| { | |
| "best_metric": 0.7945205479452054, | |
| "best_model_checkpoint": "swin-small-patch4-window7-224-finetuned-piid/checkpoint-266", | |
| "epoch": 19.51219512195122, | |
| "eval_steps": 500, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 1.25e-05, | |
| "loss": 1.4653, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.2497, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "eval_accuracy": 0.4703196347031963, | |
| "eval_loss": 1.1627486944198608, | |
| "eval_runtime": 2.4207, | |
| "eval_samples_per_second": 90.471, | |
| "eval_steps_per_second": 11.567, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.9997, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 5e-05, | |
| "loss": 0.7297, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6940639269406392, | |
| "eval_loss": 0.7559153437614441, | |
| "eval_runtime": 2.1672, | |
| "eval_samples_per_second": 101.051, | |
| "eval_steps_per_second": 12.92, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 4.8611111111111115e-05, | |
| "loss": 0.5636, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 4.722222222222222e-05, | |
| "loss": 0.6428, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "eval_accuracy": 0.6575342465753424, | |
| "eval_loss": 0.8122929334640503, | |
| "eval_runtime": 2.207, | |
| "eval_samples_per_second": 99.23, | |
| "eval_steps_per_second": 12.687, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "learning_rate": 4.5833333333333334e-05, | |
| "loss": 0.5561, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "learning_rate": 4.4444444444444447e-05, | |
| "loss": 0.5308, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7214611872146118, | |
| "eval_loss": 0.5771130323410034, | |
| "eval_runtime": 3.3532, | |
| "eval_samples_per_second": 65.311, | |
| "eval_steps_per_second": 8.35, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "learning_rate": 4.305555555555556e-05, | |
| "loss": 0.489, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 0.5384, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "eval_accuracy": 0.7168949771689498, | |
| "eval_loss": 0.6496338248252869, | |
| "eval_runtime": 2.1941, | |
| "eval_samples_per_second": 99.814, | |
| "eval_steps_per_second": 12.762, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 5.37, | |
| "learning_rate": 4.027777777777778e-05, | |
| "loss": 0.5137, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 5.85, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.4094, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7579908675799086, | |
| "eval_loss": 0.5802311301231384, | |
| "eval_runtime": 2.1787, | |
| "eval_samples_per_second": 100.517, | |
| "eval_steps_per_second": 12.852, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 6.34, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.458, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 6.83, | |
| "learning_rate": 3.611111111111111e-05, | |
| "loss": 0.3204, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 6.98, | |
| "eval_accuracy": 0.7488584474885844, | |
| "eval_loss": 0.6126295328140259, | |
| "eval_runtime": 2.3006, | |
| "eval_samples_per_second": 95.191, | |
| "eval_steps_per_second": 12.17, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 7.32, | |
| "learning_rate": 3.472222222222222e-05, | |
| "loss": 0.3744, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.3364, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7442922374429224, | |
| "eval_loss": 0.539148211479187, | |
| "eval_runtime": 2.1908, | |
| "eval_samples_per_second": 99.964, | |
| "eval_steps_per_second": 12.781, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 8.29, | |
| "learning_rate": 3.194444444444444e-05, | |
| "loss": 0.2656, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 8.78, | |
| "learning_rate": 3.055555555555556e-05, | |
| "loss": 0.278, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 8.98, | |
| "eval_accuracy": 0.7534246575342466, | |
| "eval_loss": 0.5725705027580261, | |
| "eval_runtime": 2.1671, | |
| "eval_samples_per_second": 101.058, | |
| "eval_steps_per_second": 12.921, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 9.27, | |
| "learning_rate": 2.916666666666667e-05, | |
| "loss": 0.2159, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 9.76, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.3469, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.7899543378995434, | |
| "eval_loss": 0.5220110416412354, | |
| "eval_runtime": 2.1745, | |
| "eval_samples_per_second": 100.711, | |
| "eval_steps_per_second": 12.876, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 10.24, | |
| "learning_rate": 2.6388888888888892e-05, | |
| "loss": 0.2572, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 10.73, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.2631, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 10.98, | |
| "eval_accuracy": 0.730593607305936, | |
| "eval_loss": 0.6372915506362915, | |
