curl

back to cpu

gpu fixed to cuda

using transformers pipeline and docker and uvicorn using gpu

using transformers pipeline

Create app.py