Skip to content

Conversation

@avtc
Copy link
Contributor

@avtc avtc commented Aug 24, 2025

And probably other glm4_moe models.

I have quantized with this definition the GLM-4.5-Air model with padding to support tp8 and it works in vllm with tp8.

@Qubitium
Copy link
Collaborator

@avtc LTGM and great work! I really appreciate you adding comments (documentation) to the layers def. This will help future model defs.

@Qubitium Qubitium merged commit 74e1d00 into ModelCloud:main Aug 25, 2025
1 check passed
@avtc avtc deleted the feature/glm4_moe branch September 29, 2025 07:50
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