Meta のオーディオ生成 AI モデル、AudioGen を macOS の MPS (GPU) で動かす

macOS: 14.5
ffmpeg version 7.0.1

brew install ffmpeg

mkdir AudioCraft_MPS
cd AudioCraft_MPS
git clone https://github.com/facebookresearch/audiocraft.git .

pipenv --python 3.11
pipenv shell

pip install torch==2.1.0

vi requirements.txt
#xformer<0.0.23
xformers==0.0.20

pip install -e .

audiocraft/models/encodec.py

    def decode(self, codes: torch.Tensor, scale: tp.Optional[torch.Tensor] = None):
        """Decode the given codes to a reconstructed representation, using the scale to perform
        audio denormalization if needed.

        Args:
            codes (torch.Tensor): Int tensor of shape [B, K, T]
            scale (torch.Tensor, optional): Float tensor containing the scale value.

        Returns:
            out (torch.Tensor): Float tensor of shape [B, C, T], the reconstructed audio.
        """
        emb = self.decode_latent(codes)
        #out = self.decoder(emb)
        # Below if block is added based on https://github.com/facebookresearch/audiocraft/issues/31
        if emb.device.type == 'mps':
            # XXX: Since mps-decoder does not work, cpu-decoder is used instead
            out = self.decoder.to('cpu')(emb.to('cpu')).to('mps')
        else:
            out = self.decoder(emb)

        out = self.postprocess(out, scale)
        # out contains extra padding added by the encoder and decoder
        return out

from audiocraft.models import AudioGen
from audiocraft.data.audio import audio_write
import argparse
import time

model = AudioGen.get_pretrained('facebook/audiogen-medium', device='mps') # mps で Apple Silicon の GPU を指定
model.set_generation_params(duration=5)  # [duration] 秒のファイルを生成

start = time.time()
def generate_audio(descriptions):
  wav = model.generate(descriptions)  # 引数として与えられたテキスト全ての音声を生成
  
  for idx, one_wav in enumerate(wav):
      # {idx}.wav というファイルを生成。音の大きさ loudness は -14 db LUFS で平準化
      audio_write(f'{idx}', one_wav.cpu(), model.sample_rate, strategy="loudness", loudness_compressor=True)
      print(f'{idx}.wav を生成')
      print(f'かかった時間: {round(time.time()-start, 2)} 秒')

if __name__ == "__main__":
    parser = argparse.ArgumentParser(description="Generate audio based on descriptions.")
    parser.add_argument("descriptions", nargs='+', help="List of descriptions for audio generation")
    args = parser.parse_args()
    
    generate_audio(args.descriptions)

python demos/audiogen_mps_app.py "text 1" "text 2"

python demos/audiogen_mps_app.py "heavy rain with a clap of thunder" "knocking on a wooden door" "people whispering in a cave" "racing cars passing by"
/Users/handsome/Documents/Python/AudioCraft_MPS/.venv/lib/python3.11/site-packages/torch/nn/utils/weight_norm.py:30: UserWarning: torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.
  warnings.warn("torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.")
0.wav を生成
かかった時間: 61.05 秒
1.wav を生成
かかった時間: 61.1 秒
2.wav を生成
かかった時間: 61.16 秒
3.wav を生成
かかった時間: 61.22 秒