Table 9 Hyperparameters considered in the random search and their values
From: Extended study on atomic featurization in graph neural networks for molecular property prediction
Hyperparameter | Values |
|---|---|
Number of conv. layers | 1, 3, 5 |
Number of channels in conv. layers | 16, 64, 256 |
Number of dense layers | 1, 3 |
Size of dense layers | 16, 64, 256 |
Dropout | 0.0, 0.2 |
BatchNorm | True, false |
Batch size | 8, 32, 128 |
Learning rate | 0.01, 0.001, 0.0001, 0.00001, 0.000001 |
Scheduler | No scheduler, decrease after 50% of epochs, decrease after 80% of epochs |