Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Per sample grad correctness util #532

Closed

Conversation

psolikov
Copy link
Contributor

Types of changes

  • Bug fix (non-breaking change which fixes an issue)
  • New feature (non-breaking change which adds functionality)
  • Breaking change (fix or feature that would cause existing functionality to change)
  • Docs change / refactoring / dependency upgrade

Motivation and Context / Related issue

Implementation of the utility described in #484.

Refactored the code to avoid code duplicates.

How Has This Been Tested (if it applies)

Added the new utility as a test case for existing tests stored in tests.grad_samples.

Checklist

  • The documentation is up-to-date with the changes I made.
  • I have read the CONTRIBUTING document and completed the CLA (see CONTRIBUTING).
  • All tests passed, and additional code has been covered with new tests.

@facebook-github-bot facebook-github-bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Oct 28, 2022
@facebook-github-bot
Copy link
Contributor

@psolikov has updated the pull request. You must reimport the pull request before landing.

@facebook-github-bot
Copy link
Contributor

@facebook-github-bot has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

@facebook-github-bot
Copy link
Contributor

@psolikov has updated the pull request. You must reimport the pull request before landing.

@facebook-github-bot
Copy link
Contributor

@facebook-github-bot has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

Copy link
Contributor

@ffuuugor ffuuugor left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks a lot for the amazing work!

It would be really great, if you can also update the relevant tutorial to mention this util exists and how to use it (https://github.com/pytorch/opacus/blob/main/tutorials/guide_to_grad_sampler.ipynb)

As for the test errors, you can ignore all the coveralls. The rest - your unit tests are failing because functorch tests is run on the nightly pytorch - see an example how we handle it

opacus/utils/per_sample_gradients_utils.py Show resolved Hide resolved
opacus/utils/per_sample_gradients_utils.py Outdated Show resolved Hide resolved
opacus/utils/per_sample_gradients_utils.py Outdated Show resolved Hide resolved
opacus/utils/per_sample_gradients_utils.py Show resolved Hide resolved
opacus/utils/per_sample_gradients_utils.py Outdated Show resolved Hide resolved
opacus/utils/per_sample_gradients_utils.py Outdated Show resolved Hide resolved
opacus/utils/per_sample_gradients_utils.py Outdated Show resolved Hide resolved
opacus/utils/per_sample_gradients_utils.py Outdated Show resolved Hide resolved
opacus/tests/grad_samples/sequence_bias_test.py Outdated Show resolved Hide resolved
opacus/tests/grad_samples/sequence_bias_test.py Outdated Show resolved Hide resolved
Pavel Solikov and others added 10 commits October 31, 2022 17:17
Summary:
Pull Request resolved: pytorch#520

sr stands for sampling rate, which is now legacy code. Now, it's just sample_rate = 1 / len(data_loader). This has been fixed in the example, by setting batch size to be 60000 * 0.004 = 240 (thanks to https://github.com/ffuuugor for the clarification).

On another note, when running with DP, the following error is thrown
```
AttributeError: Can't pickle local object 'wrap_collate_with_empty.<locals>.collate'
```

For now, a temporary fix (based on IBM/Project_CodeNet#21 (comment)) is to make num_workers=0 in the dataset loaders. This commit does that.

Reviewed By: ffuuugor

Differential Revision: D40253037

fbshipit-source-id: 99984f8963a4efea6829d109bb81acff0e587c93
Summary:
*The investigation part for this PR was done by alexandresablayrolles, thanks for figuring out the reason the tests were failing*

## Background
Current implementation of functorch-based per sample gradients fails on modules which have both trainable non-recursive parameters and standard submodules, e.g. below
```
class LinearWithExtraParam(nn.Module):
    def __init__(self, in_features: int, out_features: int, hidden_dim: int = 8):
        super().__init__()
        self.fc = nn.Linear(in_features, hidden_dim)
        self.extra_param = nn.Parameter(torch.randn(hidden_dim, out_features))

    def forward(self, x):
        x = self.fc(x)
        x = x.matmul(self.extra_param)
        return x
```

The reason is - functorch hook actually computes gradients for recursive submodules too. The problem is, normal hooks are also attached to these submodules. GradSampleModule then sees two grad_sample tensors, thinks it needs to accumulate and adds them up together

## Solution(s)

There are essentially two ways we can fix this: either make functorch compute per sample gradients for non-recursive parameters only or don't attach normal hooks to submodules where the parent module is handled by functorch.

