Magma

A collection of automation workflows designed to simplify machine learning tasks in HPC.

Table of Contents

• Prerequisites
• Resources Deployment
  • Helm Chart for GitHub Runners
  • MLflow Deployment
• Usage
• Contributing
• License

Prerequisites

Before running any of the GitHub Actions workflows in this repository, the following resources need to be deployed and configured:

1. Self-hosted GitHub Runners: Deployed using a Helm chart for scalability and management.
2. MLflow Server: Set up for real-time tracking and monitoring of machine learning experiments.

Ensure that you have access to a Kubernetes cluster and Helm installed in your local environment. The cluster should have sufficient resources for deploying the GitHub runners and the MLflow server.

How to Execute the Workflow:

To execute the workflow, visit the following link: langtech-bsc/launch2hpc