Fixes FalconRewardPredictionPolicyTest

PiperOrigin-RevId: 515585162 Change-Id: I1b6485ad513fe6255d0b91e81ceba5afbc52df51
tensorflow · Mar 10, 2023 · 95db333 · 95db333
1 parent 7d8fa0e
commit 95db333
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/tf_agents/bandits/policies/falcon_reward_prediction_policy_test.py b/tf_agents/bandits/policies/falcon_reward_prediction_policy_test.py
@@ -151,7 +151,7 @@ def split_fn(obs):
           time_step_spec=self._time_step_with_mask_spec,
           action_spec=self._action_spec,
           reward_network=DummyNet(self._obs_spec),
-          exploitation_coefficient=0.0,
+          exploitation_coefficient=exploitation_coefficient,
           num_samples_list=num_samples_list,
           emit_policy_info=(utils.InfoFields.LOG_PROBABILITY,),
           observation_and_action_constraint_splitter=split_fn)