GRACE-2L-OAM-L
Predictions
Convex hull distance prediction errors projected onto elements
2 He 0
10 Ne 0
18 Ar 0
36 Kr 0
54 Xe 0.04
57 La 0.0345
58 Ce 0.0326
59 Pr 0.0346
60 Nd 0.0323
61 Pm 0.0324
62 Sm 0.0336
63 Eu 0.0623
64 Gd 0.0445
65 Tb 0.0325
66 Dy 0.0347
67 Ho 0.0309
68 Er 0.0313
69 Tm 0.0315
70 Yb 0.0779
71 Lu 0.0353
86 Rn 0
89 Ac 0.0304
90 Th 0.0432
91 Pa 0.0448
92 U 0.0552
93 Np 0.0905
94 Pu 0.199
95 Am 0
96 Cm 0
97 Bk 0
98 Cf 0
99 Es 0
100 Fm 0
101 Md 0
102 No 0
103 Lr 0
118 Og 0
57-71 La-Lu Lanthanides
89-103 Ac-Lr Actinides
Trained By
- Yury Lysogorskiy (ICAMS, Ruhr University Bochum)
- Anton Bochkarev (ICAMS, Ruhr University Bochum)
Model Info
- Model Version 0.5.2
- Model Type UIP
- Targets EFSG
- Openness OSOD
- Train Task S2EF
- Test Task IS2RE-SR
- Trained for Benchmark Yes
Training Set
OMat24: 101M structures from 3.23M materials
Subsampled Alexandria: 10.4M structures from 3.23M materials
MPtrj: 1.58M structures from 146k materials
Hyperparameters
- max_force:
0.03 - max_steps:
500 - ase_optimizer:
"FIRE" - cell_filter:
"FrechetCellFilter" - graph_construction_radius:
6 - max_neighbors:
null