| { | |
| "best_metric": 1.1400723457336426, | |
| "best_model_checkpoint": "./vit-base/checkpoint-100", | |
| "epoch": 10.0, | |
| "global_step": 100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.00018, | |
| "loss": 1.1117, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 0.00016, | |
| "loss": 0.1567, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 0.9874135255813599, | |
| "eval_runtime": 0.2906, | |
| "eval_samples_per_second": 55.053, | |
| "eval_steps_per_second": 6.882, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 0.00014, | |
| "loss": 0.0041, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 0.00012, | |
| "loss": 0.0007, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 0.9884828329086304, | |
| "eval_runtime": 0.2687, | |
| "eval_samples_per_second": 59.539, | |
| "eval_steps_per_second": 7.442, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0005, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 8e-05, | |
| "loss": 0.0004, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 1.1330695152282715, | |
| "eval_runtime": 0.2768, | |
| "eval_samples_per_second": 57.807, | |
| "eval_steps_per_second": 7.226, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 6e-05, | |
| "loss": 0.0003, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 4e-05, | |
| "loss": 0.0003, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 1.141401767730713, | |
| "eval_runtime": 0.2772, | |
| "eval_samples_per_second": 57.724, | |
| "eval_steps_per_second": 7.216, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 2e-05, | |
| "loss": 0.0003, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.0003, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 1.1400723457336426, | |
| "eval_runtime": 0.2691, | |
| "eval_samples_per_second": 59.468, | |
| "eval_steps_per_second": 7.433, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 100, | |
| "total_flos": 3.09973514747904e+16, | |
| "train_loss": 0.12750993132591248, | |
| "train_runtime": 23.8317, | |
| "train_samples_per_second": 16.784, | |
| "train_steps_per_second": 4.196 | |
| } | |
| ], | |
| "max_steps": 100, | |
| "num_train_epochs": 10, | |
| "total_flos": 3.09973514747904e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |