Docker Images for Machine Learning and MLOps

This repository provides ready-to-run Docker images and Docker Compose configurations for applications essential in Machine Learning and MLOps workflows.

Repository Structure

Jupyterlab

Contains various JupyterLab configurations in Docker for both CPU and Nvidia GPU machines.

• jupyterlab/jupyterlab-cpu - Run Jupyterlab with popular scientific packages. Based on jupyter/scipy-notebook Docker image.
• jupyterlab/jupyterlab-cpu-multienv - Run JupyterLab with multiple conda environments, each preinstalled with packages for Machine Learning.
• jupyterlab/jupyterlab-gpu - Run Jupyterlab with Nvidia GPU drivers. Based in nvcr.io/nvidia/pytorch Docker image.

MLOps Infrastructure

Contains various Docker configurations for running DevOps and Monitoring applications useful for Machine Learning projects.

• mlops-infra/gitlab - Set up a self-hosted GitLab instance for version control, CI/CD, and project management in ML workflows.
• mlops-infra/prometheus-grafana - Deploy Prometheus for metrics collection and Grafana for visualization, enabling comprehensive monitoring of ML infrastructure and applications.
• mlops-infra/pypi-server - Run a private PyPI server to host and manage custom Python packages for your ML projects.
• mlops-infra/nginx-certbot - Configure Nginx as a reverse proxy with automatic SSL certificate management using Certbot, ensuring secure access to your ML services.

Usage

Each subdirectory contains its own README with specific instructions for building and running the Docker images or Docker Compose configurations.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.