A one-click to transcribe, summarize, and create outline of large speech audio using Insanely Fast Whisper.
Harewoosh is a limitless transcriber web UI. It designed to transcribe huge speech/audio file without worries. Uploading, transcribing, and summarizing large video/audio file are no longer a hassle. It's not like public endpoint which mostly limited to 25 MB of file to transcribe. HareWoosh is simply no limit.
HareWoosh designed to quickly capture ideas of a speech, it will create summary and outline of the transcribed file. Leveraging Insanely Fast Whisper with whisper-large-v3-turbo, the process is very fast.
- Install Insanely Fast Whisper as described on it's repo.
- Clone this repo:
git clone https://github.com/Koesn/harewoosh. - Run HareWoosh:
uvicorn main:app --host 0.0.0.0 --port 8001 --reload.
I haven't yet documented it's requirements, so for now you can add it manually using pip install <module>.