Skip to content

Conversation

tolgacangoz
Copy link
Contributor

@tolgacangoz tolgacangoz commented Jun 20, 2024

Proposes to fix #8477

Similar to #3673

@dg845 @sayakpaul @a-r-r-o-w

@tolgacangoz tolgacangoz marked this pull request as draft June 20, 2024 19:38
@tolgacangoz tolgacangoz marked this pull request as ready for review July 8, 2024 13:06
@a-r-r-o-w a-r-r-o-w requested a review from sayakpaul July 13, 2024 15:37
Copy link
Member

@sayakpaul sayakpaul left a comment

Choose a reason for hiding this comment

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

Thanks!

@HuggingFaceDocBuilderDev

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

@tolgacangoz
Copy link
Contributor Author

I realized there has already been an unwrap_model() function. I preferred it. It considers the torch.compile() case as well.

Copy link
Member

@a-r-r-o-w a-r-r-o-w left a comment

Choose a reason for hiding this comment

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

thanks! will merge once CI is green 🤗

@tolgacangoz tolgacangoz requested a review from a-r-r-o-w July 17, 2024 13:14
@tolgacangoz
Copy link
Contributor Author

This is OK now.

@a-r-r-o-w
Copy link
Member

failing tests are unrelated

@a-r-r-o-w a-r-r-o-w merged commit c1dc2ae into huggingface:main Jul 17, 2024
@tolgacangoz
Copy link
Contributor Author

Thanks for merging!

@tolgacangoz tolgacangoz deleted the fix-multi-gpu-ct branch July 17, 2024 13:35
@tolgacangoz tolgacangoz changed the title Fix multi-gpu case for train_cm_ct_unconditional.py Fix multi-gpu case for train_cm_ct_unconditional.py in terms of unwrapping unet model Jul 18, 2024
Disty0 pushed a commit to Disty0/diffusers that referenced this pull request Jul 18, 2024
* Fix multi-gpu case

* Prefer previously created `unwrap_model()` function

For `torch.compile()` generalizability

* `chore: update unwrap_model() function to use accelerator.unwrap_model()`
sayakpaul pushed a commit that referenced this pull request Dec 23, 2024
* Fix multi-gpu case

* Prefer previously created `unwrap_model()` function

For `torch.compile()` generalizability

* `chore: update unwrap_model() function to use accelerator.unwrap_model()`
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Error in Research Projects Consistency Training Script(DistributedDataParallel Error) line 1198
4 participants