priba / siamese_ged Goto Github PK

View Code? Open in Web Editor NEW

12.0 12.0 11.0 167.93 MB

Learning Graph Distances with Message PassingNeural Networks

License: MIT License

Python 97.17% Shell 2.83%

graph graph-distance neural-message-passing

siamese_ged's People

Contributors

Stargazers

Watchers

Forkers

vishalbelsare asimwadood mjamilbashir99 nadeemkajla kapitsa2811 leizhiqin walexi johannespitz tstadje tianyuzelin asclepiusinformatica

siamese_ged's Issues

RuntimeError: The expanded size of the tensor (1) must match the existing size (64) at non-singleton dimension 0

I'm trying to run the code in the repository and have been running into a number of issues.
Please find below my environment configuration and snapshots of the error.

How to calculate similarity/distance between two graphs using custom data?

Hi,
I have a graph dataset in which for every graph I have

Node feature vector for every node in the graph
Edge matrix for every graph

How can I use your model to calculate distance/similarity between any two graphs.
PS - Dataset is completely unlabeled.

GPU out of memory

We train the model ( train_siamese_net.py ) on the HistoGraph dataset with default parameters, however the it is crashed with out of memory. Our GPU is 1080ti with 12G memory.
What is the parameter you used? and how long it takes for training ?

Recommend Projects