Add FlashMLA #74
Open
Add FlashMLA #74
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
ivarflakstad
reviewed
Mar 3, 2025
Comment on lines
+61
to
+64
| let compute_cap = compute_cap()?; | ||
| // assert compute cap is sm90 | ||
| // TODO TODO TODO | ||
| // assert!(compute_cap == 90, "Compute capability must be 90 (90a)"); |
There was a problem hiding this comment.
If you want an alternative approach I'm pretty sure you could use get_device_prop from cudarc.
The returned sys::cudaDeviceProp will contain major and minor fields.
There was a problem hiding this comment.
I think this would require us to "assume" ordinal 0, which is probably fine. Also, maybe there is a case for using the output of nvidia-smi because it's what we usually do 🤔? Not sure what is best though.
There was a problem hiding this comment.
Yeah it shouldn't matter. I was just thinking it was simpler since we already have cudarc available.
I assume the bound C functions are the same that nvidia-smi use.
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Outdated suggestions cannot be applied.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.
https://github.com/deepseek-ai/FlashMLA