csm-1b-tts-demo / requirements.txt
Константин
Refactor FastAPI application to enhance audio synthesis functionality and improve logging. Introduced model loading with error handling, updated endpoint to '/synthesize', and implemented thread-safe audio generation. Modified Dockerfile to streamline dependency installation and updated requirements.txt to include additional packages.
5f36d04
raw
history blame contribute delete
166 Bytes
fastapi
uvicorn[standard]
torch
torchaudio
huggingface_hub
transformers
tokenizers
numpy
triton
silentcipher @ git+https://github.com/SesameAILabs/silentcipher@master