Skip to content
Snippets Groups Projects
Commit 2247ce18 authored by Mina Moshfegh's avatar Mina Moshfegh
Browse files

Upload New File

parent d5f5a1bb
No related branches found
No related tags found
No related merge requests found
Hide whitespace changes
Inline Side-by-side
results/CIFAR10-AIR-__None____FGSM__.log 0 → 100644
+ 84
− 0
View file @ 2247ce18
[2025-02-20 21:49:43] [CIFAR10-AIR-('None', 'FGSM')] [INFO] === Configuration ===
[2025-02-20 21:49:43] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Config(seed=42, device='cuda', epochs=1, batch_size=128, learning_rate=0.01, momentum=0.9, weight_decay=0.0005, epsilon=0.03137254901960784, alpha=0.00784313725490196, num_steps=10, random_init=True, defense_method='AIR', lambda_SD=1.0, lambda_IR=1.0, lambda_AR=1.0, lambda_Reg=1.0, alpha_range=(0.3, 0.7), use_rdrop=True, iso_noise_std=0.01, iso_clamp_min=-3, iso_clamp_max=3, iso_p_flip=0.5, iso_flip_dim=3, iso_p_rotation=0.5, iso_max_rotation=10, iso_p_crop=0.5, iso_p_erase=0.5, dataset='CIFAR10', data_root='./data', num_workers=2, lambda_lfl=1.0, feature_lambda=1.0, freeze_classifier=True, joint_lambda=0.5, adv_lambda=0.8, feat_lambda=1.0, lambda_ewc=100.0, attack_sequence=('None', 'FGSM'))
[2025-02-20 21:49:46] [CIFAR10-AIR-('None', 'FGSM')] [INFO] --- Task 1 / 2: Attack = None ---
[2025-02-20 21:49:46] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Using AIR Defense...
[2025-02-20 21:49:47] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 0, Loss: 2.3921
[2025-02-20 21:50:51] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 100, Loss: 2.1004
[2025-02-20 21:51:56] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 200, Loss: 1.8788
[2025-02-20 21:53:00] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 300, Loss: 1.6866
[2025-02-20 21:53:58] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Epoch [1/1] - Train Loss: 1.9428
[2025-02-20 21:54:01] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Clean Test Accuracy: 28.18%
[2025-02-20 21:54:01] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Model saved at: ./models/CIFAR10_AIR_None_task0.pt
[2025-02-20 21:54:04] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Clean Accuracy after Task 1, Attack=None: 28.18%
[2025-02-20 21:59:59] [CIFAR10-AIR-('None', 'FGSM')] [INFO] === Configuration ===
[2025-02-20 21:59:59] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Config(seed=42, device='cuda', epochs=1, batch_size=128, learning_rate=0.01, momentum=0.9, weight_decay=0.0005, epsilon=0.03137254901960784, alpha=0.00784313725490196, num_steps=10, random_init=True, defense_method='AIR', lambda_SD=1.0, lambda_IR=1.0, lambda_AR=1.0, lambda_Reg=1.0, alpha_range=(0.3, 0.7), use_rdrop=True, iso_noise_std=0.01, iso_clamp_min=-3, iso_clamp_max=3, iso_p_flip=0.5, iso_flip_dim=3, iso_p_rotation=0.5, iso_max_rotation=10, iso_p_crop=0.5, iso_p_erase=0.5, dataset='CIFAR10', data_root='./data', num_workers=2, lambda_lfl=1.0, feature_lambda=1.0, freeze_classifier=True, joint_lambda=0.5, adv_lambda=0.8, feat_lambda=1.0, lambda_ewc=100.0, attack_sequence=('None', 'FGSM'))
[2025-02-20 22:00:01] [CIFAR10-AIR-('None', 'FGSM')] [INFO] --- Task 1 / 2: Attack = None ---
[2025-02-20 22:00:01] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Using AIR Defense...
[2025-02-20 22:00:03] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 0, Loss: 2.3921
[2025-02-20 22:01:07] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 100, Loss: 2.1004
[2025-02-20 22:02:11] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 200, Loss: 1.8788
[2025-02-20 22:03:16] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 300, Loss: 1.6866
[2025-02-20 22:04:13] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Epoch [1/1] - Train Loss: 1.9428
[2025-02-20 22:04:16] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Clean Test Accuracy: 28.18%
[2025-02-20 22:04:17] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Model saved at: ./models/CIFAR10_AIR_None_task0.pt
[2025-02-20 22:04:19] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Clean Accuracy after Task 1, Attack=None: 28.18%
[2025-02-20 22:06:44] [CIFAR10-AIR-('None', 'FGSM')] [INFO] === Configuration ===
[2025-02-20 22:06:44] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Config(seed=42, device='cuda', epochs=1, batch_size=128, learning_rate=0.01, momentum=0.9, weight_decay=0.0005, epsilon=0.03137254901960784, alpha=0.00784313725490196, num_steps=10, random_init=True, defense_method='AIR', lambda_SD=1.0, lambda_IR=1.0, lambda_AR=1.0, lambda_Reg=1.0, alpha_range=(0.3, 0.7), use_rdrop=True, iso_noise_std=0.01, iso_clamp_min=-3, iso_clamp_max=3, iso_p_flip=0.5, iso_flip_dim=3, iso_p_rotation=0.5, iso_max_rotation=10, iso_p_crop=0.5, iso_p_erase=0.5, dataset='CIFAR10', data_root='./data', num_workers=2, lambda_lfl=1.0, feature_lambda=1.0, freeze_classifier=True, joint_lambda=0.5, adv_lambda=0.8, feat_lambda=1.0, lambda_ewc=100.0, attack_sequence=('None', 'FGSM'))
[2025-02-20 22:06:47] [CIFAR10-AIR-('None', 'FGSM')] [INFO] --- Task 1 / 2: Attack = None ---
[2025-02-20 22:06:47] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Using AIR Defense...
