91dc4926589071164f2e762198b212b8521966a5,tf_agents/agents/dqn/dqn_agent_test.py,DqnAgentTest,testLossNStepMidMidLastFirst,#DqnAgentTest#Any#,319

Before Change

        optimizer=None,
        n_step_update=3)

    observations = [tf.constant([[1, 2], [3, 4]], dtype=tf.float32)]
    rewards = tf.constant([10, 20], dtype=tf.float32)
    discounts = tf.constant([0.9, 0.9], dtype=tf.float32)
    // MID: use ts.transition
    time_steps = ts.transition(observations, rewards, discounts)

After Change

        optimizer=None,
        n_step_update=3)

    observations = tf.constant([[1, 2], [3, 4]], dtype=tf.float32)
    rewards = tf.constant([10, 20], dtype=tf.float32)
    discounts = tf.constant([0.9, 0.9], dtype=tf.float32)
    // MID: use ts.transition
    time_steps = ts.transition(observations, rewards, discounts)

Instances