| "eval_runtime": 2.162, | |
| "eval_samples_per_second": 101.294, | |
| "eval_steps_per_second": 12.951, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 11.22, | |
| "learning_rate": 2.361111111111111e-05, | |
| "loss": 0.2701, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 11.71, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 0.1854, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.7808219178082192, | |
| "eval_loss": 0.5767889022827148, | |
| "eval_runtime": 2.2505, | |
| "eval_samples_per_second": 97.312, | |
| "eval_steps_per_second": 12.442, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "learning_rate": 2.0833333333333336e-05, | |
| "loss": 0.1905, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 12.68, | |
| "learning_rate": 1.9444444444444445e-05, | |
| "loss": 0.2212, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 12.98, | |
| "eval_accuracy": 0.7945205479452054, | |
| "eval_loss": 0.5398662090301514, | |
| "eval_runtime": 2.1593, | |
| "eval_samples_per_second": 101.424, | |
| "eval_steps_per_second": 12.967, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 13.17, | |
| "learning_rate": 1.8055555555555555e-05, | |
| "loss": 0.2011, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 13.66, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "loss": 0.1918, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.7899543378995434, | |
| "eval_loss": 0.5777022242546082, | |
| "eval_runtime": 2.273, | |
| "eval_samples_per_second": 96.346, | |
| "eval_steps_per_second": 12.318, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 14.15, | |
| "learning_rate": 1.527777777777778e-05, | |
| "loss": 0.2395, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 14.63, | |
| "learning_rate": 1.388888888888889e-05, | |
| "loss": 0.2053, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 14.98, | |
| "eval_accuracy": 0.7899543378995434, | |
| "eval_loss": 0.5608686208724976, | |
| "eval_runtime": 2.1846, | |
| "eval_samples_per_second": 100.248, | |
| "eval_steps_per_second": 12.817, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 15.12, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.1599, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 15.61, | |
| "learning_rate": 1.1111111111111112e-05, | |
| "loss": 0.1409, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.7899543378995434, | |
| "eval_loss": 0.5490272641181946, | |
| "eval_runtime": 2.1581, | |
| "eval_samples_per_second": 101.479, | |
| "eval_steps_per_second": 12.975, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 16.1, | |
| "learning_rate": 9.722222222222223e-06, | |
| "loss": 0.1098, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 16.59, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.1204, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 16.98, | |
| "eval_accuracy": 0.7671232876712328, | |
| "eval_loss": 0.609619677066803, | |
| "eval_runtime": 2.3764, | |
| "eval_samples_per_second": 92.156, | |
| "eval_steps_per_second": 11.782, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 17.07, | |
| "learning_rate": 6.944444444444445e-06, | |
| "loss": 0.1498, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 17.56, | |
| "learning_rate": 5.555555555555556e-06, | |
| "loss": 0.1467, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.7625570776255708, | |
| "eval_loss": 0.6567701697349548, | |
| "eval_runtime": 2.1666, | |
| "eval_samples_per_second": 101.078, | |
| "eval_steps_per_second": 12.923, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 18.05, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 0.2156, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 18.54, | |
| "learning_rate": 2.777777777777778e-06, | |
| "loss": 0.125, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 18.98, | |
| "eval_accuracy": 0.771689497716895, | |
| "eval_loss": 0.6355265974998474, | |
| "eval_runtime": 2.1585, | |
| "eval_samples_per_second": 101.461, | |
| "eval_steps_per_second": 12.972, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 19.02, | |
| "learning_rate": 1.388888888888889e-06, | |
| "loss": 0.1898, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 19.51, | |
| "learning_rate": 0.0, | |
| "loss": 0.114, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 19.51, | |
| "eval_accuracy": 0.776255707762557, | |
| "eval_loss": 0.6257457137107849, | |
| "eval_runtime": 2.2005, | |
| "eval_samples_per_second": 99.524, | |
| "eval_steps_per_second": 12.724, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 19.51, | |
| "step": 400, | |
| "total_flos": 5.621068142334812e+17, | |
| "train_loss": 0.3745241871476173, | |
| "train_runtime": 481.7473, | |
| "train_samples_per_second": 27.11, | |
| "train_steps_per_second": 0.83 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 400, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 5.621068142334812e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |