[doc] feat: add MiniMax M2.5 / M2.7 model support

[yaoyu-33]

Summary

• Add MiniMax-M2.5 and M2.7 to the supported models list in README and docs
• Update bridge docstring and example README to note M2.5/M2.7 compatibility
• No code changes needed — M2.5 and M2.7 share the same model_type (minimax_m2) and MiniMaxM2ForCausalLM HF class as M2

Verification

Tested on cluster cw with real FP8 checkpoints (~215GB each) on 2 nodes (16x H100, TP=1 EP=16):

Model	Round-trip conversion	Inference (100 tokens)
MiniMax-M2.5	PASS (47,864 FP8 params skipped, all others matched)	PASS — coherent output
MiniMax-M2.7	PASS (47,864 FP8 params skipped, all others matched)	PASS — coherent output

M2.5 sample output:

Artificial intelligence (AI) refers to the simulation of human intelligence in machines that are programmed to think and learn like humans.

M2.7 sample output:

Artificial intelligence (AI) is a broad field of computer science focused on creating intelligent machines that can perform tasks that typically require human intelligence, such as reasoning, learning, problem-solving, perception, and language understanding.

Test plan

• Round-trip conversion verified on real M2.5 checkpoint (TP=1, EP=16)
• Round-trip conversion verified on real M2.7 checkpoint (TP=1, EP=16)
• Inference generates coherent text for M2.5
• Inference generates coherent text for M2.7

🤖 Generated with Claude Code

Summary by CodeRabbit

• Documentation
  • Updated supported models list and documentation to include MiniMax-M2.5 and MiniMax-M2.7 variants. These newer models are compatible with existing MiniMax-M2 configurations and can be used by specifying their respective model IDs.

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 421cbabc-c118-4507-9819-9f0d08ab0109

📥 Commits

Reviewing files that changed from the base of the PR and between 7ea8a45 and 90334a5.

📒 Files selected for processing (4)

• README.md
• docs/models/llm/README.md
• examples/models/minimax_m2/README.md
• src/megatron/bridge/models/minimax_m2/minimax_m2_bridge.py

---

📝 Walkthrough

Walkthrough

Documentation updates across multiple files to clarify that MiniMax-M2, M2.5, and M2.7 models share the same bridge architecture and can be used interchangeably. Changes include updated model lists, compatibility notes, and bridge class docstring updates.

Changes

Cohort / File(s)	Summary
MiniMax Model Documentation README.md, docs/models/llm/README.md, examples/models/minimax_m2/README.md, src/megatron/bridge/models/minimax_m2/minimax_m2_bridge.py	Updated model references and documentation to indicate MiniMax-M2, M2.5, and M2.7 use the same MiniMaxM2ForCausalLM architecture. Added compatibility note in examples and expanded model list entry with specification details.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3 minutes

🚥 Pre-merge checks | ✅ 4

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly describes the main change: adding MiniMax M2.5/M2.7 model support documentation updates across README and bridge files.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Test Results For Major Changes	✅ Passed	PR contains documentation-only changes to support MiniMax-M2.5 and M2.7 models with comprehensive testing results documented including round-trip conversion and inference verification on H100 clusters.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch yuya/add-minimax-m25-m27-support

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[yaoyu-33]

/ok to test ba75994