LLM Post Processing

[elebumm]

I've started working on a feature that I think would be a life saver for me.

DemoWhisper.mp4

A post processing method that will take your transcription and run it through an OpenAI compatible API of your choosing. In my demo:

• Understands when to do a new line
• Understands when I made a mistake and to remove it from the final output
• Converts "smiley face" into an emoji.
  (Used Gemini Flash with Open Router)

I wanted to take some feedback before I develop this into a PR. Currently I have it so that the user customizes the system prompt as well as the beginning of the user prompt that the transcription will be inserted in.

For my use case above, I also needed to change up pynput to support "new line" or emojis.

This is for a video for my channel, so interested to hear everyones thoughts!