Actor-Critic 法による CartPole の実験

コレクションでコンテンツを整理 必要に応じて、コンテンツの保存と分類を行います。

pip install gym[classic_control]
pip install pyglet

# Install additional packages for visualization
sudo apt-get install -y python-opengl > /dev/null 2>&1
pip install git+https://github.com/tensorflow/docs > /dev/null 2>&1

import collections
import gym
import numpy as np
import statistics
import tensorflow as tf
import tqdm

from matplotlib import pyplot as plt
from tensorflow.keras import layers
from typing import Any, List, Sequence, Tuple


# Create the environment
env = gym.make("CartPole-v1")

# Set seed for experiment reproducibility
seed = 42
tf.random.set_seed(seed)
np.random.seed(seed)

# Small epsilon value for stabilizing division operations
eps = np.finfo(np.float32).eps.item()

2024-01-11 20:25:03.412854: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
2024-01-11 20:25:03.412898: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
2024-01-11 20:25:03.414545: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered

class ActorCritic(tf.keras.Model):
  """Combined actor-critic network."""

  def __init__(
      self, 
      num_actions: int, 
      num_hidden_units: int):
    """Initialize."""
    super().__init__()

    self.common = layers.Dense(num_hidden_units, activation="relu")
    self.actor = layers.Dense(num_actions)
    self.critic = layers.Dense(1)

  def call(self, inputs: tf.Tensor) -> Tuple[tf.Tensor, tf.Tensor]:
    x = self.common(inputs)
    return self.actor(x), self.critic(x)

num_actions = env.action_space.n  # 2
num_hidden_units = 128

model = ActorCritic(num_actions, num_hidden_units)

# Wrap Gym's `env.step` call as an operation in a TensorFlow function.
# This would allow it to be included in a callable TensorFlow graph.

def env_step(action: np.ndarray) -> Tuple[np.ndarray, np.ndarray, np.ndarray]:
  """Returns state, reward and done flag given an action."""

  state, reward, done, truncated, info = env.step(action)
  return (state.astype(np.float32), 
          np.array(reward, np.int32), 
          np.array(done, np.int32))


def tf_env_step(action: tf.Tensor) -> List[tf.Tensor]:
  return tf.numpy_function(env_step, [action], 
                           [tf.float32, tf.int32, tf.int32])

def run_episode(
    initial_state: tf.Tensor,  
    model: tf.keras.Model, 
    max_steps: int) -> Tuple[tf.Tensor, tf.Tensor, tf.Tensor]:
  """Runs a single episode to collect training data."""

  action_probs = tf.TensorArray(dtype=tf.float32, size=0, dynamic_size=True)
  values = tf.TensorArray(dtype=tf.float32, size=0, dynamic_size=True)
  rewards = tf.TensorArray(dtype=tf.int32, size=0, dynamic_size=True)

  initial_state_shape = initial_state.shape
  state = initial_state

  for t in tf.range(max_steps):
    # Convert state into a batched tensor (batch size = 1)
    state = tf.expand_dims(state, 0)

    # Run the model and to get action probabilities and critic value
    action_logits_t, value = model(state)

    # Sample next action from the action probability distribution
    action = tf.random.categorical(action_logits_t, 1)[0, 0]
    action_probs_t = tf.nn.softmax(action_logits_t)

    # Store critic values
    values = values.write(t, tf.squeeze(value))

    # Store log probability of the action chosen
    action_probs = action_probs.write(t, action_probs_t[0, action])

    # Apply action to the environment to get next state and reward
    state, reward, done = tf_env_step(action)
    state.set_shape(initial_state_shape)

    # Store reward
    rewards = rewards.write(t, reward)

    if tf.cast(done, tf.bool):
      break

  action_probs = action_probs.stack()
  values = values.stack()
  rewards = rewards.stack()

  return action_probs, values, rewards

def get_expected_return(
    rewards: tf.Tensor, 
    gamma: float, 
    standardize: bool = True) -> tf.Tensor:
  """Compute expected returns per timestep."""

