ResNet Implementation wrong layer dimension for INPUT_SIZE=224

[NiccoBiondi]

Hi, thanks for sharing your repo. When you define the ResNet model in model_resnet.py, you define the last layer of the Backbone as follows:

if input_size[0] == 112:
    self.fc = Linear(2048 * 4 * 4, 512)
else:
    self.fc = Linear(2048 * 8 * 8, 512)

I notice that when the input_size is set to 224, it raises an error about the layer dimensions that I fixed setting:

self.fc = Linear(2048 * 7 * 7, 512)

Please let me know if it’s correct or if it may require further fixing. Regards