|国家预印本平台
| 注册
首页|Training a Foundation Model for Materials on a Budget

Training a Foundation Model for Materials on a Budget

Training a Foundation Model for Materials on a Budget

来源:Arxiv_logoArxiv
英文摘要

Foundation models for materials modeling are advancing quickly, but their training remains expensive, often placing state-of-the-art methods out of reach for many research groups. We introduce Nequix, a compact E(3)-equivariant potential that pairs a simplified NequIP design with modern training practices, including equivariant root-mean-square layer normalization and the Muon optimizer, to retain accuracy while substantially reducing compute requirements. Built in JAX, Nequix has 700K parameters and was trained in 500 A100-GPU hours. On the Matbench-Discovery and MDR Phonon benchmarks, Nequix ranks third overall while requiring less than one quarter of the training cost of most other methods, and it delivers an order-of-magnitude faster inference speed than the current top-ranked model. We release model weights and fully reproducible codebase at https://github.com/atomicarchitects/nequix

Teddy Koker、Tess Smidt

计算技术、计算机技术物理学

Teddy Koker,Tess Smidt.Training a Foundation Model for Materials on a Budget[EB/OL].(2025-08-22)[2025-09-06].https://arxiv.org/abs/2508.16067.点此复制

评论