There are 0 repository under model-deploy topic.
Version, share, deploy, and monitor models.
This is the Capstone project (last of the three projects) required for fulfillment of the Nanodegree Machine Learning Engineer with Microsoft Azure from Udacity. In this project, we use a dataset external to Azure ML ecosystem. Azure Machine Learning Service and Jupyter Notebook is used to train models using both Hyperdrive and Auto ML and then the best of these models is deployed as an HTTP REST endpoint. The model endpoint is also tested to verify if it is working as intented by sending an HTTP POST request. Azure ML Studio graphical interface is not used in the entire exercise to encourage use of code which is better suited for automation and gives a data scientist more control over their experiment.
In this project, we use a dataset external to Azure ML ecosystem to train and deploy models using AutoML and HyperDrive services.
This is second of the three projects required for fulfillment of the Nanodegree Machine Learning Engineer with Microsoft Azure from Udacity. In this project, we create, publish and consume a Pipeline. We also explore ML model deployment as an HTTP REST API endpoint, swagger API documentation, apache benchmarking of the deployed endpoint and consumption of the endpoint using JSON documents as HTTP POST request.