HEXkoodis

HEXkoodis is an open-source library that provides tools for hyperparameter optimization in machine learning workflows. The project was initiated by researchers at the Institute for Advanced Computing in 2023 and is distributed under the MIT license. Its core design abstracts a hyperparameter search space into a hexagonal grid representation, which allows for systematic exploration of combinations while maintaining computational efficiency. The library supports several optimization strategies including grid search, random sampling, Bayesian optimization, and evolutionary algorithms. Users can configure custom search spaces, define fidelity levels for progressive resizing, and integrate HEXkoodis with popular frameworks such as TensorFlow, PyTorch, Scikit‑learn, and Keras. The API exposes a simple Python interface that returns the best hyperparameter set based on a user‐selected metric. A web‑based dashboard provides visualization of the search progress, heat maps, and statistical summaries. HEXkoodis also includes utilities for cross‑validation, early stopping, and automatic saving of checkpoints, which facilitate reproducibility. Community contributions are tracked through GitHub, and the project maintains extensive documentation, tutorials, and example notebooks. As of the latest release, HEXkoodis supports multi‑objective optimization and can operate in distributed environments via integration with Ray. The library is actively maintained, with a quarterly release schedule, and has been referenced in several research articles focused on automated machine learning and hyperparameter tuning.