torch
torchaudio
soundfile
transformers
bitsandbytes>0.37.0
vinorm
cached_path
huggingface_hub
gradio
accelerate>=0.33.0
click
datasets
ema_pytorch>=0.5.2
gradio>=3.45.2
hydra-core>=1.3.0
jieba
librosa
matplotlib
numpy<=1.26.4
pydub
pypinyin
safetensors
soundfile
tomli
torchdiffeq
tqdm>=4.65.0
transformers_stream_generator
vocos
wandb
x_transformers>=1.31.14