Add a Traditional Chinese reading guide - #40
Conversation
|
Thank you for supporting Traditional Chinese. My suggestion is to avoid adding Chinese code and instead announce the change in English. |
|
Thank you for the guidance. I identified the scope issue and have now corrected it. The latest update removes the Chinese changes from source-code comments and restores the code exactly to I also shortened I also recognize that I published the earlier iteration too quickly. Going forward, I will complete and review the full change locally before pushing an update or replying. Thank you again for your patience and guidance. |
|
Thanks for addressing the scope issue and updating the PR. I reviewed the latest changes, and everything looks good now. The PR is appropriately limited to the two Markdown files, the source code matches The English announcement and the positioning of the Traditional Chinese guide as optional, with the English README remaining canonical, are clear. LGTM. Approved from my side. Thanks for the careful follow-up! |
Motivation
To make Apollo a little easier to approach for readers in Taiwan and other Traditional Chinese readers, and hopefully help more people discover this open-source project, this PR adds a small optional reading guide. The English README remains the canonical source for complete and current technical information.
I appreciate the maintainers' guidance on keeping the change focused. If this additional guide does not fit the project's current documentation plans, I would be happy to revise or withdraw it.
Summary
README.zh-TW.mdas a concise Traditional Chinese reading guide rather than a full localization.Traditional Chinese guidenavigation link toREADME.md.upstream/main.Validation
python -m py_compile inference.py tests/test_inference.py look2hear/utils/parser_utils.py look2hear/utils/pylogger.py train.pypython -m unittest discover -s tests -v: 14/14 passedpython -m pip check: no broken requirementsgit diff --checkupstream/main: two Markdown files onlyThis PR does not modify inference or training behavior, checkpoints, audio fixtures, generated artifacts, or any private audio.