This diff implements the latter option (reasoning below), for demo purposes the former option can be seen in pytorch#531

For the pure code perspective the former option (let's call it "non-recursive functorch") is more appealing to me. It better fits the existing paradigm and matches normal hooks behaviour - all of the existing code only deals with the immediate non-recursive parameters.
However, it doesn't make much sense from the efficiency perspective. "non-recursive functorch" would do all the work to compute per-sample gradients for its submodules, only for them to be filtered out at the very last stage.
Alternative option (a.k.a. "functorch for subtrees") does involve a bit more convoluted

This has a noticeable effect on performance.
Below is the results of MNIST benchmarks with different configurations. I've tested this with different configurations, because at the end of the day, the impact on performance depends on how deep are subtrees

* Standard model- our model from MNIST example, standard layers only (2 conv + 2 linear). No overhead expected, functorch doesn't kick in
* Mid-level model - leaf nodes (two linear layers) have one extra param and are computed with functorch. Overhead: 2x Linear hook
* Extreme model - root model have one extra param and needs to be handled by functorch. Overhead: 2x linear hook + 2x conv hook

| Mode                               | non-recursive functorch | functorch for subtrees |
|:-----------------------:|:------------------------:|:-----------------------:|
| Standard model (CPU)  |  138s                                 | 136s                                |
| Standard model (GPU)  |  149s                                 | 150s                                |
| Mid-level model (CPU)  |  157s                                 | 150s                                |
| Mid-level model (GPU)  |  100s                                 | 97s                                 |
| Extreme model (CPU)    |  207s                                 | 172s                               |
| Extreme model (GPU)    |  101s                                 | 94s                                |

Pull Request resolved: pytorch#510

Reviewed By: alexandresablayrolles

Differential Revision: D39579487

Pulled By: ffuuugor

fbshipit-source-id: 1b089bd04ab110174a1f2ebb371380eb2ce76054
@facebook-github-bot
Copy link
Contributor

@psolikov has updated the pull request. You must reimport the pull request before landing.

Copy link
Contributor

@ffuuugor ffuuugor left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Great, thanks for updating the PR and addressing the comments!

I still have some comments (see inline), but we're definitely getting closer to landing this. Thanks again for your work!

P.S: Some of the test fails look related to your changes, please double check

opacus/tests/grad_samples/common.py Outdated Show resolved Hide resolved
tutorials/guide_to_grad_sampler.ipynb Show resolved Hide resolved
opacus/tests/grad_samples/dp_multihead_attention_test.py Outdated Show resolved Hide resolved
@facebook-github-bot
Copy link
Contributor

@psolikov has updated the pull request. You must reimport the pull request before landing.

@facebook-github-bot
Copy link
Contributor

@psolikov has updated the pull request. You must reimport the pull request before landing.

@facebook-github-bot
Copy link
Contributor

@psolikov has updated the pull request. You must reimport the pull request before landing.

@facebook-github-bot
Copy link
Contributor

@psolikov has updated the pull request. You must reimport the pull request before landing.

Pavel Solikov added 2 commits November 8, 2022 15:47
# Conflicts:
#	opacus/data_loader.py
#	opacus/tests/grad_samples/common.py
#	opacus/tests/grad_samples/conv1d_test.py
#	opacus/tests/grad_samples/conv2d_test.py
#	opacus/tests/grad_samples/embedding_test.py
#	opacus/tests/grad_samples/group_norm_test.py
#	opacus/tests/grad_samples/linear_test.py
@psolikov psolikov changed the base branch from ffuuugor_522 to main November 8, 2022 16:05
@facebook-github-bot
Copy link
Contributor

@psolikov has updated the pull request. You must reimport the pull request before landing.

@facebook-github-bot
Copy link
Contributor

@psolikov has updated the pull request. You must reimport the pull request before landing.

@facebook-github-bot
Copy link
Contributor

@facebook-github-bot has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

@facebook-github-bot
Copy link
Contributor

@psolikov has updated the pull request. You must reimport the pull request before landing.

@facebook-github-bot
Copy link
Contributor

@facebook-github-bot has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

Copy link
Contributor

@ffuuugor ffuuugor left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Cool! Thanks a lot for finalising the work!

@facebook-github-bot
Copy link
Contributor

@psolikov has updated the pull request. You must reimport the pull request before landing.

@facebook-github-bot
Copy link
Contributor

@facebook-github-bot has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator.

@facebook-github-bot
Copy link
Contributor

@ffuuugor merged this pull request in a0a31ba.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. Merged
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

3 participants