Table 6.

Hyperparameters and layer Configurations for 2D ResNet.

Hyperparameter	Value
Input size	(32, 25, 256, 43)
Batch size	32
Max epochs	100
Learning rate	5e-5
Optimizer	Adam
Conv2d	in_channels:25, out_channels:64,kernel_size:7,stride:2,padding:3
BatchNorm2d	num_features: 64
ReLU	–
MaxPool2d	kernel_size: 3, stride: 2, padding: 1
BasicBlock Conv2d	in_channels: 64, out_channels: 64, kernel_size: 3, stride: 1, padding: 1
Layer1	output_channels: 64
Layer2	output_channels: 128, stride: 2
Layer3	output_channels: 256, stride: 2
Layer4	output_channels: 512, stride: 2
AdaptiveAvgPool2d	output size: 1
Linear	in_features: 512, out_features: 3
Final output size	(32, 3)