text
stringlengths 5
1.13k
|
|---|
2020-02-04 00:32:28 Iteration 0 Training Loss: 9.612e-01 Loss in Target Net: 1.076e+00 |
2020-02-04 00:35:38 Iteration 50 Training Loss: 2.282e-01 Loss in Target Net: 1.253e-01 |
2020-02-04 00:38:48 Iteration 100 Training Loss: 1.985e-01 Loss in Target Net: 8.299e-02 |
2020-02-04 00:42:00 Iteration 150 Training Loss: 1.873e-01 Loss in Target Net: 8.526e-02 |
2020-02-04 00:45:11 Iteration 200 Training Loss: 1.793e-01 Loss in Target Net: 7.569e-02 |
2020-02-04 00:48:23 Iteration 250 Training Loss: 1.748e-01 Loss in Target Net: 7.862e-02 |
2020-02-04 00:51:31 Iteration 300 Training Loss: 1.714e-01 Loss in Target Net: 6.784e-02 |
2020-02-04 00:54:42 Iteration 350 Training Loss: 1.725e-01 Loss in Target Net: 7.365e-02 |
2020-02-04 00:57:54 Iteration 400 Training Loss: 1.667e-01 Loss in Target Net: 7.830e-02 |
2020-02-04 01:01:06 Iteration 450 Training Loss: 1.690e-01 Loss in Target Net: 7.230e-02 |
2020-02-04 01:04:17 Iteration 500 Training Loss: 1.669e-01 Loss in Target Net: 7.018e-02 |
2020-02-04 01:07:29 Iteration 550 Training Loss: 1.656e-01 Loss in Target Net: 7.834e-02 |
2020-02-04 01:10:41 Iteration 600 Training Loss: 1.679e-01 Loss in Target Net: 6.863e-02 |
2020-02-04 01:13:52 Iteration 650 Training Loss: 1.665e-01 Loss in Target Net: 7.421e-02 |
2020-02-04 01:17:04 Iteration 700 Training Loss: 1.635e-01 Loss in Target Net: 6.784e-02 |
2020-02-04 01:20:16 Iteration 750 Training Loss: 1.677e-01 Loss in Target Net: 6.483e-02 |
2020-02-04 01:23:28 Iteration 800 Training Loss: 1.639e-01 Loss in Target Net: 7.447e-02 |
2020-02-04 01:26:40 Iteration 850 Training Loss: 1.660e-01 Loss in Target Net: 7.217e-02 |
2020-02-04 01:29:52 Iteration 900 Training Loss: 1.640e-01 Loss in Target Net: 7.168e-02 |
2020-02-04 01:33:04 Iteration 950 Training Loss: 1.648e-01 Loss in Target Net: 6.945e-02 |
2020-02-04 01:36:16 Iteration 1000 Training Loss: 1.603e-01 Loss in Target Net: 6.431e-02 |
2020-02-04 01:39:29 Iteration 1050 Training Loss: 1.605e-01 Loss in Target Net: 5.960e-02 |
2020-02-04 01:42:42 Iteration 1100 Training Loss: 1.620e-01 Loss in Target Net: 6.086e-02 |
2020-02-04 01:45:54 Iteration 1150 Training Loss: 1.632e-01 Loss in Target Net: 6.443e-02 |
2020-02-04 01:49:06 Iteration 1200 Training Loss: 1.617e-01 Loss in Target Net: 6.589e-02 |
2020-02-04 01:52:18 Iteration 1250 Training Loss: 1.645e-01 Loss in Target Net: 6.490e-02 |
2020-02-04 01:55:33 Iteration 1300 Training Loss: 1.585e-01 Loss in Target Net: 7.363e-02 |
2020-02-04 01:58:45 Iteration 1350 Training Loss: 1.623e-01 Loss in Target Net: 6.972e-02 |
2020-02-04 02:01:57 Iteration 1400 Training Loss: 1.607e-01 Loss in Target Net: 6.328e-02 |
2020-02-04 02:05:09 Iteration 1450 Training Loss: 1.629e-01 Loss in Target Net: 6.163e-02 |
2020-02-04 02:08:15 Iteration 1499 Training Loss: 1.663e-01 Loss in Target Net: 6.486e-02 |
Evaluating against victims networks |
DPN92 |
Using Adam for retraining |
Files already downloaded and verified |
2020-02-04 02:09:27, Epoch 0, Iteration 7, loss 0.201 (0.285), acc 94.231 (92.800) |
2020-02-04 02:14:29, 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:[-0.39842477, -0.4590174, -1.5002476, -3.5157194, -0.5763492, -4.503169, -0.8914132, -1.8805921, 14.159386, 0.24538963], Poisons' Predictions:[8, 8, 8, 8, 8] |
2020-02-04 02:20:03 Epoch 59, Val iteration 0, acc 93.400 (93.400) |
2020-02-04 02:20:52 Epoch 59, Val iteration 19, acc 93.400 (92.900) |
* Prec: 92.90000228881836 |
-------- |
------SUMMARY------ |
TIME ELAPSED (mins): 96 |
TARGET INDEX: 5 |
DPN92 1 |
Namespace(chk_path='chk-black-end2end', chk_subdir='poisons', device='cuda', dset_path='datasets', end2end=True, eval_poison_path='', gpu='6', lr_decay_epoch=[30, 45], mode='mean', model_resume_path='model-chks', nearest=False, net_repeat=3, 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=6, 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-3Repeat/1500/6 |
Selected base image indices: [213, 225, 227, 247, 249] |
2020-02-04 00:32:33 Iteration 0 Training Loss: 9.941e-01 Loss in Target Net: 1.328e+00 |
