I have problem while apply cam on my own custom model, should I change my model to resnet?

[JohnnyM1ng]

Hi torch-cam provider XD
I want to apply cam to my own model(a SE-ResNet which has 6 input channels instead of 3 in ResNet).But I met some problems when I follow your example.
model = se_resnet18(6, 2).eval() cam_extractor = SmoothGradCAMpp(model)
And a RuntimeError shows up like this
RuntimeError: Given groups=1, weight of size [64, 6, 7, 7], expected input[1, 3, 224, 224] to have 6 channels, but got 3 channels instead
I have read one similar discussion ##135 it seems can not solve my problem.Have I missed anything?
Thank you sincerely.

[JohnnyM1ng]

Ok，I have figured out what is wrong, I should set the input_shape XD
cam_extractor = SmoothGradCAMpp(model, input_shape=(6, 64, 64))

[frgfm]

Hey @JohnnyM1ng 👋

Yup exactly, starting to think I should add this as a suggestion in the error message when the shapes don't match haha
Glad you found the answer either way!