Fix adversarial image visualizer with canonical batches #227
Conversation
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
mzweilin
changed the title
mzweilin
changed the title
15 tasks
dxoigmn
approved these changes
Sep 28, 2023
| self.target = batch["target"] | ||
| # Save canonical input and target for on_train_end | ||
| self.input = batch[0] | ||
| self.target = batch[1] | ||
|
|
||
| def on_train_end(self, trainer, model): | ||
| # FIXME: We should really just save this to outputs instead of recomputing adv_input |
There was a problem hiding this comment.
Hopefully we do this one day...
Comment on lines
-33
to
+36
| visualizer.on_train_batch_end(trainer, model, outputs, batch, 0) | ||
| visualizer.on_train_end(trainer, model) | ||
| visualizer.on_train_batch_end(trainer, adversary, outputs, batch, 0) | ||
| visualizer.on_train_end(trainer, adversary) |
There was a problem hiding this comment.
Does this mean this no longer works with universal perturbations?
There was a problem hiding this comment.
Actually, this writes to disk (which is a bad idea). I wish this wrote to tensorboard. I swear there was a PR that did that...
There was a problem hiding this comment.
OK. I will revisit the visualizer later when I have a chance.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What does this PR do?
This PR fixes the adversarial image visualizer after canonicalizing adversarial batches.
Adversary.forward()to get adversarial examples.Type of change
Please check all relevant options.
Testing
Please describe the tests that you ran to verify your changes. Consider listing any relevant details of your test configuration.
pytestCUDA_VISIBLE_DEVICES=0 python -m mart experiment=CIFAR10_CNN_Adv trainer=gpu trainer.precision=16reports 70% (21 sec/epoch).CUDA_VISIBLE_DEVICES=0,1 python -m mart experiment=CIFAR10_CNN_Adv trainer=ddp trainer.precision=16 trainer.devices=2 model.optimizer.lr=0.2 trainer.max_steps=2925 datamodule.ims_per_batch=256 datamodule.world_size=2reports 70% (14 sec/epoch).Before submitting
pre-commit run -acommand without errorsDid you have fun?
Make sure you had fun coding 🙃