New Features
- Transcript Exporter: Can be used to save predicted transcript to
vtt,srt,json,tsv,txt. (Doc: https://github.com/shashikg/WhisperS2T/blob/main/docs.md#write-transcripts-to-a-file) - Prebuilt docker images: Released ready to use prebuilt docker images. (Doc: https://github.com/shashikg/WhisperS2T?tab=readme-ov-file#from-docker-container)
- Option to use single
lang_codeortasksinstead of list -- when all the audio files belongs to same language/task. #27
Bug Fixing
- Fix without VAD
transcribefunction by @shashikg in #15 (Doc: https://github.com/shashikg/WhisperS2T/blob/main/docs.md#run-without-vad-model) - Fix issue with silent file by @shashikg in #12
- Fixed missing dependency and tensorrt-llm failures by @shashikg in #32
Full Changelog: v1.3.0...v1.3.1
Contributors
shashikg