Method to save my Mother in Law time she takes to complete dictations.
Uses OpenAI Whisper to trancribe the audio to text. Pipe that through llama3 to refactor output and catch mistakes made by whisper.
Formats the net output into a docx file.
TODO
- Add a cli: I want to be able to pass a remote directory and process all files in that directory.
- Determine the best ollama model for my purposed. What runs well on my pc and what takes the context the best.
- Write some fancy tests