🧩 RoboML Research Template

A lightweight, opinionated template for running machine learning and robotics research projects. It provides a standardized structure for experiments, data handling, logging, and evaluation, with ready-to-use configs and scripts for reproducible workflows.

Projects Built with RoboML Research Template

Projects that used RoboML template:

Let me know if you use this template!!

Features

• Hydra Configuration Management: Modular config structure with separate files for model, training, and wandb settings
• Type-Safe Configs: Dataclass schemas for configuration validation and IDE support
• Wandb Integration: Built-in experiment tracking and logging
• Fast Package Management: Uses uv for quick dependency installation
• Development Tools: Pre-commit hooks with ruff and black for code quality
• Auto-formatting: Ruff automatically formats Python files on save (VS Code/Cursor)

Setup

Prerequisites

• Python 3.8+
• uv package manager

Installation

1. Clone the repository:

git clone <repository-url>
cd roboml-research-template

2. Install dependencies:

uv venv
source .venv/bin/activate
make install

Note: This template installs dependencies directly rather than using pip install -e ".[dev]" because it's structured as a template repository, not an installable Python package.

3. (Optional) Install pre-commit hooks:

make pre-commit-install

Try it! It's simple

python main.py

Configuration

Configuration files are located in the config/ directory:

• defaults.yaml: Main defaults file that imports all configs
• model.yaml: Model architecture parameters
• training.yaml: Training hyperparameters
• wandb.yaml: Wandb project settings
• config.yaml: General experiment settings

Override config values via command line:

python main.py training.learning_rate=0.01 training.batch_size=64

Development

• Lint code: make lint
• Format code: make format
• Run tests: make test
• Clean cache: make clean

Project Structure

roboml-research-template/
├── config/              # Hydra configuration files
│   ├── config_schema.py # Dataclass schemas
│   ├── defaults.yaml    # Main defaults
│   ├── model.yaml       # Model config
│   ├── training.yaml    # Training config
│   ├── wandb.yaml       # Wandb config
│   └── config.yaml      # General config
├── networks/            # Network architectures
├── learner/             # Learning algorithms
├── tests/               # Test files
└── main.py              # Entry point

Contributing

Contributions are welcome! Please see CONTRIBUTING.md for guidelines.

License

See LICENSE file for details.