Hello,
I’m pretty new to deep learning, I would be happy to receive any kind of help.
I have an exercise that I need to train a Convolutional Neural Network (CNN) on the CIFAR10 dataset. After building 31 layers I have these layers:
self.conv10 = nn.Conv2d(502, 502, 3, padding=1)
self.bn10 = nn.BatchNorm2d(502)
self.relu10 = nn.ReLU()
self.maxpool5 = nn.MaxPool2d(2, stride = 2)
self.avgpool1 = nn.AdaptiveAvgPool2d(???)
self.dropout = nn.Dropout(0.48254838509851644)
self.fc1 = nn.Linear(??? ,2039)
self.relu11 = nn.ReLU()
self.fc2 = nn.Linear(2039,???)
there are three parameters that I need to compute (all the ???) - one in the AdaptiveAvgPool2d, one in the input of the first Linear, and one in the output of the second Linear. How exactly do I compute these parameters?
Thanks