Free and Open-source Speech2Text SaaS – interviews and fine-tuning models (app.staging.voxtir.com)

A simple service intended for transcribing interviews and allowing for easy export to word. Turns out it works quite well for fine-tuning speech to text models. Hosted on AWS using founder credits. Uses whisper and Pyannote.

Code for fine-tuning can be found here https://github.com/Voxtir/whisper-fine-tuning-voxtir

and general source code found here https://github.com/Voxtir/voxtir/issues

Rolling it out to students and journalists. Feel free to reach out if you want to discuss it, contribute, or something else.