d2l-pytorch
d2l-pytorch copied to clipboard
[possible typo] AdaptiveAvgPool2d or AdaptiveMaxPool2d in NiN?
in Ch09_Modern_Convolutional_Networks/Network_in_Network(NiN).ipynb defines the NiN net model with:
#Global Average Pooling can be achieved by AdaptiveMaxPool2d with output size = (1,1) self.avg1 = nn.AdaptiveMaxPool2d((1,1))
Is it a typo of nn.AdaptiveAvgPool2d or some deeper reasons behind this? Cause using Avg will train a lot slower and get bad train/val accuracy like 0.460 (epoch 5)
Yes, that's a typo in the notebook. Thanks for pointing it out. Would you like to send in a PR fixing it and updating the results after running the notebook?