Repository navigation
Conversation
Locale tags such as en-US, zh-CN, or pt_BR previously raised Unsupported language even though the primary language is supported. Resolve them to the matching Whisper language code, including on English-only tokenizers that otherwise ignore the language argument.
Codex Review Summary: ✅ CompletedThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
get_tokenizer(..., language=...)accepted ISO 639 codes (en) and a few language names (english,mandarin), but rejected the locale tags users commonly pass from system locales and web APIs (en-US,zh-CN,pt_BR). That also broke English-only tokenizers, which ignore the language after validation and still raisedUnsupported language.This resolves a tag to its primary subtag when that subtag is a supported Whisper language (or a known alias), and leaves an unknown tag as an error.
Tests
Added regression tests in
tests/test_tokenizer.pyfor hyphen and underscore locale tags, language-name aliases, unknown tags, and the English-only tokenizer path.(
tests/test_timing.pyCUDA cases were skipped; this environment has no NVIDIA driver.)