runnrairu / generativegraph

Implementation For the paper from DeepMind

[WIP] Pytorch Implementation for Learning Deep Generative Models of Graphs

Yujia Li, Oriol Vinyals, Chris Dyer, Razvan Pascanu, Peter Battaglia (DeepMind,London,UK)
https://arxiv.org/abs/1803.03324

Usage

Enviroments (CPU/GPU)

Pull the Envrioment for this implementation.
docker pull relutropy/research

Run the enviroment.
CPU
docker run -i -t -p 1111:1111 relutropy/research /bin/bash
GPU
nvidia-docker run -i -t -p 1111:1111 relutropy/research /bin/bash

Start jupyter server by below command.
In container $ nohup jupyter notebook --allow-root >> jupyter.log 2>&1 &

Codes on Jupyter Notebooks

To get notebooks execute this command.
In container $ git clone https://github.com/shllln/GenerativeGraph.git

Welcome PR and issue.

This is prtotype.

There are some differences from the paper.
For example, h_v_init has no condition vector.
That will match the paper soon.