Commit History
-
Add four audio samples to the model card
Conversation, long-form narration, code-switched news, and a cross-lingual pair (Vietnamese reference -> English output), as 128 kbps mono mp3 under samples/ — 1.9 MB in total, and outside the download allow-list, so the Python loader never fetches them. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013TVfDo6Z3J1E43Jv9MLQDbff40271
