Project Overview

In this project, you will apply the skills you have acquired in this course to operationalize a Machine Learning Microservice API.

You are given a pre-trained, sklearn model that has been trained to predict housing prices in Boston according to several features, such as average rooms in a home and data about highway access, teacher-to-pupil ratios, and so on. You can read more about the data, which was initially taken from Kaggle, on the data source site. This project tests your ability to operationalize a Python flask app—in a provided file, app.py—that serves out predictions (inference) about housing prices through API calls. This project could be extended to any pre-trained machine learning model, such as those for image recognition and data labeling.

The final implementation of the project will showcase your abilities to operationalize production microservices.

---

Setup the Environment

• Create a virtualenv and activate it

python3 -m venv ~/.devops
source ~/.devops/bin/activate

• Run make install to install the necessary dependencies

The files in this repository

• app.py
  • This is the core of the project, it consists in a microservice written in Flask (Python) that uses a pre-trained machine learning model to predict housing prices in Boston, there are only two endpoints available on this API.
• requirements.txt
  • This is where we list all the dependencies to run the project
• Makefile
  • This includes instructions on the setup, tests and linting of the project
• Dockerfile
  • Here we have all the instructions to containerize the Flask service
• Bash scripts. As a good practice, we are including all the required steps to execute as bash scripts
  • run_docker.sh (builds the image, and run the container)
  • run_kubernetes.sh (runs a docker image with kubernetes, lists the kubernetes pod(s), and forwards the container port to a host)
  • upload_docker.sh (uploads the previously build image to DockerHub)

Running app.py

1. Standalone: python app.py
2. Run in Docker: ./run_docker.sh
3. Run in Kubernetes: ./run_kubernetes.sh

Testing the microservice

Once you have the service up and running, you can run the bash script ./make_prediction.sh and it will make a POST request to the /predict endpoint with a proper payload.

The response should be a value (float) with a 200 OK HTTP Status

Uploading the docker image to DockerHub

If you haven't login yet, it will ask you so do: docker login and use your credentials from DockerHub

Once you do that, you will be able to upload the image by running the script:

./upload_docker.sh