e2cnn
e2cnn copied to clipboard
QUVA-Lab
E(2)-Equivariant CNNs Library for Pytorch
#45 This the issue for which this pull request is being raised.
I'm working on optimizing my model inference, trying conversion to torchscript as a first step. When I call torch.jit.script() on my model, I hit: ``` name = '_weights_ranges', item =...
Hi! I am trying to export a network containing linear layers, but I'm getting the following error: File "X/site-packages/escnn/nn/modules/linear.py", line 305, in export linear.weight.data[:] = self TypeError: can't assign a...
## Modifications - Use [`np.ogrid`](https://numpy.org/doc/stable/reference/generated/numpy.ogrid.html) instead of double for loop to speed up the `build_mask` - Inspired by a StackOverflow post [link](https://stackoverflow.com/a/44874588) ## Tests & results #### Code ```python #...
Thanks @Gabri95 et al for providing this library. I however discovered a small bug, that makes `e2cnn` unusable with any `numpy` version higher or equal to 1.24. The `numpy` team...
Hello author! I want to ask, is it feasible if I use reresnet50 as the backbone of DETR and then use the extracted features for a transformer?
