Comments (3)
Hi @danushv07
Unfortunately, SequentialModule can only contain EquivariantModules, you can not add a torch.nn.Linear module in it.
If I understand correctly, you want:
- an equivariant architecture,
- followed by a pooling layer
- and a final linear layer (e.g. for classification)
You can achieve something similar this way:
net = SequentialModule(
R2Conv(in_, out_, 3, bias=False),
ReLU(out_, inplace=True),
PointwiseMaxPool(out_, kernel_size=2, stride=2),
GroupPooling(out_),
)
# this is the out_type of the last GroupPooling
final_feature_type = net.out_type
# `out_channels` invariant outputs
output_type = e2cnn.nn.FieldType(s, [s.trivial_repr]*out_channels)
# add the final linear layer as a 1x1 convolution
net.add_module('classifier', R2Conv(final_feature_type, output_type, kernel_size=1)
The final R2Conv will be a 1x1 convolution which just behaves like your torch.nn.Linear, assuming the ouput of PointwiseMaxPool is a 1x1 feature map.
You should now be able to export()
your model.
Note, however, that the output tensor will have shape B x out_channels x 1 x1
rather than B x out_channels
, so you may need to do a manual reshaping.
Is my understanding correct? Does this help?
Best,
Gabriele
from e2cnn.
Thank you for the prompt reply @Gabri95 . The fore mentioned solution does work well. However, if torch.nn.Linear
is required, the entire SequentialModule
along with the linear layer can be wrapped in a torch.nn.Module
and then used options such as .modules()
or .children()
can be used to export the required layers.
from e2cnn.
Hi @danushv07
I am not sure I understood what you mean exactly. Could you share a simple code snippet to illustrate your example?
Thanks,
Gabriele
from e2cnn.
Related Issues (20)
- wrapping pytorch operations - grid_sample HOT 4
- Import Error with Torch 1.9.0+cu111 HOT 2
- equivariant Transformer HOT 5
- ZeroPad2D on GeometricTensor
- Cannot pass weights of R2Conv as a positional argument HOT 2
- Counting FLOPs for e2cnn HOT 1
- equivariance in C8 space HOT 1
- About the equivalence of wide_resnet HOT 5
- Need a size parameter for e2cnn.R2Upsampling Class HOT 1
- about attribute R2conv.filter HOT 2
- Learning of kernels HOT 2
- O(2) group, irreps, and PyTorch DDP. HOT 2
- checking equivariance for the angles that are not 90n HOT 2
- about to set special rotation equivariant HOT 2
- Cannot import name container_abcs in python 3.6 version (e2cnn_py36)
- shriking size of output image
- Use of np.float and np.int etc
- Difference between trivial output type and regular output type with group pooling HOT 1
- Export Linear HOT 2
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from e2cnn.