Li-Ba-2022-06-12¶

The current structure dataset comprises 38305 structures generated from elemental ICSD structures and binary ICSD prototype alloy structures. A more detailed procedure is found in Phys. Rev. B 102, 174104 (2020).

Predictions using Pareto optimal MLPs

../../_images/prediction-ecoh-volume58.png

The cohesive energy and volume are obtained by performing a local structure optimization from the DFT equilibrium structure. In addition, the DFT equilibrium structure is obtained by optimizing a prototype structure included in ICSD, and the prototype is used as the structure legend in the figure. Therefore, the structure type of the converged structure is sometimes different from that shown in the legend even if the potential energy surface predicted by MLP is almost the same as the true one.

The other properties predicted by each Pareto optimal MLP are available from column Predictions in the following table.

Pareto optimals¶
Name	Time [ms] (1core/36cores)	RMSE [meV/atom]/[eV/A]	Predictions	Files
pair-1	0.027 / 0.005	79.059 / 0.1413	–	–
pair-17	0.027 / 0.005	69.047 / 0.1502	–	–
pair-33	0.028 / 0.005	20.737 / 0.0727	–	–
pair-37	0.039 / 0.005	19.616 / 0.0689	–	–
pair-49	0.041 / 0.005	13.943 / 0.0525	–	–
pair-34	0.043 / 0.008	8.7588 / 0.0475	predictions	`mlp.lammps` `input` `log`
pair-50	0.078 / 0.008	4.5090 / 0.0339	predictions	`mlp.lammps` `input` `log`
pair-51	0.123 / 0.014	4.2728 / 0.0322	predictions	`mlp.lammps` `input` `log`
pair-52	0.172 / 0.019	2.8079 / 0.0339	predictions	`mlp.lammps` `input` `log`
gtinv-737	0.735 / 0.060	2.6932 / 0.0265	predictions	`mlp.lammps` `input` `log`
gtinv-257	0.792 / 0.055	2.0963 / 0.0292	predictions	`mlp.lammps` `input` `log`
gtinv-185	0.805 / 0.056	2.0963 / 0.0292	predictions	`mlp.lammps` `input` `log`
gtinv-416	1.187 / 0.076	2.0149 / 0.0288	predictions	`mlp.lammps` `input` `log`
gtinv-269	1.464 / 0.086	1.1618 / 0.0247	predictions	`mlp.lammps` `input` `log`
gtinv-436	2.447 / 0.145	1.1440 / 0.0237	predictions	`mlp.lammps` `input` `log`
gtinv-441	3.028 / 0.181	1.0586 / 0.0233	predictions	`mlp.lammps` `input` `log`

Column “Time” shows the time required to compute the energy and forces for 1 MD step and 1 atom, which is estimated from a simulation of 10 runs for a structure with 284 atoms using a workstation with Intel(R) Xeon(R) CPU E5-2695 v4 @ 2.10GHz. Note that the MLPs should be carefully used for extreme structures. The MLPs often return meaningless values for them.

All Pareto optimal MLPs are available here.