1d4823c0ec446e93d00df8ca654db4b45b63b3d4,rllib/policy/tests/test_compute_log_likelihoods.py,,do_test_log_likelihood,#Any#Any#Any#Any#Any#Any#,18

Before Change

        if run in [dqn.DQNTrainer, sac.SACTrainer] and fw == "torch":
            continue
        print("Testing {} with framework={}".format(run, fw))
        config["eager"] = fw == "eager"
        config["use_pytorch"] = fw == "torch"

        eager_ctx = None
        if fw == "tf":
            assert not tf.executing_eagerly()
        elif fw == "eager":
            eager_ctx = eager_mode()
            eager_ctx.__enter__()
            assert tf.executing_eagerly()

        trainer = run(config=config, env=env)

        policy = trainer.get_policy()
        vars = policy.get_weights()

After Change

                           continuous=False,
                           layer_key=("fc", (0, 4)),
                           logp_func=None):
    config = config.copy()
    // Run locally.
    config["num_workers"] = 0
    // Env setup.
    if continuous:
        env = "Pendulum-v0"
        obs_batch = preprocessed_obs_batch = np.array([[0.0, 0.1, -0.1]])
    else:
        env = "FrozenLake-v0"
        config["env_config"] = {"is_slippery": False, "map_name": "4x4"}
        obs_batch = np.array([0])
        preprocessed_obs_batch = one_hot(obs_batch, depth=16)

    prev_r = None if prev_a is None else np.array(0.0)

    // Test against all frameworks.
    for fw in framework_iterator(config):
        if run in [dqn.DQNTrainer, sac.SACTrainer] and fw == "torch":
            continue

Instances