train for ORION-NEP

This is the training set for ORION-NEP. Empirical force fields remain the primary tool for large-scale molecular simulation, yet their limited flexibility and transferability often hinder predictive modeling in chemically complex condensed-phase systems. Here, we present ORION, a computationally efficient reactive machine-learning interatomic potential for systems containing C, H, O, N, S, and P, developed within the Neuroevolution Potential (NEP) framework.