2020-02-04 00:35:53 Iteration 50 Training Loss: 1.959e-01 Loss in Target Net: 3.794e-02 |
2020-02-04 00:39:19 Iteration 100 Training Loss: 1.688e-01 Loss in Target Net: 2.619e-02 |
2020-02-04 00:42:49 Iteration 150 Training Loss: 1.574e-01 Loss in Target Net: 2.776e-02 |
2020-02-04 00:46:17 Iteration 200 Training Loss: 1.503e-01 Loss in Target Net: 2.587e-02 |
2020-02-04 00:49:43 Iteration 250 Training Loss: 1.488e-01 Loss in Target Net: 2.632e-02 |
2020-02-04 00:53:05 Iteration 300 Training Loss: 1.475e-01 Loss in Target Net: 2.801e-02 |
2020-02-04 00:56:31 Iteration 350 Training Loss: 1.468e-01 Loss in Target Net: 2.422e-02 |
2020-02-04 00:59:57 Iteration 400 Training Loss: 1.451e-01 Loss in Target Net: 2.156e-02 |
2020-02-04 01:03:23 Iteration 450 Training Loss: 1.457e-01 Loss in Target Net: 2.088e-02 |
2020-02-04 01:06:50 Iteration 500 Training Loss: 1.403e-01 Loss in Target Net: 2.187e-02 |
2020-02-04 01:10:17 Iteration 550 Training Loss: 1.425e-01 Loss in Target Net: 2.150e-02 |
2020-02-04 01:13:44 Iteration 600 Training Loss: 1.410e-01 Loss in Target Net: 2.699e-02 |
2020-02-04 01:17:11 Iteration 650 Training Loss: 1.413e-01 Loss in Target Net: 2.537e-02 |
2020-02-04 01:20:38 Iteration 700 Training Loss: 1.395e-01 Loss in Target Net: 2.541e-02 |
2020-02-04 01:24:05 Iteration 750 Training Loss: 1.375e-01 Loss in Target Net: 2.290e-02 |
2020-02-04 01:27:33 Iteration 800 Training Loss: 1.392e-01 Loss in Target Net: 2.202e-02 |
2020-02-04 01:31:01 Iteration 850 Training Loss: 1.387e-01 Loss in Target Net: 1.970e-02 |
2020-02-04 01:34:28 Iteration 900 Training Loss: 1.400e-01 Loss in Target Net: 2.063e-02 |
2020-02-04 01:37:55 Iteration 950 Training Loss: 1.402e-01 Loss in Target Net: 2.499e-02 |
2020-02-04 01:41:23 Iteration 1000 Training Loss: 1.388e-01 Loss in Target Net: 2.351e-02 |
2020-02-04 01:44:50 Iteration 1050 Training Loss: 1.380e-01 Loss in Target Net: 2.719e-02 |
2020-02-04 01:48:18 Iteration 1100 Training Loss: 1.387e-01 Loss in Target Net: 2.286e-02 |
2020-02-04 01:51:46 Iteration 1150 Training Loss: 1.377e-01 Loss in Target Net: 2.578e-02 |
2020-02-04 01:55:13 Iteration 1200 Training Loss: 1.392e-01 Loss in Target Net: 2.887e-02 |
2020-02-04 01:58:40 Iteration 1250 Training Loss: 1.381e-01 Loss in Target Net: 2.547e-02 |
2020-02-04 02:02:07 Iteration 1300 Training Loss: 1.407e-01 Loss in Target Net: 3.046e-02 |
2020-02-04 02:05:33 Iteration 1350 Training Loss: 1.378e-01 Loss in Target Net: 2.615e-02 |
2020-02-04 02:08:54 Iteration 1400 Training Loss: 1.366e-01 Loss in Target Net: 2.426e-02 |
2020-02-04 02:12:34 Iteration 1450 Training Loss: 1.374e-01 Loss in Target Net: 2.816e-02 |
2020-02-04 02:16:15 Iteration 1499 Training Loss: 1.380e-01 Loss in Target Net: 2.596e-02 |
Evaluating against victims networks |
DPN92 |
Using Adam for retraining |
Files already downloaded and verified |
2020-02-04 02:17:24, Epoch 0, Iteration 7, loss 0.368 (0.529), acc 90.385 (88.800) |
2020-02-04 02:22:39, 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:[-1.4799988, 0.9696593, -3.2822611, -4.157518, 0.22720115, -1.0739338, 3.5389283, -2.147123, 9.219605, -1.5845889], Poisons' Predictions:[8, 8, 8, 8, 8] |
2020-02-04 02:28:22 Epoch 59, Val iteration 0, acc 93.600 (93.600) |
2020-02-04 02:29:13 Epoch 59, Val iteration 19, acc 93.200 (93.150) |
* Prec: 93.15000190734864 |
-------- |
------SUMMARY------ |
TIME ELAPSED (mins): 104 |
TARGET INDEX: 6 |
DPN92 1 |
Namespace(chk_path='chk-black-end2end', chk_subdir='poisons', device='cuda', dset_path='datasets', end2end=True, eval_poison_path='', gpu='7', lr_decay_epoch=[30, 45], mode='mean', model_resume_path='model-chks', nearest=False, net_repeat=3, 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=7, 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-3Repeat/1500/7 |
Selected base image indices: [213, 225, 227, 247, 249] |
2020-02-04 00:32:42 Iteration 0 Training Loss: 1.082e+00 Loss in Target Net: 1.326e+00 |
2020-02-04 00:35:59 Iteration 50 Training Loss: 2.740e-01 Loss in Target Net: 1.497e-01 |
Subsets and Splits