Car Price prediction Project

car-pred-demo.mp4

This repository contains the entire journey of building a car price prediction model, from data scraping to deployment using modern tools and practices. check out full explanation here: https://medium.com/@chidubemndukwe/ml-from-data-scraping-to-deployment-fa7ddc5fab5c

Project Overview

• Data Scraping

  • Scrapy: Developed a web scraper to extract car data, saved it as CSV.
  • PostgreSQL: Hosted the scraped data on a local PostgreSQL database.
  • GCP VM Instance: Deployed the scraper using ScrapeOps, scheduled to run daily at midnight.
  • Data Curation: Stopped the scraper upon reaching a sufficient number of entries.

• Machine Learning

  • Imported data from PostgreSQL to local storage.
  • Performed data analysis and preprocessing.
  • Built and evaluated the machine learning model.
  • Fine-tuned the model and saved it as a pickle file.

• FastAPI

Created an API for the model to process new data and make predictions. Hosted the API on port 8000.

• Gradio

  • Designed a simple UI for visualization purposes.

• Docker

  • Dockerized the API and Gradio app into a single container.
  • Pushed the container to Docker Hub.

• GitHub Actions

  • Configured CI/CD to build and push Docker images on PR to the main branch.

Repo Structure

car_pred_project
│   .dockerignore
│   .env
│   .gitignore
│   compose.yaml
│   Dockerfile
│   poetry.lock
│   pyproject.toml
│   README.md
│   requirements.txt
│   supervisord.conf
│
├───.github
│   └───workflows
│           docker-image.yml
│
├───car_pred
│   │   gradio_app.py
│   │   main.py
│   │   model.py
│   │   README.Docker.md
│   │   __init__.py
│   │
│   ├───data
│   │       car_data.csv
│   │
│   ├───model
│   │       car_prediction_model.pickle
│   │
│   ├───notebooks
│   │       import_data.ipynb
│   │       main.ipynb
│
└───tests
        __init__.py

Getting Started

Prerequisites

• Docker
• Python
• PostgreSQL
• Poetry

💻 Local Development

git clone https://github.com/Duks31/car_price-prediction 
cd car_price-prediction
poetry install

Fast API

uvicorn car_pred.main:app --reload

This will start the FastAPI server on http://localhost:8000/

Gradio

python car_pred/gradio_app.py

This will start the FastAPI server on http://localhost:7860/

🐋 Docker compose

Windows

git clone https://github.com/Duks31/car_price-prediction    
mkdir car_pred_project
cd car_pred_project
curl -o docker-compose.yml https://raw.githubusercontent.com/Duks31/car_price-prediction/main/compose.yaml
docker-compose up --build

Linux/macOS

git clone https://github.com/Duks31/car_price-prediction
cd car_price-prediction
wget -O docker-compose.yml https://raw.githubusercontent.com/Duks31/car_price-prediction/main/compose.yaml
docker-compose up --build

CI/CD with github actions

The GitHub Actions workflow is configured to build and push the Docker image to Docker Hub whenever a pull request is made to the main branch i.e. an update to the model.

The workflow file is located at .github/workflows/docker-image.yml.

Contributing

The projcet and model is definitely not perfect. Pull requests are welcome. For major changes.

License

This project is licensed under the MIT License. See the LICENSE file for more details.