{ "best_metric": 0.9803571428571428, "best_model_checkpoint": "vit-large-patch16-224-dungeon-geo-morphs-0-4-30Nov24-007/checkpoint-50", "epoch": 31.90909090909091, "eval_steps": 10, "global_step": 80, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 3.909090909090909, "grad_norm": 23.13675308227539, "learning_rate": 1e-05, "loss": 1.5474, "step": 10 }, { "epoch": 3.909090909090909, "eval_accuracy": 0.5696428571428571, "eval_loss": 1.2454607486724854, "eval_runtime": 7.4612, "eval_samples_per_second": 75.055, "eval_steps_per_second": 9.382, "step": 10 }, { "epoch": 7.909090909090909, "grad_norm": 19.284276962280273, "learning_rate": 8.611111111111112e-06, "loss": 0.8254, "step": 20 }, { "epoch": 7.909090909090909, "eval_accuracy": 0.8767857142857143, "eval_loss": 0.6164689064025879, "eval_runtime": 8.5422, "eval_samples_per_second": 65.557, "eval_steps_per_second": 8.195, "step": 20 }, { "epoch": 11.909090909090908, "grad_norm": 16.49819564819336, "learning_rate": 7.222222222222223e-06, "loss": 0.2878, "step": 30 }, { "epoch": 11.909090909090908, "eval_accuracy": 0.9464285714285714, "eval_loss": 0.2867428958415985, "eval_runtime": 7.9963, "eval_samples_per_second": 70.033, "eval_steps_per_second": 8.754, "step": 30 }, { "epoch": 15.909090909090908, "grad_norm": 21.303470611572266, "learning_rate": 5.833333333333334e-06, "loss": 0.0848, "step": 40 }, { "epoch": 15.909090909090908, "eval_accuracy": 0.9732142857142857, "eval_loss": 0.1632217913866043, "eval_runtime": 8.3473, "eval_samples_per_second": 67.088, "eval_steps_per_second": 8.386, "step": 40 }, { "epoch": 19.90909090909091, "grad_norm": 1.3407821655273438, "learning_rate": 4.444444444444444e-06, "loss": 0.0235, "step": 50 }, { "epoch": 19.90909090909091, "eval_accuracy": 0.9803571428571428, "eval_loss": 0.10389769822359085, "eval_runtime": 7.4794, "eval_samples_per_second": 74.872, "eval_steps_per_second": 9.359, "step": 50 }, { "epoch": 23.90909090909091, "grad_norm": 0.5087811350822449, "learning_rate": 3.055555555555556e-06, "loss": 0.0077, "step": 60 }, { "epoch": 23.90909090909091, "eval_accuracy": 0.9714285714285714, "eval_loss": 0.08780629187822342, "eval_runtime": 8.6162, "eval_samples_per_second": 64.994, "eval_steps_per_second": 8.124, "step": 60 }, { "epoch": 27.90909090909091, "grad_norm": 0.22263120114803314, "learning_rate": 1.6666666666666667e-06, "loss": 0.0038, "step": 70 }, { "epoch": 27.90909090909091, "eval_accuracy": 0.9696428571428571, "eval_loss": 0.08838720619678497, "eval_runtime": 8.1691, "eval_samples_per_second": 68.551, "eval_steps_per_second": 8.569, "step": 70 }, { "epoch": 31.90909090909091, "grad_norm": 0.15753118693828583, "learning_rate": 2.7777777777777776e-07, "loss": 0.0027, "step": 80 }, { "epoch": 31.90909090909091, "eval_accuracy": 0.9714285714285714, "eval_loss": 0.08130071312189102, "eval_runtime": 8.1162, "eval_samples_per_second": 68.998, "eval_steps_per_second": 8.625, "step": 80 }, { "epoch": 31.90909090909091, "step": 80, "total_flos": 7.012786101918106e+17, "train_loss": 0.3478837160393596, "train_runtime": 509.7037, "train_samples_per_second": 6.435, "train_steps_per_second": 0.157 } ], "logging_steps": 10, "max_steps": 80, "num_input_tokens_seen": 0, "num_train_epochs": 40, "save_steps": 10, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 7.012786101918106e+17, "train_batch_size": 8, "trial_name": null, "trial_params": null }