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

GEMM custom op enablement #3046

Closed
wants to merge 1 commit into from
Closed

Conversation

samanamp
Copy link

@samanamp samanamp commented Aug 28, 2024

Summary: This diff adds custom op wrappers around matmul_fp8_block. This should make them opaque to torch.compile and prevent issues where dynamo tries to trace triton code that is meant to be precompiled. I also add registration for fake kernels so that torch.compile can properly pass faketensors through the ops.

Differential Revision: D61800794

Summary:
Inspired by the following medium article, wanted to implement block-wise benchmarking to see if are getting any perf gains.

This diff, introduces block-wise as a custom op

Differential Revision: D61800794
Copy link

netlify bot commented Aug 28, 2024

Deploy Preview for pytorch-fbgemm-docs ready!

Name Link
🔨 Latest commit 3c40eff
🔍 Latest deploy log https://app.netlify.com/sites/pytorch-fbgemm-docs/deploys/66cf5ea02bea9400089ed926
😎 Deploy Preview https://deploy-preview-3046--pytorch-fbgemm-docs.netlify.app
📱 Preview on mobile
Toggle QR Code...

QR Code

Use your smartphone camera to open QR code link.

To edit notification comments on pull requests, go to your Netlify site configuration.

@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D61800794

@facebook-github-bot
Copy link
Contributor

This pull request has been merged in a9a3713.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants