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.
Are there any scripts in this repo that I'm missing for generating those json files from the single MusicCaps CSV metafile ? Otherwise anyone can point me to a download location where this has already processed according to that structure?
Any advice would be appreciated. I want to avoid extra manual efforts that may prevent me from reproducing the reported numbers.
Hello,
I'm trying to reproduce the MusicGen evaluation metrics on MusicCaps dataset. I've downloaded the dataset and have created the manifest file according to the instructions in the repo https://github.com/facebookresearch/audiocraft/blob/main/docs/DATASETS.md.
There is a note that each wav file needs to also have an accompanying json file, with metadata, similar to the structure in https://github.com/facebookresearch/audiocraft/tree/main/dataset/example, however the dataset does not provide it in this way.
Are there any scripts in this repo that I'm missing for generating those json files from the single MusicCaps CSV metafile ? Otherwise anyone can point me to a download location where this has already processed according to that structure?
Any advice would be appreciated. I want to avoid extra manual efforts that may prevent me from reproducing the reported numbers.