-
Notifications
You must be signed in to change notification settings - Fork 0
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
add paper: Data filtering for efficient adversarial training
- Loading branch information
ecchen
committed
Mar 8, 2024
1 parent
8117cba
commit 7ac6d97
Showing
15 changed files
with
694 additions
and
0 deletions.
There are no files selected for viewing
File renamed without changes.
File renamed without changes.
File renamed without changes.
File renamed without changes.
File renamed without changes.
File renamed without changes.
File renamed without changes.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,5 @@ | ||
| # Data filtering for efficient adversarial training | ||
|
|
||
| The code for the paper [Data filtering for efficient adversarial training](https://doi.org/10.1016/j.patcog.2024.110394). | ||
|
|
||
| The details will be released soon. |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,76 @@ | ||
| 2022-08-05 13:14:14.844 [INFO] eval_pytorch:<module>:59 | [ GPU ] CUDA_VISIBLE_DEVICES: 3 | ||
| 2022-08-05 13:14:14.844 [INFO] eval_pytorch:<module>:60 | [torch] Pytorch version: 1.12.0+cu113 | ||
| 2022-08-05 13:14:14.845 [INFO] eval_pytorch:<module>:61 | [param] Listing hyper-parameters... | ||
| 2022-08-05 13:14:14.845 [INFO] eval_pytorch:<module>:62 | [param] store_path : ../AWP/trades_AWP/model-best/ | ||
| 2022-08-05 13:14:14.845 [INFO] eval_pytorch:<module>:63 | [param] ckpt_name : ./model-best.pt | ||
| 2022-08-05 13:14:14.845 [INFO] eval_pytorch:<module>:64 | [param] norm : Linf | ||
| 2022-08-05 13:14:14.845 [INFO] eval_pytorch:<module>:65 | [param] epsilon : 0.031373 | ||
| 2022-08-05 13:14:14.846 [INFO] eval_pytorch:<module>:75 | [param] dataset : CIFAR10 | ||
| 2022-08-05 13:14:14.846 [INFO] eval_pytorch:<module>:76 | [param] model_depth: 34 | ||
| 2022-08-05 13:14:14.846 [INFO] eval_pytorch:<module>:77 | [param] model_widen: 10 | ||
| 2022-08-05 13:14:18.908 [WARNING] eval_pytorch:<module>:112 | using the original data | ||
| 2022-08-05 13:14:22.522 [INFO] eval_pytorch:<module>:135 | MD5 checksum: 274deb91b91e16a86466a5fc86a27b10 | ||
| 2022-08-05 13:14:22.733 [INFO] eval_pytorch:<module>:136 | SHA1 checksum: 9bb5639c033de0ecd58c567d533b584eb722060b | ||
| 2022-08-05 13:14:36.035 [INFO] other_utils:log:13 | initial accuracy: 86.54% | ||
| 2022-08-05 13:15:58.752 [INFO] other_utils:log:13 | apgd-ce - 1/18 - 149 out of 500 successfully perturbed | ||
| 2022-08-05 13:17:15.184 [INFO] other_utils:log:13 | apgd-ce - 2/18 - 132 out of 500 successfully perturbed | ||
| 2022-08-05 13:18:31.619 [INFO] other_utils:log:13 | apgd-ce - 3/18 - 136 out of 500 successfully perturbed | ||
| 2022-08-05 13:19:48.043 [INFO] other_utils:log:13 | apgd-ce - 4/18 - 147 out of 500 successfully perturbed | ||
| 2022-08-05 13:21:04.464 [INFO] other_utils:log:13 | apgd-ce - 5/18 - 160 out of 500 successfully perturbed | ||
| 2022-08-05 13:22:20.876 [INFO] other_utils:log:13 | apgd-ce - 6/18 - 146 out of 500 successfully perturbed | ||
| 2022-08-05 13:23:37.293 [INFO] other_utils:log:13 | apgd-ce - 7/18 - 149 out of 500 successfully perturbed | ||
| 2022-08-05 13:24:53.688 [INFO] other_utils:log:13 | apgd-ce - 8/18 - 136 out of 500 successfully perturbed | ||
| 2022-08-05 13:26:10.135 [INFO] other_utils:log:13 | apgd-ce - 9/18 - 150 out of 500 successfully perturbed | ||
| 2022-08-05 13:27:26.555 [INFO] other_utils:log:13 | apgd-ce - 10/18 - 153 out of 500 successfully perturbed | ||
