SPUS PDE-UNet-36M-v2 (FNS-KF) PDE Solving Model

This is a 36M-parameter UNet-based model repository hosted on Hugging Face by siddik-lanl (Los Alamos National Laboratory), version 2 of the SPUS PDE-UNet series with FNS-KF configuration. The model targets partial differential equation solving using deep learning methods, falling under neural operator approaches for scientific machine learning. The repository contains model weights for PDE surrogate modeling tasks.