{ "best_metric": 0.9821428571428571, "best_model_checkpoint": "vit-large-patch16-224-dungeon-geo-morphs-0-4-28Nov24-006/checkpoint-60", "epoch": 32.0, "eval_steps": 10, "global_step": 80, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 4.0, "grad_norm": 24.072080612182617, "learning_rate": 9.861111111111112e-06, "loss": 1.6194, "step": 10 }, { "epoch": 4.0, "eval_accuracy": 0.6214285714285714, "eval_loss": 1.232202172279358, "eval_runtime": 7.6519, "eval_samples_per_second": 73.185, "eval_steps_per_second": 9.148, "step": 10 }, { "epoch": 8.0, "grad_norm": 17.304100036621094, "learning_rate": 8.472222222222223e-06, "loss": 0.7978, "step": 20 }, { "epoch": 8.0, "eval_accuracy": 0.925, "eval_loss": 0.5918735861778259, "eval_runtime": 7.3658, "eval_samples_per_second": 76.027, "eval_steps_per_second": 9.503, "step": 20 }, { "epoch": 12.0, "grad_norm": 11.8755464553833, "learning_rate": 7.083333333333335e-06, "loss": 0.2576, "step": 30 }, { "epoch": 12.0, "eval_accuracy": 0.9678571428571429, "eval_loss": 0.2720572352409363, "eval_runtime": 7.6632, "eval_samples_per_second": 73.077, "eval_steps_per_second": 9.135, "step": 30 }, { "epoch": 16.0, "grad_norm": 2.4268698692321777, "learning_rate": 5.694444444444445e-06, "loss": 0.0723, "step": 40 }, { "epoch": 16.0, "eval_accuracy": 0.9785714285714285, "eval_loss": 0.15481743216514587, "eval_runtime": 7.6991, "eval_samples_per_second": 72.736, "eval_steps_per_second": 9.092, "step": 40 }, { "epoch": 20.0, "grad_norm": 1.032953143119812, "learning_rate": 4.305555555555556e-06, "loss": 0.0202, "step": 50 }, { "epoch": 20.0, "eval_accuracy": 0.9767857142857143, "eval_loss": 0.10659553855657578, "eval_runtime": 8.3338, "eval_samples_per_second": 67.196, "eval_steps_per_second": 8.4, "step": 50 }, { "epoch": 24.0, "grad_norm": 0.3209587633609772, "learning_rate": 2.916666666666667e-06, "loss": 0.0067, "step": 60 }, { "epoch": 24.0, "eval_accuracy": 0.9821428571428571, "eval_loss": 0.07470609247684479, "eval_runtime": 8.1625, "eval_samples_per_second": 68.606, "eval_steps_per_second": 8.576, "step": 60 }, { "epoch": 28.0, "grad_norm": 0.22312191128730774, "learning_rate": 1.527777777777778e-06, "loss": 0.0035, "step": 70 }, { "epoch": 28.0, "eval_accuracy": 0.9767857142857143, "eval_loss": 0.07538952678442001, "eval_runtime": 7.554, "eval_samples_per_second": 74.133, "eval_steps_per_second": 9.267, "step": 70 }, { "epoch": 32.0, "grad_norm": 0.1788581758737564, "learning_rate": 1.3888888888888888e-07, "loss": 0.0027, "step": 80 }, { "epoch": 32.0, "eval_accuracy": 0.9785714285714285, "eval_loss": 0.07296328991651535, "eval_runtime": 7.9971, "eval_samples_per_second": 70.026, "eval_steps_per_second": 8.753, "step": 80 }, { "epoch": 32.0, "step": 80, "total_flos": 7.012786101918106e+17, "train_loss": 0.34752421248704196, "train_runtime": 465.741, "train_samples_per_second": 6.871, "train_steps_per_second": 0.172 } ], "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 }