Lyrics missing

[Verryx-02]

Problem:

It's not possible to get lyrics for all songs.
Some are missing or unavailable.

[Verryx-02]

Possible Solution:

Use a model (e.g., OpenAI Whisper ) that can extract text from the audio track.
Whisper can run locally, is open source, and is distributed under the MIT license.

[Verryx-02]

Problem:

We need to figure out if there's a reliable enough model that can run on all PCs.
Alternatively, this feature could be offered only to those with certain types of hardware.
One thing is certain: no calls will be made to the OpenAI API. The audio tracks must be processed locally.