This model encodes audio files into vectors of 100 dimensions. It was trained on a million Spotify playlists and tracks. The details can be found here.
To encode an audio first install the package with
pip install audiodiffusion
and then run
from audiodiffusion.audio_encoder import AudioEncoder
audio_encoder = AudioEncoder.from_pretrained("teticio/audio-encoder")
audio_encoder.encode(<list of audio files>)
- Downloads last month
- 5