| 2022-08-05 13:28:42.976 [INFO] other_utils:log:13 | apgd-ce - 11/18 - 144 out of 500 successfully perturbed | ||
| 2022-08-05 13:29:59.389 [INFO] other_utils:log:13 | apgd-ce - 12/18 - 140 out of 500 successfully perturbed | ||
| 2022-08-05 13:31:15.791 [INFO] other_utils:log:13 | apgd-ce - 13/18 - 159 out of 500 successfully perturbed | ||
| 2022-08-05 13:32:32.217 [INFO] other_utils:log:13 | apgd-ce - 14/18 - 130 out of 500 successfully perturbed | ||
| 2022-08-05 13:33:48.631 [INFO] other_utils:log:13 | apgd-ce - 15/18 - 165 out of 500 successfully perturbed | ||
| 2022-08-05 13:35:05.045 [INFO] other_utils:log:13 | apgd-ce - 16/18 - 144 out of 500 successfully perturbed | ||
| 2022-08-05 13:36:21.463 [INFO] other_utils:log:13 | apgd-ce - 17/18 - 149 out of 500 successfully perturbed | ||
| 2022-08-05 13:36:48.202 [INFO] other_utils:log:13 | apgd-ce - 18/18 - 41 out of 154 successfully perturbed | ||
| 2022-08-05 13:36:48.202 [INFO] other_utils:log:13 | robust accuracy after APGD-CE: 61.24% (total time 1332.2 s) | ||
| 2022-08-05 13:47:30.659 [INFO] other_utils:log:13 | apgd-t - 1/13 - 39 out of 500 successfully perturbed | ||
| 2022-08-05 13:58:19.215 [INFO] other_utils:log:13 | apgd-t - 2/13 - 25 out of 500 successfully perturbed | ||
| 2022-08-05 14:09:05.150 [INFO] other_utils:log:13 | apgd-t - 3/13 - 29 out of 500 successfully perturbed | ||
| 2022-08-05 14:19:53.440 [INFO] other_utils:log:13 | apgd-t - 4/13 - 25 out of 500 successfully perturbed | ||
| 2022-08-05 14:30:37.473 [INFO] other_utils:log:13 | apgd-t - 5/13 - 34 out of 500 successfully perturbed | ||
| 2022-08-05 14:41:15.466 [INFO] other_utils:log:13 | apgd-t - 6/13 - 36 out of 500 successfully perturbed | ||
| 2022-08-05 14:51:59.074 [INFO] other_utils:log:13 | apgd-t - 7/13 - 29 out of 500 successfully perturbed | ||
| 2022-08-05 15:02:33.482 [INFO] other_utils:log:13 | apgd-t - 8/13 - 41 out of 500 successfully perturbed | ||
| 2022-08-05 15:13:15.251 [INFO] other_utils:log:13 | apgd-t - 9/13 - 31 out of 500 successfully perturbed | ||
| 2022-08-05 15:23:52.353 [INFO] other_utils:log:13 | apgd-t - 10/13 - 34 out of 500 successfully perturbed | ||
| 2022-08-05 15:34:25.614 [INFO] other_utils:log:13 | apgd-t - 11/13 - 36 out of 500 successfully perturbed | ||
| 2022-08-05 15:45:09.937 [INFO] other_utils:log:13 | apgd-t - 12/13 - 27 out of 500 successfully perturbed | ||
| 2022-08-05 15:47:59.642 [INFO] other_utils:log:13 | apgd-t - 13/13 - 8 out of 124 successfully perturbed | ||
| 2022-08-05 15:47:59.643 [INFO] other_utils:log:13 | robust accuracy after APGD-T: 57.30% (total time 9203.6 s) | ||
| 2022-08-05 16:03:08.041 [INFO] other_utils:log:13 | fab-t - 1/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 16:18:16.559 [INFO] other_utils:log:13 | fab-t - 2/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 16:33:25.204 [INFO] other_utils:log:13 | fab-t - 3/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 16:48:34.252 [INFO] other_utils:log:13 | fab-t - 4/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 17:03:43.067 [INFO] other_utils:log:13 | fab-t - 5/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 17:18:52.190 [INFO] other_utils:log:13 | fab-t - 6/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 17:34:01.195 [INFO] other_utils:log:13 | fab-t - 7/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 17:49:09.736 [INFO] other_utils:log:13 | fab-t - 8/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 18:04:18.557 [INFO] other_utils:log:13 | fab-t - 9/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 18:19:27.193 [INFO] other_utils:log:13 | fab-t - 10/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 18:34:35.867 [INFO] other_utils:log:13 | fab-t - 11/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 18:41:38.436 [INFO] other_utils:log:13 | fab-t - 12/12 - 0 out of 230 successfully perturbed | ||
