Welcome to galaxyHackers documentation!

GalaxyHackers is an innovative toolkit for astronomical data analysis, with a focus on galaxy cluster detection. The project provides utilities for building, training, and evaluating machine learning models specifically tailored to the unique challenges of astrophysical datasets.

Key Features

• Fetches and processes astronomical data.

• Annotates and visualizes galaxy images dynamically.

• Trains and evaluates YOLO models for galaxy and cluster detection.

Features

• Supports a variety of deep learning models: ResNet18, EfficientNet, DenseNet, ViTL16, etc.

• Integrates with Comet ML for tracking training experiments.

• Includes automatic segmentation plot generation after model training.

• Provides flexibility in choosing optimizers, learning rate schedulers, and hyperparameters.

galaxyHackers Overview

• Overview
  • Introduction
  • Architecture Overview
  • Next Steps

Get Started

• Installation
  • Cloning the Repository
  • Setting Up the Environment
• Usage
  • Preparing the Dataset
  • Training Models
  • Output and Logs
• Development
  • Contribution Guidelines
  • Code Style
  • Workflow for Adding Features
  • Best Practices

Extra

• Changelog
  • v0.1 (May 7, 2024)
• API Reference
  • galaxy

GitHub Repository: https://github.com/pelancha/galaxyHackers