You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I found that if I use my own data set and change the class number to 7, an error will be reported:
UserWarning: Using a target size (torch.Size([4, 7, 64, 64])) that is different to the input size (torch.Size([4, 6, 64, 64])). This will likely lead to incorrect results due to broadcasting. Please ensure they have the same size
...
RuntimeError: The size of tensor a (6) must match the size of tensor b (7) at non-singleton dimension 1.
This is because the MCC output code writes the shape as a fixed value 6. The following is the debug report,
I found that if I use my own data set and change the class number to 7, an error will be reported:
This is because the MCC output code writes the shape as a fixed value 6. The following is the debug report,
L14:
def __init__(self, net_name, gpu_id): -> def __init__(self, net_name, gpu_id, num_class):
L25:
L160: self.output_layer = SCAModule(self.project_out, 6) -> self.output_layer = SCAModule(self.project_out, num_class)
L36: self.net = CrowdCounter(self.net_name, self.gpu_id) -> self.net = CrowdCounter(self.net_name, self.gpu_id, self.num_classes)
The text was updated successfully, but these errors were encountered: