GPTFF: A high-accuracy out-of-the-box universal AI force field for arbitrary inorganic materials
arXiv:2402.19327 · doi:10.1016/j.scib.2024.08.039
Abstract
This study introduces a novel AI force field, namely graph-based pre-trained transformer force field (GPTFF), which can simulate arbitrary inorganic systems with good precision and generalizability. Harnessing a large trove of the data and the attention mechanism of transformer algorithms, the model can accurately predict energy, atomic forces, and stress with Mean Absolute Error (MAE) values of 32 meV/atom, 71 meV/Å, and 0.365 GPa, respectively. The dataset used to train the model includes 37.8 million single-point energies, 11.7 billion force pairs, and 340.2 million stresses. We also demonstrated that GPTFF can be universally used to simulate various physical systems, such as crystal structure optimization, phase transition simulations, and mass transport.
References in corpus (7)
- A Spectral Analysis Method for Automated Generation of Quantum-Accurate Interatomic Potentials
- A Universal Graph Deep Learning Interatomic Potential for the Periodic Table
- Atomistic Line Graph Neural Network for Improved Materials Property Predictions
- Towards Universal Neural Network Potential for Material Discovery Applicable to Arbitrary Combination of 45 Elements
- Unified Graph Neural Network Force-field for the Periodic Table
- Discrepancies and Error Evaluation Metrics for Machine Learning Interatomic Potentials
- A universal model for the formation energy prediction of inorganic compounds
Cited by in corpus (10)
- A practical guide to machine learning interatomic potentials -- Status and future
- The Evolution of Machine Learning Potentials for Molecules, Reactions and Materials
- AI-Driven Accelerated Discovery of Intercalation-type Cathode Materials for Magnesium Batteries
- DPmoire: A tool for constructing accurate machine learning force fields in moiré systems
- Pre-training, fine-tuning, and distillation (PFD): Automatically generating machine learning force fields from universal models
- Efficient moment tensor machine-learning interatomic potential for accurate description of defects in Ni-Al Alloys
- Machine Learning Potentials for Alloys: A Detailed Workflow to Predict Phase Diagrams and Benchmark Accuracy
- Predicting neutron experiments from first principles: A workflow powered by machine learning
- FastCHGNet: Training one Universal Interatomic Potential to 1.5 Hours with 32 GPUs
- FastTrack: a fast method to evaluate mass transport in solid leveraging universal machine learning interatomic potential