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

Question: MLXClipTextEncoder run time #9

Open
jonny7737 opened this issue Nov 4, 2024 · 2 comments
Open

Question: MLXClipTextEncoder run time #9

jonny7737 opened this issue Nov 4, 2024 · 2 comments

Comments

@jonny7737
Copy link

Some times this node take 2 seconds to run and other times it takes 15-20 seconds to run.

Any idea why?

@thoddnn
Copy link
Owner

thoddnn commented Nov 5, 2024

Hey @jonny7737, can you send your workflow? Does the difference you see occur with the same prompt?

@jonny7737
Copy link
Author

I test by modifying the prompt text. I change 'a beautiful woman' to 'an old man'.

What I see is that one will clip encode in 2 - 3 seconds and the next one will take 15 - 25 ( sometimes 35) seconds. There does not seem to be a correlation between the prompt text and the time. It is, in my tests, related to the order in which the prompts are encoded: man first or woman first.

I hope this is clear.

MLX-Test.json

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

No branches or pull requests

2 participants