Support head dim 256#329
Closed
woct0rdho wants to merge 2 commits into
Closed
Conversation
woct0rdho
force-pushed
the
head_dim_256
branch
from
f6c9c5c to
c049753
Compare
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Head dim 256 is needed for models like AuraFlow. Recently new models based on AuraFlow are emerging and I'd like to help it get adopted.
In
transpose_pad_permute_cuda, we need to reduceCTA_SIZEon device to 32 so the assertCTA_SIZE * HEAD_DIM <= 8192is satisfied.I've tested on RTX 3090 (sm86) and RTX 4090 (sm89). Currently sm90 is not supported due to constraint of WGMMA size. Let me still put it here in case anyone needs it.