Skip to content

support minimax_m2 ignore layer: block_sparse_moe.gate#1508

Open
xin3he wants to merge 1 commit intomainfrom
xin3he-patch-1
Open

support minimax_m2 ignore layer: block_sparse_moe.gate#1508
xin3he wants to merge 1 commit intomainfrom
xin3he-patch-1

Conversation

@xin3he
Copy link
Contributor

@xin3he xin3he commented Mar 6, 2026

Description

support minimax_m2 ignore layer: block_sparse_moe.gate

Type of Change

  • Bug fix
  • New feature
  • Documentation update
  • Performance improvement
  • Code refactoring
  • Other (please specify):

Related Issues

Fixes or relates to #

Checklist Before Submitting

  • My code has been tested locally.
  • Documentation has been updated as needed.
  • New or updated tests are included where applicable.

Copy link
Contributor

Copilot AI left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

Adds a predefined ignore-layer rule so AutoRound will keep the MiniMax M2 MoE gate layer in full precision during quantization, aligning behavior with models whose gate weights are directly referenced and can cause issues when quantized.

Changes:

  • Register a new predefined ignore rule for model.config.model_type == "minimax_m2".
  • Add "block_sparse_moe.gate" to the ignore-layers list for that model type.

You can also share your feedback on Copilot code review. Take the survey.

@chensuyue chensuyue added this to the 0.10.3 milestone Mar 7, 2026
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.

3 participants