[2025-02-20 22:06:48] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 0, Loss: 2.3921
[2025-02-20 22:07:52] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 100, Loss: 2.1004
[2025-02-20 22:08:57] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 200, Loss: 1.8788
[2025-02-20 22:10:01] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 300, Loss: 1.6866
[2025-02-20 22:10:59] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Epoch [1/1] - Train Loss: 1.9428
[2025-02-20 22:11:01] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Clean Test Accuracy: 28.18%
[2025-02-20 22:11:02] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Model saved at: ./models/CIFAR10_AIR_None_task0.pt
[2025-02-20 22:11:05] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Clean Accuracy after Task 1, Attack=None: 28.18%
[2025-02-20 22:16:48] [CIFAR10-AIR-('None', 'FGSM')] [INFO] === Configuration ===
[2025-02-20 22:16:48] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Config(seed=42, device='cuda', epochs=1, batch_size=128, learning_rate=0.01, momentum=0.9, weight_decay=0.0005, epsilon=0.03137254901960784, alpha=0.00784313725490196, num_steps=10, random_init=True, defense_method='AIR', lambda_SD=1.0, lambda_IR=1.0, lambda_AR=1.0, lambda_Reg=1.0, alpha_range=(0.3, 0.7), use_rdrop=True, iso_noise_std=0.01, iso_clamp_min=-3, iso_clamp_max=3, iso_p_flip=0.5, iso_flip_dim=3, iso_p_rotation=0.5, iso_max_rotation=10, iso_p_crop=0.5, iso_p_erase=0.5, dataset='CIFAR10', data_root='./data', num_workers=2, lambda_lfl=1.0, feature_lambda=1.0, freeze_classifier=True, joint_lambda=0.5, adv_lambda=0.8, feat_lambda=1.0, lambda_ewc=100.0, attack_sequence=('None', 'FGSM'))
[2025-02-20 22:16:50] [CIFAR10-AIR-('None', 'FGSM')] [INFO] --- Task 1 / 2: Attack = None ---
[2025-02-20 22:16:50] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Using AIR Defense...
[2025-02-20 22:16:52] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 0, Loss: 2.3921
[2025-02-20 22:17:56] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 100, Loss: 2.1004
[2025-02-20 22:19:00] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 200, Loss: 1.8788
[2025-02-20 22:20:05] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 300, Loss: 1.6866
[2025-02-20 22:21:02] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Epoch [1/1] - Train Loss: 1.9428
[2025-02-20 22:21:05] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Clean Test Accuracy: 28.18%
[2025-02-20 22:21:06] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Model saved at: ./models/CIFAR10_AIR_None_task0.pt
[2025-02-20 22:21:08] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Clean Accuracy after Task 1, Attack=None: 28.18%
[2025-02-20 22:26:53] [CIFAR10-AIR-('None', 'FGSM')] [INFO] === Configuration ===
[2025-02-20 22:26:53] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Config(seed=42, device='cuda', epochs=1, batch_size=128, learning_rate=0.01, momentum=0.9, weight_decay=0.0005, epsilon=0.03137254901960784, alpha=0.00784313725490196, num_steps=10, random_init=True, defense_method='AIR', lambda_SD=1.0, lambda_IR=1.0, lambda_AR=1.0, lambda_Reg=1.0, alpha_range=(0.3, 0.7), use_rdrop=True, iso_noise_std=0.01, iso_clamp_min=-3, iso_clamp_max=3, iso_p_flip=0.5, iso_flip_dim=3, iso_p_rotation=0.5, iso_max_rotation=10, iso_p_crop=0.5, iso_p_erase=0.5, dataset='CIFAR10', data_root='./data', num_workers=2, lambda_lfl=1.0, feature_lambda=1.0, freeze_classifier=True, joint_lambda=0.5, adv_lambda=0.8, feat_lambda=1.0, lambda_ewc=100.0, attack_sequence=('None', 'FGSM'))
[2025-02-20 22:26:56] [CIFAR10-AIR-('None', 'FGSM')] [INFO] --- Task 1 / 2: Attack = None ---
[2025-02-20 22:28:33] [CIFAR10-AIR-('None', 'FGSM')] [INFO] === Configuration ===
[2025-02-20 22:28:33] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Config(seed=42, device='cuda', epochs=1, batch_size=128, learning_rate=0.01, momentum=0.9, weight_decay=0.0005, epsilon=0.03137254901960784, alpha=0.00784313725490196, num_steps=10, random_init=True, defense_method='AIR', lambda_SD=1.0, lambda_IR=1.0, lambda_AR=1.0, lambda_Reg=1.0, alpha_range=(0.3, 0.7), use_rdrop=True, iso_noise_std=0.01, iso_clamp_min=-3, iso_clamp_max=3, iso_p_flip=0.5, iso_flip_dim=3, iso_p_rotation=0.5, iso_max_rotation=10, iso_p_crop=0.5, iso_p_erase=0.5, dataset='CIFAR10', data_root='./data', num_workers=2, lambda_lfl=1.0, feature_lambda=1.0, freeze_classifier=True, joint_lambda=0.5, adv_lambda=0.8, feat_lambda=1.0, lambda_ewc=100.0, attack_sequence=('None', 'FGSM'))
