This repo covers the 3 scenarios for hosting ML models in production
- Use Sagemaker Pre-built Algorithms and Hosting Container Example for using Sagemaker pre-built algorithms and images for online inference.
- Use Own Code and Sagemaker Frameworks Examples for using pre-built Sagemaker Frameworks for MXNet and Tensorflow with your own code.
- Use Own Code and Own Hosting Container Examples for using your own code and own hosting container when you either need to add new libraries to existing Sagemaker images or a new framework is to be used.