Probenahme aus dem Puffer PER

Bevor du die Klasse Prioritized Experience Buffer verwenden kannst, um deinen Agenten zu trainieren, musst du noch die Methode .sample() implementieren. Diese Methode nimmt als Argument die Größe der Stichprobe, die du ziehen willst, und gibt die gesampelten Übergänge als tensors zurück, zusammen mit ihren Indizes im Speicherpuffer und ihrem Wichtigkeitsgewicht.

In deiner Umgebung wurde ein Puffer mit einer Kapazität von 10 geladen, aus dem du Proben nehmen kannst.

Diese Übung ist Teil des Kurses

<Kurs>Deep Reinforcement Learning in Python</Kurs>

Übungsanweisungen

Berechne die Stichprobenwahrscheinlichkeit, die mit jedem Übergang verbunden ist.
Ziehe die Indizes, die den Übergängen in der Stichprobe entsprechen; np.random.choice(a, s, p=p) zieht eine Stichprobe der Größe s mit Ersatz aus dem Array a, basierend auf dem Wahrscheinlichkeitsarray p.
Berechne die Wichtigkeit, die jedem Übergang zugeordnet ist.

Interaktive praktische Übung

Versuche dich an dieser Übung, indem du diesen Beispielcode vervollständigst.

def sample(self, batch_size):
    priorities = np.array(self.priorities)
    # Calculate the sampling probabilities
    probabilities = ____ / np.sum(____)
    # Draw the indices for the sample
    indices = np.random.choice(____)
    # Calculate the importance weights
    weights = (1 / (len(self.memory) * ____)) ** ____
    weights /= np.max(weights)
    states, actions, rewards, next_states, dones = zip(*[self.memory[idx] for idx in indices])
    weights = [weights[idx] for idx in indices]
    states_tensor = torch.tensor(states, dtype=torch.float32)
    rewards_tensor = torch.tensor(rewards, dtype=torch.float32)
    next_states_tensor = torch.tensor(next_states, dtype=torch.float32)
    dones_tensor = torch.tensor(dones, dtype=torch.float32)
    weights_tensor = torch.tensor(weights, dtype=torch.float32)
    actions_tensor = torch.tensor(actions, dtype=torch.long).unsqueeze(1)
    return (states_tensor, actions_tensor, rewards_tensor, next_states_tensor,
            dones_tensor, indices, weights_tensor)

PrioritizedReplayBuffer.sample = sample
print("Sampled transitions:\n", buffer.sample(3))

Code bearbeiten und ausführen

Diese Übung ist Teil des Kurses

<Kurs>Deep Reinforcement Learning in Python</Kurs>

Hohe SchwierigkeitSchwierigkeitsgrad

4.8+

Kurs kostenlos starten

Discover how deep reinforcement learning improves upon traditional Reinforcement Learning while studying and implementing your first Deep Q Learning algorithm.

Exercise 1: Introduction to deep reinforcement learning Exercise 2: Environment and neural network setup Exercise 3: DRL training loop Exercise 4: Introduction to deep Q learning Exercise 5: Deep learning and DQN Exercise 6: The Q-Network architecture Exercise 7: Instantiating the Q-Network Exercise 8: The barebone DQN algorithm Exercise 9: Barebone DQN action selection Exercise 10: Barebone DQN loss function Exercise 11: Training the barebone DQN

Dive into Deep Q-learning by implementing the original DQN algorithm, featuring Experience Replay, epsilon-greediness and fixed Q-targets. Beyond DQN, you will then explore two fascinating extensions that improve the performance and stability of Deep Q-learning: Double DQN and Prioritized Experience Replay.

Exercise 1: DQN mit Erlebniswiedergabe Exercise 2: Die doppelendige Warteschlange Exercise 3: Puffer für Erfahrungswiedergabe Exercise 4: DQN mit Erlebniswiedergabe Exercise 5: Der vollständige DQN Algorithmus Exercise 6: Epsilon-Grausamkeit Exercise 7: Feste Q-Ziele Exercise 8: Den kompletten DQN Algorithmus implementieren Exercise 9: Doppelter DQN Exercise 10: Online-Netzwerk und Zielnetzwerk in DDQN Exercise 11: Das Doppelte trainieren DQN Exercise 12: Priorisierte Erfahrungswiedergabe Exercise 13: Priorisierter Erfahrungswiedergabepuffer Exercise 14: Probenahme aus dem Puffer PER

Aktuelle Übung

Exercise 15: DQN mit priorisierter Erfahrungswiedergabe

Learn about the foundational concepts of policy gradient methods found in DRL. You will begin with the policy gradient theorem, which forms the basis for these methods. Then, you will implement the REINFORCE algorithm, a powerful approach to learning policies. The chapter will then guide you through Actor-Critic methods, focusing on the Advantage Actor-Critic (A2C) algorithm, which combines the strengths of both policy gradient and value-based methods to enhance learning efficiency and stability.

Exercise 1: Introduction to policy gradient Exercise 2: The policy network architecture Exercise 3: Working with discrete distributions Exercise 4: Policy gradient and REINFORCE Exercise 5: Action selection in REINFORCE Exercise 6: Training the REINFORCE algorithm Exercise 7: Advantage Actor Critic Exercise 8: Critic network Exercise 9: Actor Critic loss calculations Exercise 10: Training the A2C algorithm

Explore Proximal Policy Optimization (PPO) for robust DRL performance. Next, you will examine using an entropy bonus in PPO, which encourages exploration by preventing premature convergence to deterministic policies. You'll also learn about batch updates in policy gradient methods. Finally, you will learn about hyperparameter optimization with Optuna, a powerful tool for optimizing performance in your DRL models.

Exercise 1: Proximal policy optimization Exercise 2: The clipped probability ratio Exercise 3: The clipped surrogate objective function Exercise 4: Entropy bonus and PPO Exercise 5: Entropy playground Exercise 6: Training the PPO algorithm Exercise 7: Batch updates in policy gradient Exercise 8: Minibatch and DRL Exercise 9: A2C with batch updates Exercise 10: Hyperparameter optimization with Optuna Exercise 11: Hyperparameter or not?Exercise 12: Hands-on with Optuna Exercise 13: Congratulations!