Build neural networks using TensorFlow and run it in a distributed manner.
All ps tasks and worker tasks run the same codes but we run different command.
For the ps task:
python spoof_nn_dist.py --job_name=ps --task_index=0
For worker tasks:
python spoof_nn_dist.py --job_name=worker --task_index=0
python spoof_nn_dist.py --job_name=worker --task_index=1
python spoof_nn_dist.py --job_name=worker --task_index=2
Some plots from experiments
