Hey, does this allow inferencing on CPU after the model deployment?

I am working on LVO detecton task. After  create a custom model, can i convert it into torchscript and depoy it using docker and use it for inference on device with no GPU?