This code uses generative adversarial networks to generate diverse task allocation plans for Multi-agent teams.
To change hyperparameters, check out params.py
.
Specifically, params['sim_env']
controls whether we are using the toy environment (with hand-crafted rewards) or the ergodic search environment.
To train the allocation generator and discriminator with the pre-trained reward network weight (as a surrogate approximation to speed up training), run
python train.py
To test the allocation generator, relocate trained weights as logs/test_weights/generator_weight
, and run
python test_alloc.py
(Optional) To retrain the reward network weight, run:
python train_simulation_reward.py
Put the trained weight in logs/reward_logs/reward_weight
for training.
The training data for the reward network is stored in
logs/training_data/*.npy