PaddleMaterials: An End-to-End Deep Learning Toolkit for Materials Science
Abstract
Machine learning supports materials design, atomistic simulation and experimental characterization, yet models and workflows remain fragmented. We present PaddleMaterials, an open-source framework integrating heterogeneous materials-learning tasks while preserving task-specific representations and scientific semantics. It unifies data processing, training, evaluation, pretrained-model management and workflows across six task classes: property prediction, structure generation, machine-learning interatomic potentials, electronic-structure fields, spectrum elucidation and microscopy image processing. Task interfaces record physical quantities, units, aggregation conventions and references, making cross-task composition verifiable. A shared compiler backend enables switching from dynamic-graph to compiled execution through one setting, using graph optimization, operator fusion and hardware-targeted code generation without changing model definitions. Across 65 model instances in 11 groups, compiled execution accelerated every instance in eight groups, with group median speedups of 11%-224% and group maximum speedups of 11%-336%. Case studies on LiMnO2 and toluene link structure generation or spectrum elucidation, interatomic-potential relaxation, property prediction and electron-density prediction, using same-geometry density functional theory (DFT) calculations as references. For LiMnO2, predicted formation energies for generated and relaxed structures differed from same-geometry PBE+U references by at most 7 meV per atom, while the computational cost of model steps was less than one-twentieth of a DFT single-point calculation. PaddleMaterials extends unified materials workflows beyond conventional property, potential and structure-generation tasks while supporting verifiable cross-task operations and compiled execution.
Related
- Deep-Learning Prediction of Dislocation Presence and Position at Grain Boundaries in FCC materials
- BootLoops: an LLM-driven toolkit for exact quantitative science
- Neutral atom quantum computing for materials science and quantum chemistry
- Agentic programs: an emerging form of scientific software in computational materials science
- Mined from Scientific Literature: Process Schemas for Atomic Layer Deposition and Etching in Materials Science
BibTeX
@misc{liu2026paddlematerials,
title = {{PaddleMaterials: An End-to-End Deep Learning Toolkit for Materials Science}},
author = {Xuwei Liu and Zhimin Zhang and Xiaoguang Hu and Zhongwei Liu and Yuanhui Sun and Jun Huo and Hua Wu and Haifeng Wang},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.03154},
eprint = {2611.03154},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.03154},
}
RIS
TY - UNPB
TI - PaddleMaterials: An End-to-End Deep Learning Toolkit for Materials Science
AU - Xuwei Liu
AU - Zhimin Zhang
AU - Xiaoguang Hu
AU - Zhongwei Liu
AU - Yuanhui Sun
AU - Jun Huo
AU - Hua Wu
AU - Haifeng Wang
PY - 2026
DA - 2026/11/02
PB - arXiv
DO - 10.48550/arxiv.2611.03154
UR - https://arxiv.org/abs/2611.03154
AB - Machine learning supports materials design, atomistic simulation and experimental characterization, yet models and workflows remain fragmented. We present PaddleMaterials, an open-source framework integrating heterogeneous materials-learning tasks while preserving task-specific representations and scientific semantics. It unifies data processing, training, evaluation, pretrained-model management and workflows across six task classes: property prediction, structure generation, machine-learning interatomic potentials, electronic-structure fields, spectrum elucidation and microscopy image processing. Task interfaces record physical quantities, units, aggregation conventions and references, making cross-task composition verifiable. A shared compiler backend enables switching from dynamic-graph to compiled execution through one setting, using graph optimization, operator fusion and hardware-targeted code generation without changing model definitions. Across 65 model instances in 11 groups, compiled execution accelerated every instance in eight groups, with group median speedups of 11%-224% and group maximum speedups of 11%-336%. Case studies on LiMnO2 and toluene link structure generation or spectrum elucidation, interatomic-potential relaxation, property prediction and electron-density prediction, using same-geometry density functional theory (DFT) calculations as references. For LiMnO2, predicted formation energies for generated and relaxed structures differed from same-geometry PBE+U references by at most 7 meV per atom, while the computational cost of model steps was less than one-twentieth of a DFT single-point calculation. PaddleMaterials extends unified materials workflows beyond conventional property, potential and structure-generation tasks while supporting verifiable cross-task operations and compiled execution.
ER -