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

Feat (gptq): optimizing CPU to GPU memory transfer #1009

Merged
merged 2 commits into from
Sep 12, 2024

Conversation

i-colbert
Copy link
Collaborator

No description provided.

@i-colbert i-colbert requested review from Giuseppe5 and removed request for Giuseppe5 August 27, 2024 01:12
@nickfraser
Copy link
Collaborator

I'm slightly worried about this messing with our interop with HuggingFace accelerate. Would you test this in a multi-GPU setup with accelerate? Easiest way is to use this: https://github.com/huggingface/optimum-amd/tree/main/examples/quantization/brevitas

@nickfraser
Copy link
Collaborator

I ran a small multi-GPU test with accelerate and this seems to work.

@Giuseppe5 Giuseppe5 merged commit 10dcee3 into Xilinx:dev Sep 12, 2024
337 checks passed
@i-colbert i-colbert deleted the feat/gptq branch September 12, 2024 21:45
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.

3 participants