Alok Saboo

@vabene1111 would love for you to test this approach out. I think recipes will be more reliable with such structured outputs in the long run.

Honestly, we can skip embeddings for now. Adding a vector database is a significant undertaking, and we should definitely consider it to AI-enable beets (although I am not the right...

Yes, I already have a working implementation at https://github.com/arsaboo/beets/tree/audiomuse (happy to create a draft PR, if that helps). Basically, we can get the following values: `audiomuse_embedding` can be disabled if...

Users may use their existing Ollama instance. It may be better to just add a env variable specifying the ollama endpoint.

I'm guessing there are no tags in the file.

@sgued you should use the [FromFilename](https://beets.readthedocs.io/en/stable/plugins/fromfilename.html) plugin for that. @semohr Agreed. If no metadata is provided, we can simply skip (possibly with a log).

@endcredits33 indeed, because `media` for an album does not always make sense. What would you expect if one of the tracks is vinyl and the other track is DVD-ripped?

I don't see the value of doing this. Have you tried [album_fields](https://beets.readthedocs.io/en/stable/plugins/inline.html#album-fields)?

From what I understand, reimport `-L` behaves the same as `--search-id` when the pre-existing ID is present in the DB. In your case, album_is 853460 was discovered and used. I...

Shouldn't it work even with the hash part? The identifiers across different tag sources are not similar. The probability that the `853460` album_id is used for the same album across...