I think something like this would be great as I am just getting started using super whisper . Something like the gpt library or prompt library of OpenWebUI would be amazing.
Neil Chudleigh - so I don't know what the larger user base will want but here are some examples of the ones I have set up: - JIRA Requirements: This generates markdown in sections around a user story, acceptance criteria, testing plan. Outputs in markdown format. - Internal Email: Uses an informal tone, corrects grammar/punctuation, hardcodes my email signature/name. - 1:1 Meeting: For my employees' weekly 1:1. Discards any non-work discussion, generates markdown for "Last Week", "This Week" and incorporates my clipboard contents (I copy last week's notes) to detect which items carried over from last week and marks them accordingly. - Multi-person Meeting: Outputs "key points", "takeaways", and "next steps". Listens for takeaways and assigns the person's name "if known" in a specified "- @[Name]" format - Chat: Used for Slack, primarily. Cleans up grammar/punctuation, removes crutch words (it may already do this in the voice model), formats specifically for conversational chat in mind.
Dave Weniger What kind of modes would you be looking for? What common usecases are you thinking?
I would also like the ability to use whatever model I'd like. I'm interested in testing out the Janus models from DeepSeek. And I think it's simpler/faster to have the ability to use a local model of your choice rather than wait for the devs to manually support it. That said, there may be technical reasons for now allowing custom model use at this time that I'm unaware of. ohhhh wow you said MODES not MODELS 🤦♂️️ Welp, disregard my comment. I may need to make a new request after all.