How to access the policy object in the on_train_result callback of Ray RLlib? The callback signature is on_train_result(self, *, algorithm, result, **kwargs)

RLlib
1

code is as followed:
@override(DefaultCallbacks)
def on_train_result(self, *, algorithm, result, **kwargs):
iteration = result.get(“training_iteration”, 0)
policy_map = algorithm.workers.local_worker().policy_map
policy = policy_map.get(“default_policy”)
config = (
PPOConfig()
.api_stack(enable_rl_module_and_learner=True, enable_env_runner_and_connector_v2=True)
.callbacks(IntrinsicRewardCallbacks)