paper

Automated Graph Neural Network Search under Federated Learning Framework

environment

soft：Pytorch, torch_geometric

hard：4 GPU

usage

server end

usage: server.py [-h] [--mode MODE]
                 [--code SUPERMASK SUPERMASK SUPERMASK SUPERMASK SUPERMASK SUPERMASK SUPERMASK SUPERMASK]
                 [--model {fl-rl,fl-darts,fl-agnns,fl-random}]
                 [--dataset {citeseer,physics,pubmed,corafull,cora}]
                 [--client CLIENT] [--save_dir SAVE_DIR]

optional arguments:
  -h, --help            show this help message and exit
  --mode MODE           test mode or not
  --code SUPERMASK SUPERMASK SUPERMASK SUPERMASK SUPERMASK SUPERMASK SUPERMASK SUPERMASK
                        code of son net
  --model {fl-rl,fl-darts,fl-agnns,fl-random}
                        search model
  --dataset {citeseer,physics,pubmed,corafull,cora}
                        used dataset
  --client CLIENT       the number of clients in the search
  --save_dir SAVE_DIR   the directory to save the best code and best population

client end

usage: client.py [-h] [--mode MODE]
                 [--model {fl-random,fl-darts,fl-agnns,fl-rl}]
                 [--client CLIENT]

optional arguments:
  -h, --help            show this help message and exit
  --mode MODE           test mode or not
  --model {fl-random,fl-darts,fl-agnns,fl-rl}
                        search model
  --client CLIENT       the number of clients in the search

example

We prepare cora dataset in this repository and use cora as example.

1.python server.py --mode test --code 5 0 0 0 0 0 0 2 --dataset cora
2.python client.py --mode test

Please open two terminals, input the first command line into one terminal and input the second one into another terminal.

note

1. If connection refused，please run test.py to unify the ip_port.
2. run server.py before client.py