PET-OAM-XL
Predictions
Convex hull distance prediction errors projected onto elements
Model Info
- Model Version 1.0.0
- Model Type Transformer
- Targets EFSG
- Openness OSOD
- Train Task S2EFS
- 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:
0.02 - max_steps:
500 - ase_optimizer:
"FIRE" - cell_filter:
"FrechetCellFilter" - d_pet:
640 - d_node:
2560 - d_feedforward:
1280 - d_head:
640 - num_gnn_layers:
5 - num_attention_layers:
3 - cutoff:
10 - num_neighbors_adaptive:
40 - cutoff_width:
0.5 - cutoff_function:
"Bump"
Dependencies
- upet 0.1.0