jianfch/stable-ts
A Python library that modifies OpenAI's Whisper model to produce more reliable timestamps and provides transcription, forced alignment, and audio indexing utilities.
Not currently ranked — collecting fresh signals.
star history
This library extends OpenAI’s Whisper speech recognition model to generate more reliable and accurate word-level timestamps. It provides utilities for forced alignment of audio to text, audio indexing, silence suppression, and word regrouping. The library offers both standard and “whisperless” installation options, allowing users to integrate it with existing Whisper installations or use it as a standalone tool for timestamp refinement on audio transcriptions.
Frequently asked
- What is jianfch/stable-ts?
- A Python library that modifies OpenAI's Whisper model to produce more reliable timestamps and provides transcription, forced alignment, and audio indexing utilities.
- Is stable-ts open source?
- Yes — jianfch/stable-ts is open source, released under the MIT license.
- What language is stable-ts written in?
- jianfch/stable-ts is primarily written in Python.
- How popular is stable-ts?
- jianfch/stable-ts has 2.3k stars on GitHub.
- Where can I find stable-ts?
- jianfch/stable-ts is on GitHub at https://github.com/jianfch/stable-ts.