def bugged_mlp_actor_critic()

in spinup/exercises/tf1/problem_set_2/exercise2_2.py [0:0]

• 11 lines of code
• 1 McCabe index (conditional complexity)

def bugged_mlp_actor_critic(x, a, hidden_sizes=(400,300), activation=tf.nn.relu, 
                            output_activation=tf.tanh, action_space=None):
    act_dim = a.shape.as_list()[-1]
    act_limit = action_space.high[0]
    with tf.variable_scope('pi'):
        pi = act_limit * mlp(x, list(hidden_sizes)+[act_dim], activation, output_activation)
    with tf.variable_scope('q'):
        q = mlp(tf.concat([x,a], axis=-1), list(hidden_sizes)+[1], activation, None)
    with tf.variable_scope('q', reuse=True):
        q_pi = mlp(tf.concat([x,pi], axis=-1), list(hidden_sizes)+[1], activation, None)
    return pi, q, q_pi