def make_dqn_agent(self, env, q_func, opt, explorer, rbuf, gpu):
return ResidualDQN(q_func,
opt,
rbuf,
gpu=gpu,
gamma=0.9,
explorer=explorer,
replay_start_size=100,
target_update_interval=100,
grad_scale=1e-1)
def make_dqn_agent(self, env, q_func, opt, explorer, rbuf, gpu):
return ResidualDQN(q_func,
opt,
rbuf,
gpu=gpu,
gamma=0.9,
explorer=explorer,
replay_start_size=100,
target_update_frequency=100,
episodic_update=True,
grad_scale=1e-1)