esa-pynas 0.1.0

PyNAS, a Python package for Neural Architecture Search experiments

PyNAS

PyNAS is a Python framework for Neural Architecture Search (NAS) experiments focused on resource-constrained, edge-deployable deep learning models. The project is developed by ESA Phi-lab with Little Place Lab and targets workflows where model accuracy, memory footprint, and deployment cost must be considered together.

Why PyNAS

PyNAS provides building blocks for evolutionary NAS experiments, with an emphasis on compact neural networks for onboard and edge AI use cases.

Area	What is included
Architecture generation	Utilities for generating, parsing, and rebuilding architecture codes
Search operators	Genetic mutation and single-point crossover primitives
Model blocks	Convolutional, pooling, activation, residual, classifier, and U-Net components
Training utilities	PyTorch Lightning modules, segmentation losses, metrics, and early stopping helpers
Data utilities	Hugging Face dataset download helper with retry handling

Project Status

Item	Status
Package name	esa-pynas
Version	0.1.0
Python support	>=3.11,<3.15
Distribution	Local checkout; PyPI publication planned
License	Apache 2.0
Development status	Alpha / research preview

Installation

PyNAS uses uv for reproducible dependency management. From a local checkout:

uv sync --locked

For development and verification:

uv sync --locked --dev
uv run pytest
uv run ruff format --check src scripts data tests
uv run ruff check src scripts data tests

The default lock configuration uses CPU PyTorch wheels for deterministic Windows, Linux, and macOS CI environments. For CUDA-enabled training workloads, install the PyTorch wheel index that matches your driver and hardware before running experiments.

Quick Start

Generate and inspect a candidate architecture code:

from pynas.core.architecture_builder import (
    generate_code_from_parsed_architecture,
    generate_random_architecture_code,
    parse_architecture_code,
)

architecture_code = generate_random_architecture_code()
parsed_architecture = parse_architecture_code(architecture_code)
round_tripped_code = generate_code_from_parsed_architecture(parsed_architecture)

print(architecture_code)
print(round_tripped_code)

Use the public package interface for training-related utilities:

from pynas import (
    CategoricalCrossEntropyLoss,
    GenericLightningSegmentationNetwork,
    Individual,
    calculate_iou,
)

Data

The burned-area segmentation dataset referenced by this project is hosted on Hugging Face:

• ESA-PhiLab-Edge/LPL-Burned-Area-Seg

Install the optional data dependencies and run the downloader:

uv sync --locked --extra data
uv run python data/download_hf_datasets.py

The downloader supports dataset and model repositories from the Hugging Face Hub, retry handling for transient network failures, progress reporting, and custom local output directories. Edit the repo_ids list in data/download_hf_datasets.py to change the default download targets.

Repository Layout

.
├── data/                 # Dataset download utilities and data notes
├── docs/                 # Documentation assets and static documentation files
├── examples/             # Usage examples and demos
├── notebooks/            # Experiment and walkthrough notebooks
├── papers/               # Research manuscript assets
├── scripts/              # Training and data-loading scripts
├── src/pynas/            # PyNAS Python package
│   ├── blocks/           # Neural network building blocks
│   ├── core/             # Architecture, population, configuration, and model logic
│   ├── opt/              # Evolutionary optimization operators
│   └── train/            # Losses, metrics, and training helpers
└── tests/                # Unit and integration tests

Research Context

Spaceborne edge computing enables AI-capable CubeSats to process data onboard, reduce downlink pressure, and operate with greater autonomy. These systems face strict memory, power, and latency constraints, so model design must account for both predictive performance and deployment cost.

PyNAS explores evolutionary Neural Architecture Search for this setting. The framework is designed around compact segmentation architectures and hardware-aware optimization for CubeSat-class platforms such as NVIDIA Jetson AGX Orin and Intel Movidius Myriad X targets.

Documentation

• Project documentation: sirbastiano.github.io/pynas-docs
• Source repository: github.com/ESA-PhiLab/pynas
• ESA Phi-lab: philab.esa.int
• Little Place Lab: littleplace.com

Citation

If you use PyNAS in academic work, cite the Zenodo record below. A paper-specific citation will be added when the associated manuscript is available.

@software{pynas_zenodo_15332053,
  title = {PyNAS},
  publisher = {Zenodo},
  doi = {10.5281/zenodo.15332053},
  url = {https://doi.org/10.5281/zenodo.15332053}
}

Authors

• Roberto Del Prete (Google Scholar)
• Parampuneet Thind (Google Scholar)
• Andrea Mazzeo
• Lorenzo Papa (Google Scholar)
• Matthew Whitley
• Gabriele Meoni (Google Scholar)
• Nicolas Longepe (Google Scholar)

License

This project is licensed under the Apache License 2.0. See LICENSE for the full license text.