RLEKF: An Optimizer for Deep Potential with Ab Initio Accuracy

Authors: Siyu Hu, Wentao Zhang, Qiuchen Sha, Feng Pan, Lin-Wang Wang, Weile Jia, Guangming Tan, Tong Zhao

AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental numerical experiments are performed on 13 unbiased datasets. Overall, RLEKF converges faster with slightly better accuracy. For example, a test on a typical system, bulk copper, shows that RLEKF converges faster by both the number of training epochs ( 11.67) and wall-clock time ( 1.19). Besides, we theoretically prove that the updates of weights converge and thus are against the gradient exploding problem. Experimental results verify that RLEKF is not sensitive to the initialization of weights.
Researcher Affiliation Academia 1 State Key Lab of Processors, Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China 2University of Chinese Academy of Sciences, Beijing, China 3School of Advanced Materials, Shenzhen Graduate School, Peking University, Shenzhen, China 4Institute of Semiconductors, Chinese Academy of Sciences, Beijing, China
Pseudocode Yes Algorithm 1: RLEKF( ˆY , Y DFT, win, Pin, λin); Algorithm 2: High-level Structure of Training with RLEKF
Open Source Code No The paper states, 'We implement both RLEKF and the baseline Adam in our αDynamics package' but does not provide any link or explicit statement about the public availability of this code.
Open Datasets No Data Description. We choose 13 unbiased and representative datasets of the aforementioned systems with certain specific structures (second column of Tab. 1). For each dataset, snapshots are yielded based on solving ab initio molecular trajectories via PWmat (Jia et al. 2013). While the paper describes the datasets and how they were generated, it does not provide concrete access information (link, DOI, repository) for the datasets themselves.
Dataset Splits Yes For all sub-systems, the former 80% and the latter 20% of the snapshot dataset are used for training and testing, respectively.
Hardware Specification No The paper does not explicitly mention the specific hardware (e.g., CPU, GPU models, or memory specifications) used to run the experiments.
Software Dependencies No The paper mentions implementing in 'αDynamics package' and using 'Tensor Flow and Py Torch' as general frameworks for others, but it does not specify any software dependencies with version numbers.
Experiment Setup Yes Experiment Setting. Set λ1 = 0.98, P0 = I, ν = 0.9987, w0 consistent with Dee PMD-kit (Wang et al. 2018). The network configuration is [1, 25, 25, 25] (embedding net), [400, 50, 50, 50,1] (fitting net). Here, we recommend a relatively universal setting x = 6, y = 4.