Deploying pre-trained AI model on Azure
I find that deploying any ML model on the cloud is not very straightforward. Often the given documentation is not sufficient and beginners always find themselves stuck with the given documentation. In this blog, we are going to deploy a simple ML model to the Azure cloud as a web application.
In this blog we are not going to train the ML model on Azure rather we are going to deploy a model that is trained on our local machine. For our use case, we are going to deploy a simple model on the iris dataset. The goal of the blog is to teach deployment, not model building.
Contents
- Train a simple SVM on the Iris dataset.
- Write the Dependency files that are required for the deployment.
- Deploy model on Azure.
- Test the created Endpoint.
Model building
For our case, we are going to use SVM to classify the iris dataset.
Iris data can be downloaded from here: https://archive.ics.uci.edu/ml/machine-learning-databases/iris/iris.data
Let’s build the simple SVM model first. Put the iris.data file in the same folder as the below script. Let’s call the below file iris_SVM.py and after running this file one will get a pickle file called iris.pkl that…