879775853286, |
| "eval_loss": 0.42127805948257446, |
| "eval_runtime": 131.8099, |
| "eval_samples_per_second": 74.463, |
| "eval_steps_per_second": 74.463, |
| "step": 12271 |
| }, |
| { |
| "epoch": 1.02, |
| "learning_rate": 3.491942665096372e-05, |
| "loss": 0.4506, |
| "step": 12500 |
| }, |
| { |
| "epoch": 1.22, |
| "learning_rate": 3.132881394881222e-05, |
| "loss": 0.3355, |
| "step": 15000 |
| }, |
| { |
| "epoch": 1.43, |
| "learning_rate": 2.7738201246660732e-05, |
| "loss": 0.327, |
| "step": 17500 |
| }, |
| { |
| "epoch": 1.63, |
| "learning_rate": 2.4147588544509238e-05, |
| "loss": 0.3286, |
| "step": 20000 |
| }, |
| { |
| "epoch": 1.83, |
| "learning_rate": 2.055697584235774e-05, |
| "loss": 0.3248, |
| "step": 22500 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_accuracy": 0.8537952114111055, |
| "eval_loss": 0.4055429697036743, |
| "eval_runtime": 128.0726, |
| "eval_samples_per_second": 76.636, |
| "eval_steps_per_second": 76.636, |
| "step": 24542 |
| }, |
| { |
| "epoch": 2.04, |
| "learning_rate": 1.6966363140206246e-05, |
| "loss": 0.2878, |
| "step": 25000 |
| }, |
| { |
| "epoch": 2.24, |
| "learning_rate": 1.3375750438054752e-05, |
| "loss": 0.1668, |
| "step": 27500 |
| }, |
| { |
| "epoch": 2.44, |
| "learning_rate": 9.785137735903255e-06, |
| "loss": 0.1662, |
| "step": 30000 |
| }, |
| { |
| "epoch": 2.65, |
| "learning_rate": 6.19452503375176e-06, |
| "loss": 0.1598, |
| "step": 32500 |
| }, |
| { |
| "epoch": 2.85, |
| "learning_rate": 2.6039123316002644e-06, |
| "loss": 0.1571, |
| "step": 35000 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_accuracy": 0.8572592969943963, |
| "eval_loss": 0.5091902017593384, |
| "eval_runtime": 128.5413, |
| "eval_samples_per_second": 76.357, |
| "eval_steps_per_second": 76.357, |
| "step": 36813 |
| }, |
| { |
| "epoch": 3.0, |
| "step": 36813, |
| "total_flos": 8.534430377160109e+16, |
| "train_loss": 0.33884849047837634, |
| "train_runtime": 68835.7532, |
| "train_samples_per_second": 17.115, |
| "train_steps_per_second": 0.535 |
| } |
| ], |
| "max_steps": 36813, |
| "num_train_epochs": 3, |
| "total_flos": 8.534430377160109e+16, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
