Closed DQSSSSS closed 1 year ago
Solve it, I deleted the code of the Fourier translation, so I should set allow_unused=True... Sorry...
Set allow_unused=True doesn't solve it, I redesign the input layer code: change self.affine
to weight and bias buffer object, it will keep the graph and doesn't contribute to the gradient.
When I train this model until 1000k images, I received this error message
Here are my running command:
My environment: