From: Designing Ti-6Al-4V microstructure for strain delocalization using neural networks
Optimal values | ||||
---|---|---|---|---|
Hyper-parameter | Possible values | SR | \(\textbf{E}_{\textbf{inSB}}\) | \(\textbf{d}_{\textbf{H}}\) |
Number of layers | [2, 3, 4, 5, 6] | 5 | 2 | 6 |
Learning rate | \([1,5,10,50,100]*10^{-4}\) | \(5*10^{-3}\) | \(5*10^{-3}\) | \(10^{-3}\) |
Number of epochs | [64, 128, 256, 512, 1024, 2048] | 2048 | 2048 | 256 |
Batch size | [8, 16, 32, 64] | 64 | 16 | 32 |
Size of layer 1 | [8, 16, 32, 64] | 8 | 16 | 64 |
Size of layer 2 | [16, 32, 64, 128, 256] | 128 | 256 | 256 |
Size of layer 3 | [16, 32, 64, 128, 256] | 128 | - | 128 |
Size of layer 4 | [16, 32, 64, 128, 256] | 128 | - | 256 |
Size of layer 5 | [16, 32, 64, 128] | 16 | - | 32 |
Size of layer 6 | [8, 16, 32, 64] | - | - | 16 |