| 2022-08-05 18:41:38.436 [INFO] other_utils:log:13 | robust accuracy after FAB-T: 57.30% (total time 19622.4 s) | ||
| 2022-08-05 19:01:45.922 [INFO] other_utils:log:13 | square - 1/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 19:21:53.203 [INFO] other_utils:log:13 | square - 2/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 19:41:59.006 [INFO] other_utils:log:13 | square - 3/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 20:02:03.977 [INFO] other_utils:log:13 | square - 4/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 20:22:10.998 [INFO] other_utils:log:13 | square - 5/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 20:42:18.367 [INFO] other_utils:log:13 | square - 6/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 21:02:26.304 [INFO] other_utils:log:13 | square - 7/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 21:22:34.902 [INFO] other_utils:log:13 | square - 8/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 21:42:43.920 [INFO] other_utils:log:13 | square - 9/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 22:02:53.565 [INFO] other_utils:log:13 | square - 10/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 22:23:06.879 [INFO] other_utils:log:13 | square - 11/12 - 0 out of 500 successfully perturbed | ||
| 2022-08-05 22:32:27.113 [INFO] other_utils:log:13 | square - 12/12 - 0 out of 230 successfully perturbed | ||
| 2022-08-05 22:32:27.114 [INFO] other_utils:log:13 | robust accuracy after SQUARE: 57.30% (total time 33471.1 s) | ||
| 2022-08-05 22:32:27.116 [INFO] other_utils:log:13 | max Linf perturbation: 0.03137, nan in tensor: 0, max: 1.00000, min: 0.00000 | ||
| 2022-08-05 22:32:27.116 [INFO] other_utils:log:13 | robust accuracy: 57.30% | ||
|
|
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,90 @@ | ||
| import math | ||
| import torch | ||
| import torch.nn as nn | ||
| import torch.nn.functional as F | ||
|
|
||
|
|
||
| class BasicBlock(nn.Module): | ||
| def __init__(self, in_planes, out_planes, stride, dropRate=0.0): | ||
| super(BasicBlock, self).__init__() | ||
| self.bn1 = nn.BatchNorm2d(in_planes) | ||
| self.relu1 = nn.ReLU(inplace=True) | ||
| self.conv1 = nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride, | ||
| padding=1, bias=False) | ||
| self.bn2 = nn.BatchNorm2d(out_planes) | ||
| self.relu2 = nn.ReLU(inplace=True) | ||
| self.conv2 = nn.Conv2d(out_planes, out_planes, kernel_size=3, stride=1, | ||
| padding=1, bias=False) | ||
| self.droprate = dropRate | ||
| self.equalInOut = (in_planes == out_planes) | ||
| self.convShortcut = (not self.equalInOut) and nn.Conv2d(in_planes, out_planes, kernel_size=1, stride=stride, | ||
| padding=0, bias=False) or None | ||
|
|
||
| def forward(self, x): | ||
| if not self.equalInOut: | ||
| x = self.relu1(self.bn1(x)) | ||
| else: | ||
| out = self.relu1(self.bn1(x)) | ||
| out = self.relu2(self.bn2(self.conv1(out if self.equalInOut else x))) | ||
| if self.droprate > 0: | ||
| out = F.dropout(out, p=self.droprate, training=self.training) | ||
| out = self.conv2(out) | ||
| return torch.add(x if self.equalInOut else self.convShortcut(x), out) | ||
