RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [18, 256, 3, 3]] is at version 3; expected version 2 instead. Hint: enable anomaly detection to find the operation that failed to compute its gradient, with torch.autograd.set_detect_anomaly(True).
It is a common issue for pytorch version > 1.6.0. Using pytorch 1.6.0 does not render such a problem.
While for pytorch 1.6.0, another one arises that nn.SiLU is implemented for pytorch >= 1.7.0.
Following the command in ReadMe,
RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [18, 256, 3, 3]] is at version 3; expected version 2 instead. Hint: enable anomaly detection to find the operation that failed to compute its gradient, with torch.autograd.set_detect_anomaly(True).
It is a common issue for pytorch version > 1.6.0. Using pytorch 1.6.0 does not render such a problem.
While for pytorch 1.6.0, another one arises that nn.SiLU is implemented for pytorch >= 1.7.0.