asr-demo / requirements.txt
GavinHuang's picture
fix: improve audio processing in transcribe function and add soundfile dependency
779d79b
raw
history blame
112 Bytes
torch>=1.13.0
gradio>=4.0.0
nemo_toolkit[asr]>=1.18.0
omegaconf>=2.2.0
numpy>=1.22.0
cuda-python>=12.3
soundfile