mkiol
Remote server component
App is great. I wanted to toss out the idea of a remote server component to offload the models or compute to. That way on some laptops or lower spec machines, if someone has a better rig or self hosted server they could use for some of the processing. Maybe even add a ping, port, latency, check to build upon for auto switching between a remote server and local. The work on this app is incredible and i appreciate it. I wanted to mention this as a possible idea, but I understand it might be outside the scope.
The idea is definitely interesting, but it is out of scope at the moment. Speech Note is a "fat" desktop GUI app.
Here is the feature request for an HTTP-based API for "local" integration with, for example, a Firefox extension. I want to implement this API, but the scope will be limited now. Who knows, maybe it will evolve into a full-fledged server eventually ;-)
Would love to +1 this enhancement and allow for integration to remote hosted endpoint e.g. https://api.openai.com/v1/audio/transcriptions
Whilst it may be counter to what this repo is about, it would be super useful to many to be able to run this on a laptop without a beefy GPU available.
