KAN Dataset Creation

      I programmed and parallelized a dataset creation script for training a Kolmogorov-Arnold Network (KAN). This script is parallelized to compute 10,000 randomized cubatures across 30 Feynman functions using five methods: H-Adaptive, Vegas, Suave, Divonne, and Cuhre. I prepared and debugged SLURM scripts for large-scale distributed dataset computation.