Skip to content

docs: document Qwen3-Next balance serve build - #2137

Open
zcxGGmu wants to merge 2 commits into
kvcache-ai:mainfrom
zcxGGmu:docs/qwen3-next-balance-serve-build
Open

docs: document Qwen3-Next balance serve build#2137
zcxGGmu wants to merge 2 commits into
kvcache-ai:mainfrom
zcxGGmu:docs/qwen3-next-balance-serve-build

Conversation

@zcxGGmu

@zcxGGmu zcxGGmu commented Aug 5, 2026

Copy link
Copy Markdown

Summary

  • Add the required USE_BALANCE_SERVE=1 bash ./install.sh build step to the Qwen3-Next tutorial.
  • Explain that ModuleNotFoundError: No module named 'sched_ext' means the balance_serve scheduler extension was not built in the active environment.

Refs #1514

Test Plan

  • python documentation guard check for the new Qwen3-Next balance_serve build guidance
  • git diff --check HEAD~1..HEAD

@greptile-apps

greptile-apps Bot commented Aug 5, 2026

Copy link
Copy Markdown

Greptile Summary

This PR adds inline build steps for enabling the balance_serve extension (USE_BALANCE_SERVE=1 bash ./install.sh) to the Qwen3-Next tutorial, and adds a troubleshooting note for the ModuleNotFoundError: No module named 'sched_ext' error that appears when the extension was not compiled.

  • The install block (git clonecd ktransformersgit submodule updateUSE_BALANCE_SERVE=1 bash ./install.shcd ..) is technically correct; the trailing cd .. restores the working directory so Section 4's python ktransformers/server/main.py invocation works from the parent directory.
  • The original Installation Guide link is preserved as a prerequisite cross-reference, and the troubleshooting paragraph points users at the likely root cause of the missing module error.

Confidence Score: 5/5

Documentation-only change that adds a missing build step and a troubleshooting note; no code is affected.

The change is confined to a single markdown file. The added shell commands are technically correct, the working-directory issue from the prior review is resolved with cd .., and the Installation Guide link is preserved. No logic, configuration, or runtime behavior is touched.

Files Needing Attention: No files require special attention.

Important Files Changed

Filename Overview
doc/en/Qwen3-Next.md Adds USE_BALANCE_SERVE=1 build block and sched_ext troubleshooting note to Section 3; working-directory and Installation Guide link concerns from prior review are both addressed.

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A[Reader starts Qwen3-Next setup] --> B[Section 2: Download model via huggingface-cli]
    B --> C[Section 3: Check prerequisites via Installation Guide]
    C --> D["git clone + git submodule update --init --recursive"]
    D --> E["USE_BALANCE_SERVE=1 bash ./install.sh"]
    E --> F{Build succeeded?}
    F -- Yes --> G["cd .. (restore parent dir)"]
    F -- No / ModuleNotFoundError: sched_ext --> H[Re-run install in clean environment]
    H --> E
    G --> I["Section 4: python ktransformers/server/main.py --backend_type balance_serve"]
    I --> J[Section 5: curl /v1/chat/completions]
Loading

Reviews (2): Last reviewed commit: "[docs]: align Qwen3-Next install flow" | Re-trigger Greptile

Comment thread doc/en/Qwen3-Next.md
Comment thread doc/en/Qwen3-Next.md
@zcxGGmu

zcxGGmu commented Aug 5, 2026

Copy link
Copy Markdown
Author

Thanks for the review — I pushed 65e9d45c79f8 addressing the Qwen3-Next install flow feedback.

Changes made:

  • Restored the Installation Guide cross-reference before the inline build steps.
  • Added cd .. after the source install so the following server launch command runs from the expected parent directory.

Validation:

  • python content guard — passed, confirmed the prerequisite link and working-directory step are present.
  • git diff --check — passed.

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