S1E1: Code & Deploy: Build and Deploy an ML Binary Classifier

Join me with my guest William Arias, developer advocate at GitLab, as they show you step-by-step how to code and deploy a binary classifier using machine learning with open source tools. You’ll first prototype a solution in a Jupyter Notebook, then set up a basic CI pipeline to automate a machine learning app experimentation with MLFlow. Next, you’ll refactor the application (a bit) and create an API endpoint and simple UI for it to test manually (vibe testing). Before you’re done, you’ll learn how to combine everything and deploy the application to a runtime in the cloud using CI/CD principles. 

In this session, you’ll learn how to:

• Prototype a binary classifier using machine learning in a Jupyter Notebook.

• Set up a basic continuous integration (CI) pipeline with MLFlow for app experimentation.

• Refactor the application to create an API endpoint and a simple user interface.

• Deploy the application to a cloud runtime using continuous deployment (CD) principles.

Links to repo
https://gitlab.com/gitlab-da/use-cases/devsecops-platform/deep-learning/meowsky-classifier/-/blob/main/classifier/modeling/train.py

https://gitlab.com/gitlab-da/use-cases/devsecops-platform/deep-learning/meowsky-classifier/-/blob/main/classifier/modeling/train.py