def _build_online_policy(self):
policy = policies.BCQPolicy(
a_network=self._agent_module.p_net,
q_network=self._agent_module.q_nets[0][0],
b_network=self._agent_module.b_net,
n=self._n_action_samples,
)
return policy
def _build_test_policies(self):
policy = policies.BCQPolicy(
a_network=self._agent_module.p_net,
q_network=self._agent_module.q_nets[0][0],
b_network=self._agent_module.b_net,
n=self._n_action_samples,
)
self._test_policies['main'] = policy