|
|
||
|
|
||
| class NetworkBlock(nn.Module): | ||
| def __init__(self, nb_layers, in_planes, out_planes, block, stride, dropRate=0.0): | ||
| super(NetworkBlock, self).__init__() | ||
| self.layer = self._make_layer(block, in_planes, out_planes, nb_layers, stride, dropRate) | ||
|
|
||
| def _make_layer(self, block, in_planes, out_planes, nb_layers, stride, dropRate): | ||
| layers = [] | ||
| for i in range(int(nb_layers)): | ||
| layers.append(block(i == 0 and in_planes or out_planes, out_planes, i == 0 and stride or 1, dropRate)) | ||
| return nn.Sequential(*layers) | ||
|
|
||
| def forward(self, x): | ||
| return self.layer(x) | ||
|
|
||
|
|
||
| class WideResNet(nn.Module): | ||
| def __init__(self, depth=34, num_classes=10, widen_factor=10, dropRate=0.0): | ||
| super(WideResNet, self).__init__() | ||
| nChannels = [16, 16 * widen_factor, 32 * widen_factor, 64 * widen_factor] | ||
| assert ((depth - 4) % 6 == 0) | ||
| n = (depth - 4) / 6 | ||
| block = BasicBlock | ||
| # 1st conv before any network block | ||
| self.conv1 = nn.Conv2d(3, nChannels[0], kernel_size=3, stride=1, | ||
| padding=1, bias=False) | ||
| # 1st block | ||
| self.block1 = NetworkBlock(n, nChannels[0], nChannels[1], block, 1, dropRate) | ||
| # 2nd block | ||
| self.block2 = NetworkBlock(n, nChannels[1], nChannels[2], block, 2, dropRate) | ||
| # 3rd block | ||
| self.block3 = NetworkBlock(n, nChannels[2], nChannels[3], block, 2, dropRate) | ||
| # global average pooling and classifier | ||
| self.bn1 = nn.BatchNorm2d(nChannels[3]) | ||
| self.relu = nn.ReLU(inplace=True) | ||
| self.fc = nn.Linear(nChannels[3], num_classes) | ||
| self.nChannels = nChannels[3] | ||
|
|
||
| for m in self.modules(): | ||
| if isinstance(m, nn.Conv2d): | ||
| n = m.kernel_size[0] * m.kernel_size[1] * m.out_channels | ||
| m.weight.data.normal_(0, math.sqrt(2. / n)) | ||
| elif isinstance(m, nn.BatchNorm2d): | ||
| m.weight.data.fill_(1) | ||
| m.bias.data.zero_() | ||
| elif isinstance(m, nn.Linear): | ||
| m.bias.data.zero_() | ||
|
|
||
| def forward(self, x): | ||
| out = self.conv1(x) | ||
| out = self.block1(out) | ||
| out = self.block2(out) | ||
| out = self.block3(out) | ||
| out = self.relu(self.bn1(out)) | ||
| out = F.avg_pool2d(out, 8) | ||
| out = out.view(-1, self.nChannels) | ||
| return self.fc(out) |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,73 @@ | ||
| 2022-02-11 05:32:33.098 [INFO] eval_pytorch:<module>:61 | [ GPU ] CUDA_VISIBLE_DEVICES: 2 | ||
| 2022-02-11 05:32:33.099 [INFO] eval_pytorch:<module>:62 | [torch] Pytorch version: 1.10.1+cu113 | ||
| 2022-02-11 05:32:33.099 [INFO] eval_pytorch:<module>:63 | [param] Listing hyper-parameters... | ||
| 2022-02-11 05:32:33.099 [INFO] eval_pytorch:<module>:64 | [param] dataset : CIFAR10 | ||
| 2022-02-11 05:32:33.099 [INFO] eval_pytorch:<module>:65 | [param] store_path : ../AWP/trades_AWP/model-best/ | ||
| 2022-02-11 05:32:33.099 [INFO] eval_pytorch:<module>:66 | [param] ckpt_name : ./model-best.pt | ||
| 2022-02-11 05:32:33.099 [INFO] eval_pytorch:<module>:67 | [param] norm : Linf | ||
| 2022-02-11 05:32:33.099 [INFO] eval_pytorch:<module>:68 | [param] epsilon : 0.031373 | ||
| 2022-02-11 05:32:37.689 [WARNING] eval_pytorch:<module>:113 | using the original data | ||
| 2022-02-11 05:33:00.337 [INFO] eval_pytorch:<module>:136 | MD5 checksum: 287f8e885bd44f748d21f7ad77eb7d11 | ||
| 2022-02-11 05:33:01.317 [INFO] eval_pytorch:<module>:137 | SHA1 checksum: 65cdd1d98e5132d683c29345ac2c5f3a8f33ae1a | ||
