i-Melt is a physics-guided neural network model, that combines deep neural networks with physical equations to predict the structural, thermodynamic and dynamic properties of aluminosilicate melts and glasses.

• Web app https://i-melt.streamlit.app/
• Documentation (in progress) https://i-melt.readthedocs.io/en/latest/
• Source code: https://github.com/charlesll/i-melt/tree/main/i-melt/src/
• License: https://github.com/charlesll/i-melt/LICENSE
• References: https://i-melt.readthedocs.io/en/latest/references.html
• Bug reports: https://github.com/charlesll/i-melt/issues
• Contact lelosq@ipgp.fr

Charles Le Losq, University Paris Cité, Institut de physique du globe de Paris, lelosq@ipgp.fr

Barbara Baldoni, University Paris Cité, Institut de physique du globe de Paris, baldoni@ipgp.fr

Andrew P. Valentine, Durham University, andrew.valentine@durham.ac.uk