Skip to content

About

Attempted implementation of ''Semi-Supervised Classification with Graph Convolutional Networks'' paper using PyTorch inside docker..

Resources

Stars

1 star

Watchers

1 watching

Forks

Latest commit

 

History

20 Commits

Folders and files

Repository files navigation

docker-gcn

Attempted implementation of ''Semi-Supervised Classification with Graph Convolutional Networks'' paper using PyTorch inside docker..

You need to have nvidia-docker installed.

Run the following command from inside command line from the root directory of the project.

To run the recommendation code sudo make build && sudo make recommend EXPERIMENT_NAME='Your choice of experiment name'

To run for single iteration with no early stopping tolerance and without batch normalization sudo make build && sudo make run-1

To run for single iteration with early stopping tolerance but without batch normalization sudo make build && sudo make run-2

To run for single iteration with no early stopping tolerance but with batch normalization sudo make build && sudo make run-3

To run for 100 iteration with no early stopping tolerance but with batch normalization sudo make build && sudo make run-multi

About

Attempted implementation of ''Semi-Supervised Classification with Graph Convolutional Networks'' paper using PyTorch inside docker..

Resources

Stars

1 star

Watchers

1 watching

Forks

Releases

Packages

Contributors

Languages