Skip to content

[model] add Qwen-AgentWorld-35B-A3B support#10615

Merged
hiyouga merged 1 commit into
hiyouga:mainfrom
xiaoyesoso:feat/qwen-agentworld-support
Jun 30, 2026
Merged

[model] add Qwen-AgentWorld-35B-A3B support#10615
hiyouga merged 1 commit into
hiyouga:mainfrom
xiaoyesoso:feat/qwen-agentworld-support

Conversation

@xiaoyesoso

Copy link
Copy Markdown
Contributor

What does this PR do?

Add built-in support for Qwen-AgentWorld-35B-A3B, a native language world model released by the Qwen team, based on Qwen3.5-35B-A3B (MoE, 35B total / 3B active, 256K context).

  • Register Qwen-AgentWorld-35B-A3B-Thinking in src/llamafactory/extras/constants.py, with both Hugging Face (Qwen/Qwen-AgentWorld-35B-A3B) and ModelScope (Qwen/Qwen-AgentWorld-35B-A3B) download sources.
  • Bind it to the existing qwen3 chat template — the model shares the standard Qwen3 chat format and its inference uses --reasoning-parser qwen3, so no new template is required.
  • The -Thinking suffix is used so the register_model_group helper automatically populates DEFAULT_TEMPLATE.

Reference:

Fixes #10614

Before submitting

  • Did you read the contributor guideline?
  • Did you write any new necessary tests? (No new tests added: this PR only adds an entry to the model registry that reuses the existing qwen3 template; verified locally with ruff check / format and a runtime registry smoke check.)

Register Qwen-AgentWorld-35B-A3B (with ModelScope mirror) as a
Thinking variant under the qwen3 chat template. Qwen-AgentWorld is a
native language world model based on Qwen3.5-35B-A3B (MoE).

Reference: https://github.com/QwenLM/Qwen-AgentWorld

@gemini-code-assist gemini-code-assist Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

Code Review

This pull request registers the new model group "Qwen-AgentWorld-35B-A3B-Thinking" in constants.py. Feedback recommends moving this registration to its correct alphabetical position (before Qwen2) and updating the template from "qwen3" to "qwen3_5" to ensure compatibility with Qwen3.5-specific features.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

Comment thread src/llamafactory/extras/constants.py
@hiyouga hiyouga merged commit c383c0d into hiyouga:main Jun 30, 2026
15 of 16 checks passed
@hiyouga hiyouga added the solved This problem has been already solved label Jun 30, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

solved This problem has been already solved

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[Feature] Add support for Qwen-AgentWorld-35B-A3B

2 participants