Ciclo di training DRL

Per permettere all'agente di fare esperienza ripetutamente nell'ambiente, devi impostare un ciclo di training.

Molti algoritmi di DRL condividono questa struttura di base:

Cicla sugli episodi
All'interno di ogni episodio, cicla sugli step
A ogni step, scegli un'azione, calcola la loss e aggiorna la rete

Ti vengono fornite funzioni segnaposto select_action() e calculate_loss() che consentono l'esecuzione del codice. Anche la Network e l'optimizer definiti nell'esercizio precedente sono a tua disposizione.

Questo esercizio fa parte del corso

Deep Reinforcement Learning in Python

Visualizza il corso

Istruzioni dell'esercizio

Assicurati che il ciclo esterno (sugli episodi) venga eseguito per dieci episodi.
Assicurati che il ciclo interno (sugli step) continui fino al completamento dell'episodio.
Esegui nel contesto env l'azione selezionata da select_action().
Al termine di ogni iterazione del ciclo interno, aggiorna lo stato prima di iniziare lo step successivo.

Esercizio pratico interattivo

Prova a risolvere questo esercizio completando il codice di esempio.

env = gym.make("LunarLander-v2")
# Run ten episodes
for episode in ____:
    state, info = env.reset()
    done = False    
    # Run through steps until done
    while ____:
        action = select_action(network, state)        
        # Take the action
        next_state, reward, terminated, truncated, _ = ____
        done = terminated or truncated        
        loss = calculate_loss(network, state, action, next_state, reward, done)
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()        
        # Update the state
        state = ____
    print(f"Episode {episode} complete.")

Modifica ed esegui il codice

Questo esercizio fa parte del corso

Deep Reinforcement Learning in Python

AvançadoNível de habilidade

4.8+

Inizia il corso gratis

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

Exercise 1: Introduzione al deep reinforcement learning Exercise 2: Configurazione dell'ambiente e della rete neurale Exercise 3: Ciclo di training DRL

Esercizio in corso

Exercise 4: Introduzione al deep Q learning Exercise 5: Deep learning e DQN Exercise 6: L'architettura della Q-Network Exercise 7: Istanziamento della Q-Network Exercise 8: L'algoritmo DQN essenziale Exercise 9: Selezione dell'azione con DQN essenziale Exercise 10: Funzione di loss del DQN essenziale Exercise 11: Addestrare il DQN essenziale

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 with experience replay Exercise 2: The double-ended queue Exercise 3: Experience replay buffer Exercise 4: DQN with experience replay Exercise 5: The complete DQN algorithm Exercise 6: Epsilon-greediness Exercise 7: Fixed Q-targets Exercise 8: Implementing the complete DQN algorithm Exercise 9: Double DQN Exercise 10: Online network and target network in DDQN Exercise 11: Training the double DQN Exercise 12: Prioritized experience replay Exercise 13: Prioritized experience replay buffer Exercise 14: Sampling from the PER buffer Exercise 15: DQN with prioritized experience replay

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!