Python Data Driven framework for acceleration of Phase-Field simulations

The passage describes the development of a numerical framework in Python to create and process large dataset for time-series prediction using Deep Learning algorithms. is generated by solving Cahn-Hilliard equation spinodal decomposition binary alloy labeled train Prior training, dimensionality reduction performed Auto-encoders Principal Component Analysis. identifies three distinct latent dimensions/spaces datasets. primary was running up 10,000 High-Fidelity Phase-Field simulations parallel High-Performance Computing (HPC). compatible with all major operating systems has been thoroughly tested on 3.7 later versions.