ElectrophysRL RL agent and activation recordings Training python -um dopatorch.discrete_domains.train Some notes General training and file structure inspired by Google Dopamine DQN implementation is inspired from Google Dopamine A2C implementation is inspired from ikostrikov