09e72f2042e26725632082d93df70cbc86071d38,reagent/gym/tests/preprocessors/test_replay_buffer_inserters.py,TestBasicReplayBufferInserter,test_cartpole,#TestBasicReplayBufferInserter#,12

Before Change

        obs = env.reset()
        terminal = False
        i = 0
        while not terminal and i < 5:
            action = env.action_space.sample()
            next_obs, reward, terminal, _ = env.step(action)
            replay_buffer_inserter(
                replay_buffer, obs, action, reward, terminal, log_prob=0.0
            )
            obs = next_obs

After Change

    def test_cartpole(self):
        env = gym.make("CartPole-v0")
        replay_buffer, inserted = _create_replay_buffer_and_insert(env)
        batch = replay_buffer.sample_transition_batch_tensor(indices=np.array([0]))
        npt.assert_array_almost_equal(
            inserted[0]["observation"], batch.state.squeeze(0)
        )
        npt.assert_array_almost_equal(

Instances