Skip to content

Conversation

@jlamypoirier
Copy link
Collaborator

@jlamypoirier jlamypoirier commented Jan 9, 2026

✨ Description

Allow running most of Fast-LLM with CPU only. This will speed up debugging (do most of it locally) and allow running a lot more tests with github actions.

Most test now run on cpu only, with the exception of triton, distributed, megatron and mixers with gpu-only implementation. Distributed tests could probably also run on cpu (with gloo), but that's low priority and would conflict with #440.

@jlamypoirier jlamypoirier marked this pull request as ready for review January 10, 2026 03:58
@jlamypoirier jlamypoirier requested a review from tscholak January 10, 2026 04:02
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

Successfully merging this pull request may close these issues.

2 participants