[2025-02-20 22:28:36] [CIFAR10-AIR-('None', 'FGSM')] [INFO] --- Task 1 / 2: Attack = None ---
[2025-02-20 22:44:16] [CIFAR10-AIR-('None', 'FGSM')] [INFO] === Configuration ===
[2025-02-20 22:44:16] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Config(seed=42, device='cuda', epochs=1, batch_size=128, learning_rate=0.01, momentum=0.9, weight_decay=0.0005, epsilon=0.03137254901960784, alpha=0.00784313725490196, num_steps=10, random_init=True, defense_method='AIR', lambda_SD=1.0, lambda_IR=1.0, lambda_AR=1.0, lambda_Reg=1.0, alpha_range=(0.3, 0.7), use_rdrop=True, iso_noise_std=0.01, iso_clamp_min=-3, iso_clamp_max=3, iso_p_flip=0.5, iso_flip_dim=3, iso_p_rotation=0.5, iso_max_rotation=10, iso_p_crop=0.5, iso_p_erase=0.5, dataset='CIFAR10', data_root='./data', num_workers=2, lambda_lfl=1.0, feature_lambda=1.0, freeze_classifier=True, joint_lambda=0.5, adv_lambda=0.8, feat_lambda=1.0, lambda_ewc=100.0, attack_sequence=('None', 'FGSM'))
[2025-02-20 22:44:18] [CIFAR10-AIR-('None', 'FGSM')] [INFO] --- Task 1 / 2: Attack = None ---
[2025-02-20 22:44:18] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Using AIR Defense...
[2025-02-20 22:44:20] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 0, Loss: 2.3921
[2025-02-20 22:45:24] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 100, Loss: 2.1004
[2025-02-20 22:46:28] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 200, Loss: 1.8788
[2025-02-20 22:47:33] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 300, Loss: 1.6866
[2025-02-20 22:48:30] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Epoch [1/1] - Train Loss: 1.9428
[2025-02-20 22:48:33] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Clean Test Accuracy: 28.18%
[2025-02-20 22:48:33] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Model saved at: ./models/CIFAR10_AIR_None_task0.pt
[2025-02-20 22:48:36] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Clean Accuracy after Task 1, Attack=None: 28.18%
[2025-02-20 22:48:39] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Robust Accuracy on None after Task 1: 28.18%
[2025-02-20 22:48:50] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Robust Accuracy on FGSM after Task 1: 12.54%
[2025-02-20 22:48:50] [CIFAR10-AIR-('None', 'FGSM')] [INFO] --- Task 2 / 2: Attack = FGSM ---
[2025-02-20 22:48:50] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Using AIR Defense...
[2025-02-20 22:48:50] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 0, Loss: 2.4333
[2025-02-20 22:50:05] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 100, Loss: 2.2941
[2025-02-20 22:51:19] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 200, Loss: 1.8745
[2025-02-20 22:52:34] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Batch 300, Loss: 1.4442
[2025-02-20 22:53:39] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Epoch [1/1] - Train Loss: 1.9750
[2025-02-20 22:53:42] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Clean Test Accuracy: 50.12%
[2025-02-20 22:53:42] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Model saved at: ./models/CIFAR10_AIR_FGSM_task1.pt
[2025-02-20 22:53:45] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Clean Accuracy after Task 2, Attack=FGSM: 50.12%
[2025-02-20 22:53:48] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Robust Accuracy on None after Task 2: 50.12%
[2025-02-20 22:53:58] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Robust Accuracy on FGSM after Task 2: 21.41%
[2025-02-20 22:53:58] [CIFAR10-AIR-('None', 'FGSM')] [INFO] === All tasks completed! ===
[2025-02-20 22:53:58] [CIFAR10-AIR-('None', 'FGSM')] [INFO] === Final Results ===
[2025-02-20 22:53:58] [CIFAR10-AIR-('None', 'FGSM')] [INFO] []
[2025-02-20 22:53:58] [CIFAR10-AIR-('None', 'FGSM')] [INFO] Done!
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment