EveNet Documentation Portal¶

Welcome to the EveNet knowledge base! This site hosts the same Markdown guides that live in the repository, but it formats them with navigation, search, and table-of-contents support so new collaborators can onboard quickly.

Use the navigation menu to jump straight to the guide you need, or start with the highlights below.

🚀 Quick Starts¶

• Plug-and-play setup? Follow the Quick Start path in the Getting Started tutorial to pair the official Docker image with the PyPI package.
• Hacking on the source? The same guide outlines the advanced workflow for cloning the repo and running modules directly.
• Prepping datasets? Head to the Data Preparation guide to learn how to configure preprocessing YAMLs and generate parquet shards.
• Training & inference. Consult the Training playbook and Prediction walkthrough for command-line examples and Ray configuration tips.

🧠 Reference Library¶

• Architecture deep dive. The Model Architecture overview explains the hybrid point-cloud and global feature encoders that power EveNet.
• Configuration catalogue. The Configuration reference documents every YAML option and how they interact.
• Internal utilities. Project maintainers can review internal preprocessing notes for NERSC-specific helpers.

📚 Resources¶

Here are the main resources for EveNet, collected in one place for quick access:

• 🐳 Docker image: docker.io/avencast1994/evenet:1.5
• 🐍 Python package: PyPI evenet
• 🏋️ Pretrained weights: Hugging Face Model Hub
• 📦 Pretraining dataset: Hugging Face Datasets
• 💻 Git repository: UW-EPE-ML/EveNet_Public
• 📄 arXiv paper: coming soon