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
20.5k stars 2.06k forks source link

The import of audiogen at Colab #265

Open underdogliu opened 11 months ago

underdogliu commented 11 months ago

Hi and first of all thanks for such amazing work!

So I am now starting from the very beginning by trying the model via Colab. I found that when trying to import the audiogen model and run the related example script link, I have to install via the "bloody edge" method:

pip install -U git+https://git@github.com/facebookresearch/audiocraft#egg=audiocraft

instead of simply "pip install -U audiocraft". Is this something expected? Or I got it wrong somewhere in the middle?

Below is my example script which shall be run on Colab with T4.

%%capture
!pip install torch
!pip install -U git+https://git@github.com/facebookresearch/audiocraft#egg=audiocraft

import audiocraft
from audiocraft.models.audiogen import AudioGen
from audiocraft.data.audio import audio_write
adefossez commented 11 months ago

Yes at the time we had not release the stable version of audiocraft updated with all the new models. Now it should work though as we officially released the v1.0.0 on PyPI.