text
stringlengths 5
1.13k
|
|---|
2020-02-02 11:35:37 Iteration 350 Training Loss: 2.248e-01 Loss in Target Net: 1.294e-01 |
2020-02-02 11:35:53 Iteration 400 Training Loss: 2.249e-01 Loss in Target Net: 1.233e-01 |
2020-02-02 11:36:10 Iteration 450 Training Loss: 2.152e-01 Loss in Target Net: 1.349e-01 |
2020-02-02 11:36:29 Iteration 500 Training Loss: 2.238e-01 Loss in Target Net: 1.059e-01 |
2020-02-02 11:36:47 Iteration 550 Training Loss: 2.215e-01 Loss in Target Net: 1.221e-01 |
2020-02-02 11:37:03 Iteration 600 Training Loss: 2.169e-01 Loss in Target Net: 1.256e-01 |
2020-02-02 11:37:21 Iteration 650 Training Loss: 2.172e-01 Loss in Target Net: 9.861e-02 |
2020-02-02 11:37:36 Iteration 700 Training Loss: 2.140e-01 Loss in Target Net: 1.219e-01 |
2020-02-02 11:37:53 Iteration 750 Training Loss: 2.118e-01 Loss in Target Net: 1.099e-01 |
2020-02-02 11:38:11 Iteration 800 Training Loss: 2.166e-01 Loss in Target Net: 1.137e-01 |
2020-02-02 11:38:28 Iteration 850 Training Loss: 2.144e-01 Loss in Target Net: 1.049e-01 |
2020-02-02 11:38:44 Iteration 900 Training Loss: 2.135e-01 Loss in Target Net: 1.294e-01 |
2020-02-02 11:39:02 Iteration 950 Training Loss: 2.130e-01 Loss in Target Net: 9.809e-02 |
2020-02-02 11:39:18 Iteration 1000 Training Loss: 2.127e-01 Loss in Target Net: 9.209e-02 |
2020-02-02 11:39:36 Iteration 1050 Training Loss: 2.189e-01 Loss in Target Net: 9.052e-02 |
2020-02-02 11:39:53 Iteration 1100 Training Loss: 2.140e-01 Loss in Target Net: 1.175e-01 |
2020-02-02 11:40:11 Iteration 1150 Training Loss: 2.082e-01 Loss in Target Net: 1.032e-01 |
2020-02-02 11:40:27 Iteration 1200 Training Loss: 2.127e-01 Loss in Target Net: 1.318e-01 |
2020-02-02 11:40:45 Iteration 1250 Training Loss: 2.075e-01 Loss in Target Net: 1.076e-01 |
2020-02-02 11:41:02 Iteration 1300 Training Loss: 2.088e-01 Loss in Target Net: 1.143e-01 |
2020-02-02 11:41:20 Iteration 1350 Training Loss: 2.080e-01 Loss in Target Net: 9.678e-02 |
2020-02-02 11:41:36 Iteration 1400 Training Loss: 2.177e-01 Loss in Target Net: 1.062e-01 |
2020-02-02 11:41:54 Iteration 1450 Training Loss: 2.113e-01 Loss in Target Net: 1.183e-01 |
2020-02-02 11:42:11 Iteration 1499 Training Loss: 2.095e-01 Loss in Target Net: 1.349e-01 |
Evaluating against victims networks |
DPN92 |
Using Adam for retraining |
Files already downloaded and verified |
2020-02-02 11:42:21, Epoch 0, Iteration 7, loss 0.318 (0.504), acc 92.308 (89.000) |
2020-02-02 11:43:18, Epoch 30, Iteration 7, loss 0.000 (0.000), acc 100.000 (100.000) |
Target Label: 6, Poison label: 8, Prediction:8, Target's Score:[-0.32191554, -1.0705252, -0.123313524, -2.8186245, 1.31438, -1.9395487, 4.5642676, -2.1060102, 4.83693, -2.0429184], Poisons' Predictions:[8, 8, 8, 8, 8] |
2020-02-02 11:44:17 Epoch 59, Val iteration 0, acc 93.400 (93.400) |
2020-02-02 11:44:25 Epoch 59, Val iteration 19, acc 93.200 (92.940) |
* Prec: 92.9400016784668 |
-------- |
------SUMMARY------ |
TIME ELAPSED (mins): 8 |
TARGET INDEX: 19 |
DPN92 1 |
