Shypd.aiThe largest open index of AI toolsMatGL is an open-source graph deep learning library for materials science, enabling the development and sharing of materials graph deep learning models. It provides architectures like M3GNet and TensorNet for predicting material properties.
Trending
About
MatGL (Materials Graph Library) is an open-source graph deep learning library specifically designed for materials science. It leverages mathematical graphs to represent atomic collections, allowing for the development of surrogate models that consistently deliver exceptional performance in predicting material properties. The library aims to be an extensible platform for creating and sharing materials graph deep learning models. MatGL includes implementations of various architectures such as M3GNet, MEGNet, CHGNet, QET, and TensorNet, with a transition towards PyTorch Geometric (PyG) as the default backend. It offers pre-trained models, a command-line interface for quick relaxations and predictions, and integration with Hugging Face Hub for model sharing and loading. MatGL is a collaborative effort between Materialyze.AI and Intel Labs, forming part of the broader MatML ecosystem.
Capabilities
Pricing & Plans
Open Source
Free
FAQs
Trending