| 2022-02-11 05:33:21.838 [INFO] other_utils:log:13 | initial accuracy: 86.10% | ||
| 2022-02-11 05:37:31.347 [INFO] other_utils:log:13 | apgd-ce - 1/18 - 150 out of 500 successfully perturbed | ||
| 2022-02-11 05:41:35.934 [INFO] other_utils:log:13 | apgd-ce - 2/18 - 136 out of 500 successfully perturbed | ||
| 2022-02-11 05:45:40.689 [INFO] other_utils:log:13 | apgd-ce - 3/18 - 139 out of 500 successfully perturbed | ||
| 2022-02-11 05:49:45.250 [INFO] other_utils:log:13 | apgd-ce - 4/18 - 149 out of 500 successfully perturbed | ||
| 2022-02-11 05:53:49.968 [INFO] other_utils:log:13 | apgd-ce - 5/18 - 148 out of 500 successfully perturbed | ||
| 2022-02-11 05:57:54.800 [INFO] other_utils:log:13 | apgd-ce - 6/18 - 136 out of 500 successfully perturbed | ||
| 2022-02-11 06:01:59.610 [INFO] other_utils:log:13 | apgd-ce - 7/18 - 136 out of 500 successfully perturbed | ||
| 2022-02-11 06:06:04.420 [INFO] other_utils:log:13 | apgd-ce - 8/18 - 130 out of 500 successfully perturbed | ||
| 2022-02-11 06:10:09.260 [INFO] other_utils:log:13 | apgd-ce - 9/18 - 140 out of 500 successfully perturbed | ||
| 2022-02-11 06:14:13.939 [INFO] other_utils:log:13 | apgd-ce - 10/18 - 150 out of 500 successfully perturbed | ||
| 2022-02-11 06:18:18.689 [INFO] other_utils:log:13 | apgd-ce - 11/18 - 153 out of 500 successfully perturbed | ||
| 2022-02-11 06:22:23.327 [INFO] other_utils:log:13 | apgd-ce - 12/18 - 140 out of 500 successfully perturbed | ||
| 2022-02-11 06:26:27.981 [INFO] other_utils:log:13 | apgd-ce - 13/18 - 161 out of 500 successfully perturbed | ||
| 2022-02-11 06:30:32.668 [INFO] other_utils:log:13 | apgd-ce - 14/18 - 135 out of 500 successfully perturbed | ||
| 2022-02-11 06:34:37.528 [INFO] other_utils:log:13 | apgd-ce - 15/18 - 158 out of 500 successfully perturbed | ||
| 2022-02-11 06:38:42.350 [INFO] other_utils:log:13 | apgd-ce - 16/18 - 132 out of 500 successfully perturbed | ||
| 2022-02-11 06:42:47.087 [INFO] other_utils:log:13 | apgd-ce - 17/18 - 154 out of 500 successfully perturbed | ||
| 2022-02-11 06:43:43.047 [INFO] other_utils:log:13 | apgd-ce - 18/18 - 25 out of 110 successfully perturbed | ||
| 2022-02-11 06:43:43.047 [INFO] other_utils:log:13 | robust accuracy after APGD-CE: 61.38% (total time 4221.2 s) | ||
| 2022-02-11 07:18:15.151 [INFO] other_utils:log:13 | apgd-t - 1/13 - 34 out of 500 successfully perturbed | ||
| 2022-02-11 07:53:24.270 [INFO] other_utils:log:13 | apgd-t - 2/13 - 15 out of 500 successfully perturbed | ||
| 2022-02-11 08:27:58.437 [INFO] other_utils:log:13 | apgd-t - 3/13 - 29 out of 500 successfully perturbed | ||
| 2022-02-11 09:02:30.998 [INFO] other_utils:log:13 | apgd-t - 4/13 - 26 out of 500 successfully perturbed | ||
| 2022-02-11 09:36:43.824 [INFO] other_utils:log:13 | apgd-t - 5/13 - 31 out of 500 successfully perturbed | ||
| 2022-02-11 10:11:16.419 [INFO] other_utils:log:13 | apgd-t - 6/13 - 28 out of 500 successfully perturbed | ||
| 2022-02-11 10:45:30.324 [INFO] other_utils:log:13 | apgd-t - 7/13 - 26 out of 500 successfully perturbed | ||
| 2022-02-11 11:19:49.398 [INFO] other_utils:log:13 | apgd-t - 8/13 - 30 out of 500 successfully perturbed | ||
| 2022-02-11 11:54:18.904 [INFO] other_utils:log:13 | apgd-t - 9/13 - 24 out of 500 successfully perturbed | ||
| 2022-02-11 12:28:28.030 [INFO] other_utils:log:13 | apgd-t - 10/13 - 29 out of 500 successfully perturbed | ||
| 2022-02-11 13:02:40.310 [INFO] other_utils:log:13 | apgd-t - 11/13 - 27 out of 500 successfully perturbed | ||
