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.

Sign up for GitHub

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

Jump to bottom

[Performance] why so many bubbles between steps when running llava-one-vision? #2076

Open

sleepwalker2017 opened this issue Nov 18, 2024 · 0 comments

sleepwalker2017 commented Nov 18, 2024

Hi, I'm running llava-on-vision-7B using sglang on H100.

The GPU utilization is low, I can see the following timeline.

It seems a lot of time is wasted in memory management and data deserialization. Is that the fact?

I'm not familiar with sglang codes. I wonder what exactly is the bubbles doing? Thank you.

Attached is the profiling result.
1731919135.1918306.trace.json

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment