sagemaker-demos

This repository is mainteaind by three authors:

• Muhyun Kim
• Yuval Fernbach
• Cyrus Vahid

The repository's aim is to implement most common use-cases of deep learning as SageMaker custom algorithms.

The code in this repository is not uptimized for performance and scale and is rather aiming to provide demos and tutorials.

#Introduction to AmazonSageMaker Python SDK SageMaker Python SDK is an open source library for training and deploying machine learning models on Amazon SageMaker.

Classes

• Estimator: Encapsulate training on SageMaker. Can be fit() to run training, then the resulting model deploy() ed to a SageMaker Endpoint.

  • EstimatorBase: Abstract class from which all estimators are derived
  • Framework: Superclass fro frameworks such as MXNet and TensorFlow.

• MXNet Estimator: With MXNet Estimators, you can train and host MXNet models on Amazon SageMaker.
  - attach: Create a SageMaker MXNetModel object that can be deployed to an Endpoint. - create_model: Create an Estimator bound to an existing training job. After attaching, if the training job is in a Complete status, it can be deployed to create a SageMaker Endpoint and return a Predictor. - train_image: EstimatorBase.fit method, which does the model training, calls this method to find the image to use for model training. The methods returns a docker image

• Model: Encapsulate built ML models. Can be deploy() ed to a SageMaker Endpoint.
  

  • deploy: Deploy this Model to an Endpoint and optionally return a Predictor
  • prepare_container_def: Return a dict created by sagemaker.container_def() for deploying this model to a specified instance type

• MXNet Model:
  

• Predictor: Provide real-time inference and transformation using Python data-types against a SageMaker Endpoint
  

• Session: Provides a collection of convience methods for working with SageMaker resources.