Graphormer is a deep learning package that allows researchers and developers to train custom models for molecule modeling tasks. It aims to accelerate the research and application in AI for molecule science, such as material discovery, drug discovery, etc. Project website.

The model checkpoints for our paper Predicting equilibrium distributions for molecular systems with deep learning is available under directory Distributional-Graphormer.

Advanced pre-trained versions of Graphormer are available exclusively on Azure Quantum Elements.

Model checkpoints for redox potential prediction as produced in Mixed Computational/Experimental Screening for Aqueous Organic Redox Flow Battery Negolytes and further utilized in Closed-Loop Molecular Design with Calibrated Deference are available under directory orfb-design.

Data Summary

https://huggingface.co/microsoft/graphormer/blob/main/data_summary_card.md