I am trying to implement a simple linear network.
The input tensor size is (B,3,64,64)
My network is defined like this.
tensorSize = x.size()
input = tensorSize[0] * tensorSize[1] * tensorSize[2]
self.linear_one = torch.nn.Linear(64, input)
self.linear_two = torch.nn.Linear(input, 32)
self.linear_output = torch.nn.Linear(32, 6)
self.layer_in = self.linear_one(x)
self.layer_in_two = self.linear_two(self.layer_in)
self.layer_out = self.linear_output(self.layer_in_two)
My output is size (B,3,64,6) but it needs to be (B,6)
Why is my network not outputting the correct results?
source https://stackoverflow.com/questions/76984392/the-size-of-tensor-a-100-must-match-the-size-of-tensor-b-64-at-non-singleton
Comments
Post a Comment