Add GPU support for inference service

Change Literal to str, because PyArrow doesn't know about literals

Clean up docs. Fix test script incorrect path.

Default to CPU-only version of torch. Improve docs.

Fix docker image not working in kubernetes. Fix int width not set in input scherma. Add makefile and test image.

Finalize Request and Response schema and update model accordingly.

fix for async methods as per PR

fix circular deps issue

fix circular deps issue

removed verbosity

refactored codebase

commit README.md

initial commit