RuntimeError: mat1 and mat2 shapes cannot be multiplied (1x1 and 2048x1000)

1

I am new to pytorch

class RN50_End(torch.nn.Module):
def init(self, rn50):
super(RN50_End, self).init()
self.avgpool = rn50.avgpool
self.fc = rn50.fc

def forward(self, x):
    x = self.avgpool(x)
    x = torch.flatten(x, 1)
    x = self.fc(x)
    return x

end_x = torch.rand([112, 2048, 7, 7 ], dtype=torch.float32).contiguous()

r50end = RN50_End(rn50)
for i, (images, target) in enumerate(cal_loader):
output = r50end(images)

2

I don’t know how rn50 is defined but the error is raised as a shape mismatch in self.fc as this layer seems to expect an activation input with 2048 features while x has a single feature only.