This repository accompanies the paper presenting "Physics-informed machine learning for predicting temperature-dependent chemical properties". By combining established physics-based equations, such as the Arrhenius equation, with machine learning models, this approach encodes temperature dependence directly into the predictive framework. The model predicts the chemistry-dependent coefficients of the equation, enabling accurate and generalizable predictions across diverse chemistries and temperature ranges. The methodology has been validated using experimental data and benchmarked against two different base models.