Member-only story
Deploy your first Machine Learning Model
Deploying models is not as scary as you might think and it takes only few minutes to get up and running.
Before we deploy our model, we need to build it, train it, test it and finalize it. For this example, I am going to give an example of the simplest model so you won’t be lost in the model itself.
Our problem with machine learning:
- Our client app may not be in Python itself, it may be a Node.js, React, Angular, Ruby On Rails app.
- So how do we integrate two different applications?
We will build a machine learning model as a RESTful API service and deploy the API.
You can also see the image to get an idea. Building an API server for our machine learning model, we follow the principles of microservices. It may not seem reasonable for small apps, but for larger applications, following best practices will help to form maintainable codebase.
For this example, we will build a simple NLP model to extract named entities from a paragraph of text and give results as JSON output. We will use Python library called Spacy for this purpose. The figure may show Tornado…
