Releases: JuergenFleiss/atrain_core
v1.2.0
We are pleased to announce the release of aTrain_core v1.2.0
This release offers many performance improvements and new features. The highlights include:
Upgraded models for faster and more accurate transcriptions:
Whisper's Turbo Model: We've integrated the faster-large-v3-turbo model, providing near real-time transcription capabilities
Upcoming:
aTrain_core powers the upcoming release of our GUI application aTrain which will allow for intuitive transcription whilst leveraging the performance of aTrain_core. It is set to be released soon on Windows, MacOS, and Debian. Keep in touch!
v1.1.0
First release of the CLI version
aTrain-core: A command line interface for the transcription of audio and video files (faster-whisper), including speaker diarization (pyannote-audio). It runs on Windows, MacOS and Debian-based Linux distributions. Supports all the newest models, including Whisper large V3 and distilled large v3 model (English, real time transcription on CPU).
