my-whisper-asr / requirements.txt
v-e-n-o-m's picture
use torchaudio. not liborsa
855f267
raw
history blame contribute delete
190 Bytes
transformers==4.38.2
torch==2.0.1
fastapi==0.103.0
uvicorn==0.23.2
pydantic==2.3.0
soundfile==0.12.1
python-multipart==0.0.9
numpy==1.26.4
timeout-decorator==0.5.0
torchaudio==2.0.2