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] Overflow risks when batch size and sequence length grows extremely large #596

Open
rchardx opened this issue Nov 8, 2024 · 1 comment

Comments

@rchardx
Copy link

rchardx commented Nov 8, 2024

For example, when batch size is 128, sequence length is like 32K, is there any possibility for FlashInfer's internal computation overflowing?

@yzh119
Copy link
Collaborator

yzh119 commented Nov 8, 2024

I'm not aware of that.

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