Training and deploying ML models in azure ml service

Today will be focusing about one of machine learning capability azure have provided in order to train your own models using most of the popular machine learning frameworks and deploying them inside docker containers in order to expose your models for serving.

What is Azure Machine Learning service?

Process inside Azure ml service.

With Azure machine Learning service you can develop train test deploy manage and track you’re machine learning model.It supports almost all open source machine learning frameworks and python packages which will be useful on the pipeline of machine learning model building.It Supports rich tools like Jupyter notebooks or the Azure Machine Learning for Visual Studio Code extension. Azure Machine Learning service also includes features that automate model generation and tuning to help you create models with ease, efficiency, and accuracy.

You can train you’re models locally or even on cloud with the compute instances and environments provided by azure.For example Azure Machine Learning Compute and Azure Databricks, and with advanced hyperparameter tuning services, you can build better models faster by using the power of the cloud.

How is Azure Machine Learning service different from Machine Learning Studio?

I wanted to talk a little about this since some people having a little misunderstanding about these two services.Since azure have released data processing capable notebooks in azure ml studio some people think it’s possible to train a model with you’re own code in azure ml studio.It’s not that way, Azure ML Studio is to train you’re model without writing any code.drag-and-drop visual work-space where you can build, test, and deploy machine learning solutions without needing to write code.So it’s easy but if you want you’re own model with own ml framework better to try Azure ml service.

In the next blog will be having a look at a real world scenario on training a model with azure ml service.

#apache #opensource #openmrs