From 9f4d8645d453cff294c551d6eb8c5dd8fcfa3803 Mon Sep 17 00:00:00 2001 From: wqzhou <33364058+WHUweiqingzhou@users.noreply.github.com> Date: Fri, 1 Nov 2024 16:11:56 +0800 Subject: [PATCH] Docs: update the docs about `DeePKS` (#5385) * update the docs about deepks * update the ref of deepks --- docs/CITATIONS.md | 2 +- docs/advanced/interface/deepks.md | 11 ++++++++--- 2 files changed, 9 insertions(+), 4 deletions(-) diff --git a/docs/CITATIONS.md b/docs/CITATIONS.md index 2294404f53..d54627292d 100644 --- a/docs/CITATIONS.md +++ b/docs/CITATIONS.md @@ -26,7 +26,7 @@ The following references are required to be cited when using ABACUS. Specificall - **If DeePKS is used:** - Wenfei Li, Qi Ou, et al. "DeePKS+ABACUS as a Bridge between Expensive Quantum Mechanical Models and Machine Learning Potentials." . + Wenfei Li, Qi Ou, et al. "DeePKS+ABACUS as a Bridge between Expensive Quantum Mechanical Models and Machine Learning Potentials." J. Phys. Chem. A 126.49 (2022): 9154-9164. - **If hybrid functional is used:** diff --git a/docs/advanced/interface/deepks.md b/docs/advanced/interface/deepks.md index be2cfafa97..3fb2cb5b4c 100644 --- a/docs/advanced/interface/deepks.md +++ b/docs/advanced/interface/deepks.md @@ -1,9 +1,14 @@ # DeePKS -[DeePKS](https://pubs.acs.org/doi/10.1021/acs.jctc.0c00872) is a machine-learning aided density funcitonal model that fits the energy difference between highly accurate but computationally demanding method and effcient but less accurate method via neural-network. As such, the trained DeePKS model can provide highly accurate energetics (and forces) with relatively low computational cost, and can therefore act as a bridge to connect expensive quantum mechanic data and machine-learning-based potentials. While the original framework of DeePKS is for molecular systems, please refer to this [reference](https://arxiv.org/abs/2206.10093) for the application of DeePKS in periodic systems. +[DeePKS](https://pubs.acs.org/doi/10.1021/acs.jctc.0c00872) is a machine-learning (ML) aided density funcitonal model that fits the energy difference between highly accurate but computationally demanding method and effcient but less accurate method via neural-network. Common high-precision methods include hybrid functionals or CCSD-T, while common low-precision methods are LDA/GGA. -Detailed instructions on installing and running DeePKS can be found on this [website](https://deepks-kit.readthedocs.io/en/latest/index.html). An [example](https://github.com/deepmodeling/deepks-kit/tree/abacus/examples/water_single_lda2pbe_abacus) for training DeePKS model with ABACUS is also provided. The DeePKS-related keywords in `INPUT` file can be found [here](http://abacus.deepmodeling.com/en/latest/advanced/input_files/input-main.html#deepks). +As such, the trained DeePKS model can provide highly accurate energetics (and forces/band gap/density) with relatively low computational cost, and can therefore act as a bridge to connect expensive quantum mechanic data and machine-learning-based potentials. +While the original framework of DeePKS is for molecular systems, please refer to this [J. Phys. Chem. A 126.49 (2022): 9154-9164](https://pubs.acs.org/doi/abs/10.1021/acs.jpca.2c05000) for the application of DeePKS in periodic systems. -> Note: Use the LCAO basis for DeePKS-related calculations +Detailed instructions on installing and running DeePKS can be found on this [website](https://deepks-kit.readthedocs.io/en/latest/index.html). The DeePKS-related keywords in `INPUT` file can be found [here](http://abacus.deepmodeling.com/en/latest/advanced/input_files/input-main.html#deepks). An [example](https://github.com/deepmodeling/deepks-kit/tree/abacus/examples/water_single_lda2pbe_abacus) for training DeePKS model with ABACUS is also provided. For practical applications, users can refer to a series of [Notebooks](https://bohrium.dp.tech/collections/1921409690). These Notebooks provide detailed instructions on how to train and use the DeePKS model using perovskite as an example. Currently, these tutorials are available in Chinese, but we plan to release corresponding English versions in the near future. + + + +> Note: DeePKS calculations can only be performed by the LCAO basis.