Issue Station
Sign In
Tools hereToolbox

vllm - ✅(Solved) Fix [Doc]: The --rope-scaling parameter has taken effect in vLLM supports YaRN [1 pull requests, 1 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
vllm-project/vllm#37886•Fetched 2026-04-08 01:17:21
View on GitHub
Comments
0
Participants
1
Timeline
1
Reactions
0
Author
labAxiaominglabAxiaoming
Participants
labAxiaominglabAxiaoming
Timeline (top)
labeled ×1

PR fix notes

PR #37430: [Docs] Add docs for context extension using the yarn method

Description (problem / solution / changelog)

<!-- markdownlint-disable -->

Purpose

add docs to demonstrates how to extend the context length of a Qwen model using the YARN method (rope_parameters) and run a simple chat example. ref https://github.com/vllm-project/vllm/issues/37886

Note: The --rope-scaling parameter used in older versions of vLLM is no longer supported. Please use the --hf-overrides method with rope_parameters instead.

Test Plan

Test Result

<details> <summary> Essential Elements of an Effective PR Description Checklist </summary>
  • The purpose of the PR, such as "Fix some issue (link existing issues this PR will resolve)".
  • The test plan, such as providing test command.
  • The test results, such as pasting the results comparison before and after, or e2e results
  • (Optional) The necessary documentation update, such as updating supported_models.md and examples for a new model.
  • (Optional) Release notes update. If your change is user facing, please update the release notes draft in the Google Doc.
</details>

Changed files

  • docs/features/context_extension.md (added, +63/-0)
RAW_BUFFERClick to expand / collapse

📚 The doc issue

vllm serve Qwen/Qwen3-8B --rope-scaling '{"rope_type":"yarn","factor":4.0 The --rope-scaling parameter has taken effect in vLLM supports YaRN, and the document can be modified. I submitted a documentation PR about Context Length, using --hf-overrides '{"rope_parameters": ... to replace --rope-scaling. Add docs for context extension using the yarn method

Suggest a potential alternative/fix

No response

Before submitting a new issue...

  • Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.

extent analysis

Fix Plan

To address the documentation issue with the --rope-scaling parameter, we will update the documentation to reflect the new --hf-overrides parameter.

Code Changes

Here are the steps to update the documentation:

  • Replace --rope-scaling with --hf-overrides in the documentation.
  • Update the JSON object to use rope_parameters instead of rope_type and factor.
### Updated Documentation
To use YaRN with vLLM, run the following command:
```bash
vllm serve Qwen/Qwen3-8B --hf-overrides '{"rope_parameters": {"rope_type": "yarn", "factor": 4.0}}'

Verification

To verify the fix, check the documentation for the correct usage of the --hf-overrides parameter.

Extra Tips

  • Make sure to test the updated documentation with the new --hf-overrides parameter.
  • Consider adding a note to the documentation to explain the change from --rope-scaling to --hf-overrides.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…
#model compatibility#GPU setup#container setup#orchestration issue#cache issue

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING