aparatext
aparatext
> Something like [whispering](https://github.com/shirayu/whispering) perhaps Do we need real-time? All we have to do is listen to a wake word, record the audio, stop when there's no more voice activity,...
Thanks to the InnerTube project and their wrappers, it's now possible! LuanRT/YouTube.js and tombulled/innertube might give you some details about implementing the InnerTube port for NewPipe.
> I'm curious - why is there no bridge for Viber, is there just a lack of interest/resources, or is there something more insidious? They are intent on wasting any...
> One strong engineering constraint I see is that all of this should be in native code, better in c++, to be merged within the libkiwix. While most of the...
I would like to outline some considerations on implementing this feature. First-of-all, how are we going to store the embeddings? One vector costs 6145 bytes and there are 6,821,862 articles...
Hmm.... Didn't Cohere make their models free to call (albeit with a rate limit)? This could make using this much much more viable for scraping a few small pages. EDIT:...
> I'd love to see support for OpenAssistent models. Perhaps someone could train some seq2seq model precisely for this task...
Thanks for the response. I'll look into it. I just wanted to create an issue for tracking the progress.
I think fetching any data should return a Result so we could propagate any parsing errors. As for parsing the InnerTube responses, maybe iterators could parse everything, returning an Option...
Is there any kind of InnerTube schema that can be used to validate the JSON?