Add xformers attention to VAE #1507
Merged
Conversation
This file contains hidden or 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
|
The documentation is not available anymore as the PR was closed or merged. |
patil-suraj
reviewed
Dec 2, 2022
|
Hi @patil-suraj , thanks for the help! I applied the changes and it's a lot simpler now :) |
patil-suraj
approved these changes
Dec 3, 2022
|
Tried it out, works really well, merging! |
tcapelle
pushed a commit
to tcapelle/diffusers
that referenced
this pull request
Dec 12, 2022
* Add xformers attention to VAE * Simplify VAE xformers code * Update src/diffusers/models/attention.py Co-authored-by: Ilmari Heikkinen <ilmari@fhtr.org> Co-authored-by: Suraj Patil <surajp815@gmail.com>
sliard
pushed a commit
to sliard/diffusers
that referenced
this pull request
Dec 21, 2022
* Add xformers attention to VAE * Simplify VAE xformers code * Update src/diffusers/models/attention.py Co-authored-by: Ilmari Heikkinen <ilmari@fhtr.org> Co-authored-by: Suraj Patil <surajp815@gmail.com>
yoonseokjin
pushed a commit
to yoonseokjin/diffusers
that referenced
this pull request
Dec 25, 2023
* Add xformers attention to VAE * Simplify VAE xformers code * Update src/diffusers/models/attention.py Co-authored-by: Ilmari Heikkinen <ilmari@fhtr.org> Co-authored-by: Suraj Patil <surajp815@gmail.com>
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Outdated suggestions cannot be applied.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.
This PR adds xformers memory-efficient attention to the VAE to make the attention parts of the VAE require less memory.
This can help with issues like #1434 to some extent, but another part of the problem are the convolution layers. Running the VAE in channels_last memory format helps, but you're still dealing with RAM use in the tens of GB . At least it's not hundreds of GB. To run the VAE on low VRAM, switch the up_blocks / down_blocks to run on the CPU.