New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
problem in rotate_impl #29
Comments
#29 will probably fix. Can you send the code you are running so I can add a regression test? |
Not yet fixed, but I wanted to continue on that, which ended up in a separate PR building on yours because I wasn't able to push to your branch - hope that's ok ... #31 ... unfortunately that doesn't fix it yet (((but personally I'm happy to see that I'm not the only one who gets confused over whether to use Here is some test code - 2 versions, on different levels of abstraction:
|
I am not sure it's supposed to work with 2d tensors. Images in torch are represented by 3d tensors in the CHW. I added tests to your PR. |
yeah me neither ... but I checked, it failed with 3d as well ... - although it's true, the error is different ... |
OK, should e working with the last fix I pushed to your PR now |
it does, - awesome! |
The matrix returned from
get_inverse_affine_matrix
is 1-dimensional, and so cannot be reshaped like this:Would have liked to create a PR but this gets rather complex - in Python the value returned from
_get_inverse_affine_matrix
is a list as well, but it seems that in further processing we're doing something differently from Python:The text was updated successfully, but these errors were encountered: