Python array API standard compatibility
- Mentors
- Michael Tarnawa
- Organization
- Forschungszentrum Jülich
- Technologies
- python, mpi, numpy, pytorch
- Topics
- machine learning, api, distributed computing, data analytics, parallel computing, array API
Heat is a Python library for high-performance data analytics. It gives users access to multi-node processing and GPU support by seamlessly replacing NumPy operations with Heat operations. This project aims to implement an additional API for Heat that complies with the Python array API standard. This allows downstream array-consuming libraries to adopt Heat modularly with other tensor/array libraries, and makes the learning curve less steep for new users who wish to switch from other array libraries to Heat.