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

[WIP][Feature] Implement native fused MoE layer #121

Closed
wants to merge 2 commits into from

Conversation

yiz-liu
Copy link

@yiz-liu yiz-liu commented Feb 20, 2025

What this PR does / why we need it?

Enable Expert-Parallel for ascend devices.

Does this PR introduce any user-facing change?

Enable EP
export VLLM_TEST_ENABLE_EP=1

Disable MLA (not supported right now)
export VLLM_MLA_DISABLE=1

How was this patch tested?

git clone https://github.com/yiz-liu/vllm.git
cd vllm
VLLM_TARGET_DEVICE=empty pip install -v -e .

Then install vllm-ascend with this PR.

@yiz-liu yiz-liu changed the title [Feature] Implement native fused MoE layer [WIP][Feature] Implement native fused MoE layer Feb 20, 2025
@wangxiyuan
Copy link
Collaborator

see: #127

@wangxiyuan wangxiyuan closed this Feb 21, 2025
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