feat: add MiniMax-M2.5 model support#443
Open
octo-patch wants to merge 1 commit intoGradientHQ:mainfrom
Open
Conversation
Add MiniMax-M2.5, the latest open-weight model in the MiniMax-M2 family (456B total / 46B active parameters, 1M token context), to the supported models list. M2.5 uses the same MiniMaxM2ForCausalLM architecture as M2 and M2.1, so no model code changes are needed. Changes: - Add MiniMaxAI/MiniMax-M2.5 -> mlx-community/MiniMax-M2.5-4bit in MODELS - Update README model table to list M2.5 as the latest in the family
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Add MiniMax-M2.5 to the supported models list. M2.5 is the latest open-weight model in the MiniMax-M2 family:
MiniMaxM2ForCausalLMarchitecture as M2 and M2.1Changes
MiniMaxAI/MiniMax-M2.5→mlx-community/MiniMax-M2.5-4bitentry to theMODELSdict instatic_config.pyWhy no code changes needed
M2.5 shares the same architecture (
MiniMaxM2ForCausalLM) as M2 and M2.1, so the existing model code (src/parallax/models/minimax.py) and sglang monkey patches (minimax_m2_model.py) work without modification.Test plan
MiniMaxM2ForCausalLMarchitecture (same as M2/M2.1)mlx-community/MiniMax-M2.5-4bitexists on HuggingFace