Containerizing Huggingface Transformers For Gpu Inference With Docker And Fastapi On Aws