Scientiests have increasingly relied on machine learning models to forecast the properties of materials, including electronic band gaps and mechanical characteristics.