Namespace(chk_path='chk-black-end2end', chk_subdir='poisons', device='cuda', dset_path='datasets', end2end=True, eval_poison_path='', gpu='2', lr_decay_epoch=[30, 45], mode='mean', model_resume_path='model-chks', nearest=False, net_repeat=1, num_per_class=50, original_grad=True, poison_decay_ites=[], poison_decay_ratio=0.1, poison_epsilon=0.1, poison_ites=1500, poison_label=8, poison_lr=0.04, poison_momentum=0.9, poison_num=5, poison_opt='adam', resume_poison_ite=0, retrain_bsize=64, retrain_epochs=60, retrain_lr=0.0001, retrain_momentum=0.9, retrain_opt='adam', retrain_wd=0.0005, subs_chk_name=['ckpt-%s-4800-dp0.200-droplayer0.000-seed1226.t7', 'ckpt-%s-4800-dp0.250-droplayer0.000-seed1226.t7', 'ckpt-%s-4800-dp0.300-droplayer0.000.t7'], subs_dp=[0.2, 0.25, 0.3], subset_group=0, substitute_nets=['DPN92', 'SENet18', 'ResNet50', 'ResNeXt29_2x64d'], target_index=2, target_label=6, target_net=['DPN92'], test_chk_name='ckpt-%s-4800.t7', tol=1e-06, train_data_path='datasets/CIFAR10_TRAIN_Split.pth') |
Path: chk-black-end2end/mean/1500/2 |
Selected base image indices: [213, 225, 227, 247, 249] |
2020-02-02 10:49:23 Iteration 0 Training Loss: 1.016e+00 Loss in Target Net: 1.452e+00 |
2020-02-02 10:49:42 Iteration 50 Training Loss: 2.394e-01 Loss in Target Net: 8.576e-02 |
2020-02-02 10:50:01 Iteration 100 Training Loss: 2.124e-01 Loss in Target Net: 5.872e-02 |
2020-02-02 10:50:19 Iteration 150 Training Loss: 2.000e-01 Loss in Target Net: 4.674e-02 |
2020-02-02 10:50:36 Iteration 200 Training Loss: 1.969e-01 Loss in Target Net: 5.076e-02 |
2020-02-02 10:50:55 Iteration 250 Training Loss: 1.897e-01 Loss in Target Net: 5.726e-02 |
2020-02-02 10:51:13 Iteration 300 Training Loss: 1.901e-01 Loss in Target Net: 4.108e-02 |
2020-02-02 10:51:33 Iteration 350 Training Loss: 1.864e-01 Loss in Target Net: 4.620e-02 |
2020-02-02 10:51:51 Iteration 400 Training Loss: 1.843e-01 Loss in Target Net: 4.098e-02 |
2020-02-02 10:52:08 Iteration 450 Training Loss: 1.848e-01 Loss in Target Net: 4.736e-02 |
2020-02-02 10:52:26 Iteration 500 Training Loss: 1.832e-01 Loss in Target Net: 4.105e-02 |
2020-02-02 10:52:44 Iteration 550 Training Loss: 1.808e-01 Loss in Target Net: 4.203e-02 |
2020-02-02 10:53:02 Iteration 600 Training Loss: 1.807e-01 Loss in Target Net: 3.900e-02 |
2020-02-02 10:53:19 Iteration 650 Training Loss: 1.798e-01 Loss in Target Net: 4.331e-02 |
2020-02-02 10:53:36 Iteration 700 Training Loss: 1.801e-01 Loss in Target Net: 4.290e-02 |
2020-02-02 10:53:53 Iteration 750 Training Loss: 1.816e-01 Loss in Target Net: 3.878e-02 |
2020-02-02 10:54:10 Iteration 800 Training Loss: 1.788e-01 Loss in Target Net: 4.390e-02 |
2020-02-02 10:54:28 Iteration 850 Training Loss: 1.785e-01 Loss in Target Net: 5.144e-02 |
2020-02-02 10:54:45 Iteration 900 Training Loss: 1.822e-01 Loss in Target Net: 4.157e-02 |
2020-02-02 10:55:02 Iteration 950 Training Loss: 1.800e-01 Loss in Target Net: 3.893e-02 |
2020-02-02 10:55:20 Iteration 1000 Training Loss: 1.776e-01 Loss in Target Net: 3.691e-02 |
2020-02-02 10:55:38 Iteration 1050 Training Loss: 1.814e-01 Loss in Target Net: 3.576e-02 |
2020-02-02 10:55:56 Iteration 1100 Training Loss: 1.726e-01 Loss in Target Net: 4.014e-02 |
2020-02-02 10:56:14 Iteration 1150 Training Loss: 1.794e-01 Loss in Target Net: 3.278e-02 |
2020-02-02 10:56:34 Iteration 1200 Training Loss: 1.775e-01 Loss in Target Net: 3.432e-02 |
