search

facebookresearch / audiocraft

Audiocraft is a library for audio processing and generation with deep learning. It features the state-of-the-art EnCodec audio compressor / tokenizer, along with MusicGen, a simple and controllable music generation LM with textual and melodic conditioning.
MIT License
20.15k stars 2.01k forks source link

How can I use a wav file that is different from the target music as my melody condition, when i fine-tune musicgen-melody? #434

Open Apple-jun opened 3 months ago

Apple-jun commented 3 months ago

Now I know that when I fine-tune musicgen-melody, the form of the dataset is the same as the form of the other text2music models, that is, the target audio and the condition audio to be converted to the chromatogram are the same. So I would like to ask how to make them different? I gave it a try but found the code hard to modify. I would appreciate any help.

UtsahaJoshi commented 2 weeks ago

@Apple-jun I'm currently working on a project where I need to use vocal audio as the input and generate instrumental music as the output using the musicgen-melody model. I understand that typically the target audio and condition audio are the same, but I need them to be different in my case. Any progress regarding this matter?