| { | |
| "best_metric": 0.896875, | |
| "best_model_checkpoint": "swinv2-tiny-patch4-window8-256-finalterm/checkpoint-160", | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.0463645458221436, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.397, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.35625, | |
| "eval_loss": 1.3485567569732666, | |
| "eval_runtime": 2.7657, | |
| "eval_samples_per_second": 115.703, | |
| "eval_steps_per_second": 3.616, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 9.14601993560791, | |
| "learning_rate": 5e-05, | |
| "loss": 1.2224, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.509375, | |
| "eval_loss": 1.0746052265167236, | |
| "eval_runtime": 2.8349, | |
| "eval_samples_per_second": 112.879, | |
| "eval_steps_per_second": 3.527, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 8.623628616333008, | |
| "learning_rate": 4.722222222222222e-05, | |
| "loss": 0.8883, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.709375, | |
| "eval_loss": 0.7385674715042114, | |
| "eval_runtime": 3.0143, | |
| "eval_samples_per_second": 106.162, | |
| "eval_steps_per_second": 3.318, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 7.014606475830078, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.73, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.71875, | |
| "eval_loss": 0.6787489056587219, | |
| "eval_runtime": 3.0319, | |
| "eval_samples_per_second": 105.543, | |
| "eval_steps_per_second": 3.298, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 5.60139799118042, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 0.6132, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.75, | |
| "eval_loss": 0.5557254552841187, | |
| "eval_runtime": 2.7949, | |
| "eval_samples_per_second": 114.494, | |
| "eval_steps_per_second": 3.578, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 6.750443458557129, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.5297, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.828125, | |
| "eval_loss": 0.4563722610473633, | |
| "eval_runtime": 2.8163, | |
| "eval_samples_per_second": 113.622, | |
| "eval_steps_per_second": 3.551, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 8.682096481323242, | |
| "learning_rate": 3.611111111111111e-05, | |
| "loss": 0.4811, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.859375, | |
| "eval_loss": 0.4000449776649475, | |
| "eval_runtime": 2.8732, | |
| "eval_samples_per_second": 111.373, | |
| "eval_steps_per_second": 3.48, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 4.585944652557373, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.484, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.853125, | |
| "eval_loss": 0.40824002027511597, | |
| "eval_runtime": 2.7794, | |
| "eval_samples_per_second": 115.134, | |
| "eval_steps_per_second": 3.598, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 5.7387237548828125, | |
| "learning_rate": 3.055555555555556e-05, | |
| "loss": 0.4525, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.890625, | |
| "eval_loss": 0.3567732274532318, | |
| "eval_runtime": 2.8097, | |
| "eval_samples_per_second": 113.893, | |
| "eval_steps_per_second": 3.559, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 5.609104633331299, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.429, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.875, | |
| "eval_loss": 0.3661341071128845, | |
| "eval_runtime": 2.8278, | |
| "eval_samples_per_second": 113.164, | |
| "eval_steps_per_second": 3.536, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 5.6711225509643555, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.4383, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.85, | |
| "eval_loss": 0.4051894247531891, | |
| "eval_runtime": 2.947, | |
| "eval_samples_per_second": 108.586, | |
| "eval_steps_per_second": 3.393, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 6.238234996795654, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.4586, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.88125, | |
| "eval_loss": 0.3318403363227844, | |
| "eval_runtime": 2.8475, | |
| "eval_samples_per_second": 112.381, | |
| "eval_steps_per_second": 3.512, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 6.806089401245117, | |
| "learning_rate": 1.9444444444444445e-05, | |
| "loss": 0.4036, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.865625, | |
| "eval_loss": 0.35395950078964233, | |
| "eval_runtime": 2.8595, | |
| "eval_samples_per_second": 111.906, | |
| "eval_steps_per_second": 3.497, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 4.199438095092773, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.3652, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.871875, | |
| "eval_loss": 0.3941846787929535, | |
| "eval_runtime": 2.8862, | |
| "eval_samples_per_second": 110.871, | |
| "eval_steps_per_second": 3.465, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 4.736207485198975, | |
| "learning_rate": 1.388888888888889e-05, | |
| "loss": 0.3822, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.890625, | |
| "eval_loss": 0.3420585095882416, | |
| "eval_runtime": 2.8296, | |
| "eval_samples_per_second": 113.091, | |
| "eval_steps_per_second": 3.534, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 5.99350643157959, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.3564, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.896875, | |
| "eval_loss": 0.31735172867774963, | |
| "eval_runtime": 2.7873, | |
| "eval_samples_per_second": 114.807, | |
| "eval_steps_per_second": 3.588, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 4.5202741622924805, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.3635, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.884375, | |
| "eval_loss": 0.3356662690639496, | |
| "eval_runtime": 2.8073, | |
| "eval_samples_per_second": 113.988, | |
| "eval_steps_per_second": 3.562, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 5.483897686004639, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 0.3645, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.89375, | |
| "eval_loss": 0.3204434812068939, | |
| "eval_runtime": 3.0124, | |
| "eval_samples_per_second": 106.229, | |
| "eval_steps_per_second": 3.32, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 7.001428604125977, | |
| "learning_rate": 2.777777777777778e-06, | |
| "loss": 0.3576, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.878125, | |
| "eval_loss": 0.33071213960647583, | |
| "eval_runtime": 2.9125, | |
| "eval_samples_per_second": 109.87, | |
| "eval_steps_per_second": 3.433, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 5.229546070098877, | |
| "learning_rate": 0.0, | |
| "loss": 0.3532, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.878125, | |
| "eval_loss": 0.3238596022129059, | |
| "eval_runtime": 2.976, | |
| "eval_samples_per_second": 107.525, | |
| "eval_steps_per_second": 3.36, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 200, | |
| "total_flos": 8.32925255860224e+17, | |
| "train_loss": 0.5535187506675721, | |
| "train_runtime": 481.721, | |
| "train_samples_per_second": 53.143, | |
| "train_steps_per_second": 0.415 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.32925255860224e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |