Skip to content

fix: move WhisperX off yanked release#27

Merged
MapleEve merged 1 commit intomainfrom
fix/whisperx-yanked
May 6, 2026
Merged

fix: move WhisperX off yanked release#27
MapleEve merged 1 commit intomainfrom
fix/whisperx-yanked

Conversation

@MapleEve
Copy link
Copy Markdown
Owner

@MapleEve MapleEve commented May 6, 2026

Summary

  • move WhisperX from the yanked 3.1.x package series to whisperx==3.3.1
  • pin compatible pyannote/audio-adjacent dependencies that stay on the current numpy<2 and SciPy 1.11.x baseline
  • document the dependency change in both changelogs

Why 3.3.1

  • WhisperX 3.1.5 and 3.1.6 are yanked on PyPI
  • 3.3.1 is non-yanked and still works with the current torch 2.6 / numpy<2 dependency lane
  • 3.3.3+ moves to numpy>=2, and the latest 3.8.x line also moves to a newer torch / pyannote stack, so that should be a separate dependency-platform upgrade

Validation

  • public release scan passed
  • git diff --check passed
  • targeted unit tests passed: tests/unit/test_provider_registry.py, tests/unit/test_pipeline_model_lifecycle.py, tests/unit/test_huggingface_models.py
  • dependency dry-run for the WhisperX/pyannote path passed when excluding the existing optional DeepFilterNet local Rust build path

Notes

  • Full local requirements dry-run on macOS still stops at the existing DeepFilterLib source build because the local host has no Rust toolchain; this is unrelated to the WhisperX yanked fix.

Copilot AI review requested due to automatic review settings May 6, 2026 15:07
@github-actions
Copy link
Copy Markdown

github-actions Bot commented May 6, 2026

👍 @MapleEve

Thank you for raising your pull request and contributing to voscript.
Please make sure you have followed our contributing guidelines. We will review it as soon as possible.
If you encounter any problems, please feel free to connect with us.
非常感谢您提出拉取请求并为 voscript 做出贡献,请确保您已经遵循了我们的贡献指南,我们会尽快审查它。
如果您遇到任何问题,请随时与我们联系。

@codecov
Copy link
Copy Markdown

codecov Bot commented May 6, 2026

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 77.59%. Comparing base (70aa0f6) to head (e346b55).
✅ All tests successful. No failed tests found.

Additional details and impacted files
@@           Coverage Diff           @@
##             main      #27   +/-   ##
=======================================
  Coverage   77.59%   77.59%           
=======================================
  Files          75       75           
  Lines        2906     2906           
=======================================
  Hits         2255     2255           
  Misses        651      651           
Flag Coverage Δ
unit 77.59% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

@chatgpt-codex-connector
Copy link
Copy Markdown

💡 Codex Review

whisperx==3.3.1

P1 Badge Keep CTranslate2 compatible with the CUDA image

In the Docker image this service builds from pytorch/pytorch:2.6.0-cuda12.4-cudnn9-runtime, this pin pulls in WhisperX 3.3.1's ctranslate2<4.5.0 requirement. CTranslate2 4.5.0 is the release that switched the Python wheel to cuDNN 9 support, so the resolver now selects 4.4.x against a cuDNN 9 base image; any GPU ASR path that lazy-loads faster_whisper.WhisperModel on cuda will fail to initialize the CTranslate2 backend instead of using the container's CUDA stack.

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Copy link
Copy Markdown

Copilot AI left a comment

Choose a reason for hiding this comment

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

Pull request overview

This PR updates the dependency set used for WhisperX-based forced alignment to avoid installing yanked WhisperX 3.1.x releases, while keeping the project on the existing numpy<2 / SciPy 1.11.x baseline.

Changes:

  • Bump WhisperX from 3.1.6 (yanked) to whisperx==3.3.1.
  • Update/pin pyannote and adjacent dependencies (pyannote.audio==3.3.2, bounded pyannote.*, pandas<3, faster-whisper==1.1.0) to stay compatible with the current numeric stack.
  • Document the change in both English and Chinese changelogs under Unreleased.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated no comments.

File Description
doc/changelog.zh.md Adds an Unreleased entry documenting the WhisperX/pyannote dependency move and numeric stack constraints.
doc/changelog.en.md Adds the corresponding Unreleased entry in English describing the same dependency changes.
app/requirements.txt Moves off yanked WhisperX 3.1.x to whisperx==3.3.1 and pins/bounds related dependencies to maintain numpy<2 / SciPy 1.11.x compatibility.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

@MapleEve MapleEve merged commit f61061c into main May 6, 2026
13 checks passed
@MapleEve MapleEve deleted the fix/whisperx-yanked branch May 6, 2026 15:16
@github-actions
Copy link
Copy Markdown

github-actions Bot commented May 6, 2026

❤️ Great PR @MapleEve ❤️

The growth of project is inseparable from user feedback and contribution, thanks for your contribution!
项目的成长离不开用户反馈和贡献,感谢您的贡献!

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