Multiphysics and Multiscale Mechanics Research Group (M3RG), IIT Delhi

MatSciBERT

Material Science Aware Language Model

rigid_body_dynamics_graph

benchmarking_graph

FNO-StressStrain

Using Fourier Neural Operator to measure the physical fields for a mode-I failure of 2D composites

MDBENCHGNN

ML-DOS

Implementation of unsupervised GNN (GraphSAGE) to generate node representations for disordered systems. OPTICS clustering of obtained node embeddings reveals structure-dynamics correlation.

DiSCoMaT

MaScQA

Physics-Informed-GNNs

# Graph Neural Networks

cod-bench

We present cod-bench containing 12 operators and 10 datasets.

LGNN

Lagrangian Graph Neural Network

SRIMP_Hardness

StrainEvolution

StriderNET

Striding through the energy landscape

BroGNet

MatSci-IE-Challenges

MatSci-LLM

ML-Conductivity

structure_factor

The repository sharing the codes for meta data analysis of glass science research papers published in Elsevier journals

Cementron

afm-ai

chgs-ai

graph_neural_ODE

HGNN

Hamiltonian Graph Neural Network

matsciml

Open MatSci ML Toolkit is a single framework for prototyping and scaling out deep learning models for materials discovery, built on top of OpenCatalyst, PyTorch Lightning, and the Deep Graph Library.

PeriDyn

Understanding-the-Compositional-Control

Understanding the Compositional Control on Electrical, Mechanical, Optical, and Physical Properties of Inorganic Glasses with Interpretable Machine Learning