cherisezhu / gil Goto Github PK
View Code? Open in Web Editor NEWThe implementation of our NeurIPS 2020 paper "Graph Geometry Interaction Learning" (GIL)
The implementation of our NeurIPS 2020 paper "Graph Geometry Interaction Learning" (GIL)
Solved! thanks
Hi, when I try to run your code, several problems rise due to the environment.
Thanks very much!
Hey @CheriseZhu, do you also have a variant implemented for Graph classification. I am looking to use this code for the same, and it will be of great help if you can let me know?
No module named 'geoopt.manifolds.poincare'
Thanks for the beautiful work.
But I run into 'tensor size'/'out of bound' error for link prediction task on all datasets, while all is okay with node classification.
This message showed under your example input (say, the disease_lp one):
Traceback (most recent call last):
File "train.py", line 163, in
train(args)
File "train.py", line 96, in train
embeddings = model.encode(data['features'], data['adj_train_norm'])
File "/home/server/code/embedding_methods/GIL/GIL/models/base_models.py", line 34, in encode
h = self.encoder.encode(x, adj)
File "/home/server/code/embedding_methods/GIL/GIL/models/encoders.py", line 341, in encode
return super(GIL, self).encode((x_hyp, x), adj)
File "/home/server/code/embedding_methods/GIL/GIL/models/encoders.py", line 22, in encode
output, _ = self.layers.forward(input)
File "/home/server/miniconda3/envs/hy-torch/lib/python3.6/site-packages/torch/nn/modules/container.py", line 139, in forward
input = module(input)
File "/home/server/miniconda3/envs/hy-torch/lib/python3.6/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
return forward_call(*input, **kwargs)
File "/home/server/code/embedding_methods/GIL/GIL/layers/hyp_layers.py", line 217, in forward
x = self.conv(input_h)
File "/home/server/miniconda3/envs/hy-torch/lib/python3.6/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl
return forward_call(*input, **kwargs)
File "/home/server/code/embedding_methods/GIL/GIL/layers/hyp_layers.py", line 286, in forward
out = self.propagate(edge_index, x=log_x, num_nodes=x.size(0), original_x=x)
File "/home/server/miniconda3/envs/hy-torch/lib/python3.6/site-packages/torch_geometric/nn/conv/message_passing.py", line 262, in propagate
kwargs)
File "/home/server/miniconda3/envs/hy-torch/lib/python3.6/site-packages/torch_geometric/nn/conv/message_passing.py", line 170, in collect
self.set_size(size, dim, data)
File "/home/server/miniconda3/envs/hy-torch/lib/python3.6/site-packages/torch_geometric/nn/conv/message_passing.py", line 135, in set_size
(f'Encountered tensor with size {src.size(self.node_dim)} in '
ValueError: Encountered tensor with size 8 in dimension -2, but expected size 2665.
When it's the airport dataset for lp, the error is
RuntimeError: INDICES element is out of DATA bounds, id=1975 axis_dim=8
Do you have any idea what's the problem?
Many thanks!
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.