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.

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

Supports cpu without avx2 instruction set #1142

Closed
wsxiaoys opened this issue Dec 29, 2023 · 1 comment · May be fixed by #1716
Closed

Supports cpu without avx2 instruction set #1142

wsxiaoys opened this issue Dec 29, 2023 · 1 comment · May be fixed by #1716
Labels
enhancement New feature or request

Comments

@wsxiaoys
Copy link
Member

Starting from version 0.5.0, the switch to llama.cpp in inference means that, by default, Tabby is compiled with AVX2 support. Consequently, Tabby may not function properly on CPUs lacking AVX2 support.

To address this issue, we could explore potential improvements by offering a separately compiled binary or awaiting an upstream llama.cpp patch that enables dynamic dispatching.
@wsxiaoys wsxiaoys added the enhancement New feature or request label Dec 29, 2023
This was referenced Dec 29, 2023
Closed
Closed
@wsxiaoys wsxiaoys changed the title Supports avx2 in tabby Supports cpu without avx2 instruction set Dec 29, 2023
@wsxiaoys wsxiaoys mentioned this issue Feb 6, 2024
Closed
@pinbraerts pinbraerts mentioned this issue Mar 25, 2024
Open
@wsxiaoys
Copy link
Member Author

This is now possible by using http backend: https://tabby.tabbyml.com/docs/administration/model/

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
enhancement New feature or request
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

build(llama): manipulate cmake flags via cargo features pinbraerts/tabby
1 participant
@wsxiaoys