Deploying an ML Model with FastAPI

Repository for the 3rd project in the Udacity Machine Learning DevOps Engineer Nanodegree

Project Description

The project had the following requirements for completion:

• Setting up a GitHub Action for continuous integration
• Training a Machine Learning model to predict an individual's salary using census data
• Evaluating model performance on data slices
• Creating a REST API using the FastAPI library that implements a POST method that does model inference
• Writing unit tests for src functions and test cases for API
• Deploying API on Render.com

File Descriptions

• src/ - package containing functions written for model training and evaluation
• train_model.py - script to train LGBM model on census data
• model_card.md - model card of trained model
• training_output/ - contains model, encoder, and binarizer objects returned by train_model.py
• test_src.py - unit tests for src code
• get_slice_metrics.py - script to get model metrics on data slices for categorical features
• slice_output/ - contains output of get_slice_metrics.py
• app/
  • main.py - code for creating API
  • test_main.py - test cases for API
  • live_post.py - live example for API post method
• screenshots/ - contains required screenshots

Link to Render Dashboard

Link to GitHub Repository