414041c6dd3e2361a3cc9de1c646dd9114004fe8,rllib/utils/exploration/tests/test_curiosity.py,TestCuriosity,test_curiosity_on_large_frozen_lake,#TestCuriosity#,110

Before Change

                result = trainer.train()
                print(result)
                if result["episode_reward_mean"] >= 0.001:
                    print("Learnt something!")
                    learnt = True
                    break
            trainer.stop()
            self.assertTrue(learnt)

After Change

            config["exploration_config"] = {
                "type": "StochasticSampling",
            }
            trainer = ppo.PPOTrainer(config=config)
            rewards_wo = 0.0
            for _ in range(num_iterations):
                result = trainer.train()
                rewards_wo += result["episode_reward_mean"]
                print(result)
            trainer.stop()
            self.assertTrue(rewards_wo == 0.0)

Instances