| { | |
| "best_metric": 0.5754128098487854, | |
| "best_model_checkpoint": "./vit-base/checkpoint-100", | |
| "epoch": 20.0, | |
| "global_step": 180, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 0.00018888888888888888, | |
| "loss": 1.5669, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 0.00017777777777777779, | |
| "loss": 0.2294, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 0.0267, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "learning_rate": 0.00015555555555555556, | |
| "loss": 0.002, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.56, | |
| "learning_rate": 0.00014444444444444444, | |
| "loss": 0.0008, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "learning_rate": 0.00013333333333333334, | |
| "loss": 0.0005, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.78, | |
| "learning_rate": 0.00012222222222222224, | |
| "loss": 0.0004, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.89, | |
| "learning_rate": 0.00011111111111111112, | |
| "loss": 0.0003, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0003, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 11.11, | |
| "learning_rate": 8.888888888888889e-05, | |
| "loss": 0.0003, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 11.11, | |
| "eval_accuracy": 0.8095238095238095, | |
| "eval_loss": 0.5754128098487854, | |
| "eval_runtime": 1.0617, | |
| "eval_samples_per_second": 19.78, | |
| "eval_steps_per_second": 2.826, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 12.22, | |
| "learning_rate": 7.777777777777778e-05, | |
| "loss": 0.0002, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 13.33, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 0.0002, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 14.44, | |
| "learning_rate": 5.555555555555556e-05, | |
| "loss": 0.0002, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 15.56, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.0002, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.0002, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 17.78, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.0002, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 18.89, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.0002, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.0002, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 180, | |
| "total_flos": 5.42468236068864e+16, | |
| "train_loss": 0.10162481290883281, | |
| "train_runtime": 67.7444, | |
| "train_samples_per_second": 10.333, | |
| "train_steps_per_second": 2.657 | |
| } | |
| ], | |
| "max_steps": 180, | |
| "num_train_epochs": 20, | |
| "total_flos": 5.42468236068864e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |