vocos mlx下载vocos mlx源代码下载

vocos mlx

Ai源码

0.0.7

下载

VOCOS - MLX

使用MLX框架实现VOCOS。 VOCOS允许从MEL频谱图或Encodec代币中高质量重建音频。

VOCOS：缩小时间域与基于傅立叶的神经声码编码器之间的差距，以进行高质量音频合成

纸[ABS] [PDF]

安装

要在推理模式下使用VOCOS，请使用：

pip install vocos-mlx

用法

MEL频谱图

 from vocos_mlx import Vocos , load_audio , log_mel_spectrogram

vocos = Vocos . from_pretrained ( "lucasnewman/vocos-mel-24khz" )

# reconstruct
audio = load_audio ( "audio.wav" , 24_000 )
reconstructed_audio = vocos ( audio )

# decode from mel spec
mel_spec = log_mel_spectrogram ( audio , n_mels = 100 )
decoded_audio = vocos . decode ( mel_spec )

Eccodec

 from vocos_mlx import Vocos , load_audio

vocos = Vocos . from_pretrained ( "lucasnewman/vocos-encodec-24khz" )

# reconstruct
audio = load_audio ( "audio.wav" , 24_000 )
reconstructed_audio = vocos ( audio , bandwidth_id = 3 )

# decode with encodec codes
codes = vocos . get_encodec_codes ( audio , bandwidth_id = 3 )
decoded_audio = vocos . decode_from_codes ( codes , bandwidth_id = 3 )

欣赏

AWNI HANNUN用于MLX的参考Encodec实现。

引用

 @article{siuzdak2023vocos,
  title={Vocos: Closing the gap between time-domain and Fourier-based neural vocoders for high-quality audio synthesis},
  author={Siuzdak, Hubert},
  journal={arXiv preprint arXiv:2306.00814},
  year={2023}
}