This package is code to deploy a Sagemaker endpoint. The trained models are based off of previously run Sagemaker training jobs. Code for running Sagemaker training jobs can be found here.
Sagemaker endpoints utilize docker containers stored in AWS Elastic Container Registry to execute. Code for deploying the container based inference jobs can be found in the run_container.ipynb
Steps for running a inference job are as follows:
- Build container and upload to AWS ECR using the build_and_push.sh script
!cd container; ./build_and_push.sh gnn-inference
- Refer to run_container.ipynb on deploying Sagemaker endpoint and running inference