Branch: master
Find file History
giuseppefutia and ylfdq1118 [Tutorial] Update prerequisites of README (#380)
* Update prerequisites of README

* dependencies for pytorch models

* dependencies for mxnet models

* minor
Latest commit 6c3dba8 Feb 12, 2019
Permalink
Type Name Latest commit message Commit time
..
Failed to load latest commit information.
README.md [Tutorial] Update prerequisites of README (#380) Feb 12, 2019
configure.py [Model] DGMG Training with Batch Size 1 (#161) Nov 22, 2018
cycles.py
main.py
main_batch.py
model.py
model_batch.py
utils.py

README.md

Learning Deep Generative Models of Graphs

This is an implementation of Learning Deep Generative Models of Graphs by Yujia Li, Oriol Vinyals, Chris Dyer, Razvan Pascanu, Peter Battaglia.

Dependencies

Usage

  • Train with batch size 1: python main.py
  • Train with batch size larger than 1: python main_batch.py.

Performance

90% accuracy for cycles compared with 84% accuracy reported in the original paper.

Speed

On AWS p3.2x instance (w/ V100), one epoch takes ~526s for batch size 1 and takes ~238s for batch size 10.

Acknowledgement

We would like to thank Yujia Li for providing details on the implementation.