91b4b8d0f63b52d51e019c06070c0d0195dd2005,test/test_summaries.py,TestSummaries,test_summaries,#TestSummaries#,33

Before Change

            os.rmdir(path=self.__class__.directory)

        // TODO: "dropout"
        reward_estimation = dict(horizon=dict(
            type="decaying", unit="timesteps", decay="linear", initial_value=2.0, decay_steps=3,
            final_value=4.0
        ))
        baseline_policy = dict(network=dict(type="auto", size=8, depth=1, rnn=1))
        baseline_objective = "value"
        baseline_optimizer = "adam"
        preprocessing = dict(reward=dict(type="clipping", upper=0.25))

After Change

        baseline_objective = "value"
        baseline_optimizer = "adam"
        preprocessing = dict(reward=dict(type="clipping", upper=0.25))
        exploration = dict(
            type="exponential", unit="episodes", num_steps=3, initial_value=2.0, decay_rate=0.5
        )

        agent, environment = self.prepare(
            summarizer=dict(
                directory=self.__class__.directory, labels=[

Instances