| 2022-02-11 13:36:53.112 [INFO] other_utils:log:13 | apgd-t - 12/13 - 25 out of 500 successfully perturbed | ||
| 2022-02-11 13:47:00.036 [INFO] other_utils:log:13 | apgd-t - 13/13 - 5 out of 138 successfully perturbed | ||
| 2022-02-11 13:47:00.037 [INFO] other_utils:log:13 | robust accuracy after APGD-T: 58.09% (total time 29618.2 s) | ||
| 2022-02-11 14:35:54.853 [INFO] other_utils:log:13 | fab-t - 1/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-11 15:24:49.647 [INFO] other_utils:log:13 | fab-t - 2/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-11 16:13:44.217 [INFO] other_utils:log:13 | fab-t - 3/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-11 17:02:38.775 [INFO] other_utils:log:13 | fab-t - 4/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-11 17:51:34.339 [INFO] other_utils:log:13 | fab-t - 5/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-11 18:40:31.251 [INFO] other_utils:log:13 | fab-t - 6/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-11 19:29:28.811 [INFO] other_utils:log:13 | fab-t - 7/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-11 20:18:26.582 [INFO] other_utils:log:13 | fab-t - 8/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-11 21:07:27.107 [INFO] other_utils:log:13 | fab-t - 9/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-11 21:56:27.485 [INFO] other_utils:log:13 | fab-t - 10/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-11 22:45:28.873 [INFO] other_utils:log:13 | fab-t - 11/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-11 23:15:39.622 [INFO] other_utils:log:13 | fab-t - 12/12 - 0 out of 309 successfully perturbed | ||
| 2022-02-11 23:15:39.624 [INFO] other_utils:log:13 | robust accuracy after FAB-T: 58.09% (total time 63737.8 s) | ||
| 2022-02-12 00:24:18.455 [INFO] other_utils:log:13 | square - 1/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-12 01:32:58.302 [INFO] other_utils:log:13 | square - 2/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-12 02:41:37.447 [INFO] other_utils:log:13 | square - 3/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-12 03:50:16.588 [INFO] other_utils:log:13 | square - 4/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-12 04:58:54.419 [INFO] other_utils:log:13 | square - 5/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-12 06:07:33.118 [INFO] other_utils:log:13 | square - 6/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-12 07:16:11.055 [INFO] other_utils:log:13 | square - 7/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-12 08:24:53.178 [INFO] other_utils:log:13 | square - 8/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-12 09:33:26.522 [INFO] other_utils:log:13 | square - 9/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-12 10:42:00.672 [INFO] other_utils:log:13 | square - 10/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-12 11:50:32.460 [INFO] other_utils:log:13 | square - 11/12 - 0 out of 500 successfully perturbed | ||
| 2022-02-12 12:32:41.405 [INFO] other_utils:log:13 | square - 12/12 - 0 out of 309 successfully perturbed | ||
| 2022-02-12 12:32:41.406 [INFO] other_utils:log:13 | robust accuracy after SQUARE: 58.09% (total time 111559.6 s) | ||
| 2022-02-12 12:32:41.438 [INFO] other_utils:log:13 | max Linf perturbation: 0.03137, nan in tensor: 0, max: 1.00000, min: 0.00000 | ||
| 2022-02-12 12:32:41.438 [INFO] other_utils:log:13 | robust accuracy: 58.09% |
Oops, something went wrong.