2020-02-02 10:56:51 Iteration 1250 Training Loss: 1.792e-01 Loss in Target Net: 3.539e-02 |
2020-02-02 10:57:08 Iteration 1300 Training Loss: 1.813e-01 Loss in Target Net: 3.766e-02 |
2020-02-02 10:57:25 Iteration 1350 Training Loss: 1.787e-01 Loss in Target Net: 3.713e-02 |
2020-02-02 10:57:42 Iteration 1400 Training Loss: 1.785e-01 Loss in Target Net: 3.731e-02 |
2020-02-02 10:57:59 Iteration 1450 Training Loss: 1.764e-01 Loss in Target Net: 3.777e-02 |
2020-02-02 10:58:16 Iteration 1499 Training Loss: 1.820e-01 Loss in Target Net: 4.502e-02 |
Evaluating against victims networks |
DPN92 |
Using Adam for retraining |
Files already downloaded and verified |
2020-02-02 10:58:25, Epoch 0, Iteration 7, loss 0.533 (0.400), acc 84.615 (91.600) |
2020-02-02 10:59:23, Epoch 30, Iteration 7, loss 0.000 (0.000), acc 100.000 (100.000) |
Target Label: 6, Poison label: 8, Prediction:8, Target's Score:[-1.0399786, -2.378268, -3.5537567, -0.7901129, -1.6844044, -1.4772218, 4.363379, -2.4780853, 8.476478, 0.9623476], Poisons' Predictions:[8, 8, 8, 8, 8] |
2020-02-02 11:00:24 Epoch 59, Val iteration 0, acc 91.600 (91.600) |
2020-02-02 11:00:31 Epoch 59, Val iteration 19, acc 93.800 (92.800) |
* Prec: 92.80000190734863 |
-------- |
------SUMMARY------ |
TIME ELAPSED (mins): 8 |
TARGET INDEX: 2 |
DPN92 1 |
Namespace(chk_path='chk-black-end2end', chk_subdir='poisons', device='cuda', dset_path='datasets', end2end=True, eval_poison_path='', gpu='0', lr_decay_epoch=[30, 45], mode='mean', model_resume_path='model-chks', nearest=False, net_repeat=1, num_per_class=50, original_grad=True, poison_decay_ites=[], poison_decay_ratio=0.1, poison_epsilon=0.1, poison_ites=1500, poison_label=8, poison_lr=0.04, poison_momentum=0.9, poison_num=5, poison_opt='adam', resume_poison_ite=0, retrain_bsize=64, retrain_epochs=60, retrain_lr=0.0001, retrain_momentum=0.9, retrain_opt='adam', retrain_wd=0.0005, subs_chk_name=['ckpt-%s-4800-dp0.200-droplayer0.000-seed1226.t7', 'ckpt-%s-4800-dp0.250-droplayer0.000-seed1226.t7', 'ckpt-%s-4800-dp0.300-droplayer0.000.t7'], subs_dp=[0.2, 0.25, 0.3], subset_group=0, substitute_nets=['DPN92', 'SENet18', 'ResNet50', 'ResNeXt29_2x64d'], target_index=20, target_label=6, target_net=['DPN92'], test_chk_name='ckpt-%s-4800.t7', tol=1e-06, train_data_path='datasets/CIFAR10_TRAIN_Split.pth') |
Path: chk-black-end2end/mean/1500/20 |
Selected base image indices: [213, 225, 227, 247, 249] |
2020-02-02 11:46:25 Iteration 0 Training Loss: 1.014e+00 Loss in Target Net: 1.245e+00 |
2020-02-02 11:46:42 Iteration 50 Training Loss: 2.512e-01 Loss in Target Net: 5.185e-02 |
2020-02-02 11:46:58 Iteration 100 Training Loss: 2.135e-01 Loss in Target Net: 5.193e-02 |
2020-02-02 11:47:14 Iteration 150 Training Loss: 2.032e-01 Loss in Target Net: 4.463e-02 |
2020-02-02 11:47:33 Iteration 200 Training Loss: 1.977e-01 Loss in Target Net: 3.837e-02 |
2020-02-02 11:47:51 Iteration 250 Training Loss: 1.930e-01 Loss in Target Net: 3.762e-02 |
2020-02-02 11:48:09 Iteration 300 Training Loss: 1.885e-01 Loss in Target Net: 3.782e-02 |
2020-02-02 11:48:25 Iteration 350 Training Loss: 1.829e-01 Loss in Target Net: 3.293e-02 |
2020-02-02 11:48:44 Iteration 400 Training Loss: 1.790e-01 Loss in Target Net: 2.920e-02 |
Subsets and Splits