Table 2 Hyperparameters and setup for different modules in the QALO optimization simulation

designed system

8 × 8 × 8

lattice constant

3.29 Å

QALO iterations

700

Temperature

300 K

learning rate

0.05

latent space

8

epochs

10,000

optimizer

adam

# of shots for each QA

5

# of QA calls in each QA module (n)

5

growth rate of λ

0.01

λ^max

1000

cutoff

4.6 Å

bispectrum coefficient

6