  n = tf.shape(rewards)[0]
  returns = tf.TensorArray(dtype=tf.float32, size=n)

  # Start from the end of `rewards` and accumulate reward sums
  # into the `returns` array
  rewards = tf.cast(rewards[::-1], dtype=tf.float32)
  discounted_sum = tf.constant(0.0)
  discounted_sum_shape = discounted_sum.shape
  for i in tf.range(n):
    reward = rewards[i]
    discounted_sum = reward + gamma * discounted_sum
    discounted_sum.set_shape(discounted_sum_shape)
    returns = returns.write(i, discounted_sum)
  returns = returns.stack()[::-1]

  if standardize:
    returns = ((returns - tf.math.reduce_mean(returns)) / 
               (tf.math.reduce_std(returns) + eps))

  return returns

huber_loss = tf.keras.losses.Huber(reduction=tf.keras.losses.Reduction.SUM)

def compute_loss(
    action_probs: tf.Tensor,  
    values: tf.Tensor,  
    returns: tf.Tensor) -> tf.Tensor:
  """Computes the combined Actor-Critic loss."""

  advantage = returns - values

  action_log_probs = tf.math.log(action_probs)
  actor_loss = -tf.math.reduce_sum(action_log_probs * advantage)

  critic_loss = huber_loss(values, returns)

  return actor_loss + critic_loss

optimizer = tf.keras.optimizers.Adam(learning_rate=0.01)


@tf.function
def train_step(
    initial_state: tf.Tensor, 
    model: tf.keras.Model, 
    optimizer: tf.keras.optimizers.Optimizer, 
    gamma: float, 
    max_steps_per_episode: int) -> tf.Tensor:
  """Runs a model training step."""

  with tf.GradientTape() as tape:

    # Run the model for one episode to collect training data
    action_probs, values, rewards = run_episode(
        initial_state, model, max_steps_per_episode) 

    # Calculate the expected returns
    returns = get_expected_return(rewards, gamma)

    # Convert training data to appropriate TF tensor shapes
    action_probs, values, returns = [
        tf.expand_dims(x, 1) for x in [action_probs, values, returns]] 

    # Calculate the loss values to update our network
    loss = compute_loss(action_probs, values, returns)

  # Compute the gradients from the loss
  grads = tape.gradient(loss, model.trainable_variables)

  # Apply the gradients to the model's parameters
  optimizer.apply_gradients(zip(grads, model.trainable_variables))

  episode_reward = tf.math.reduce_sum(rewards)

  return episode_reward

%%time

min_episodes_criterion = 100
max_episodes = 10000
max_steps_per_episode = 500

# `CartPole-v1` is considered solved if average reward is >= 475 over 500 
# consecutive trials
reward_threshold = 475
running_reward = 0

# The discount factor for future rewards
gamma = 0.99

# Keep the last episodes reward
episodes_reward: collections.deque = collections.deque(maxlen=min_episodes_criterion)

t = tqdm.trange(max_episodes)
for i in t:
    initial_state, info = env.reset()
    initial_state = tf.constant(initial_state, dtype=tf.float32)
    episode_reward = int(train_step(
        initial_state, model, optimizer, gamma, max_steps_per_episode))

    episodes_reward.append(episode_reward)
    running_reward = statistics.mean(episodes_reward)


    t.set_postfix(
        episode_reward=episode_reward, running_reward=running_reward)

    # Show the average episode reward every 10 episodes
    if i % 10 == 0:
      pass # print(f'Episode {i}: average reward: {avg_reward}')

    if running_reward > reward_threshold and i >= min_episodes_criterion:  
        break

print(f'\nSolved at episode {i}: average reward: {running_reward:.2f}!')

0%|          | 0/10000 [00:00<?, ?it/s]/tmpfs/src/tf_docs_env/lib/python3.9/site-packages/gym/utils/passive_env_checker.py:233: DeprecationWarning: `np.bool8` is a deprecated alias for `np.bool_`.  (Deprecated NumPy 1.24)
  if not isinstance(terminated, (bool, np.bool8)):
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
I0000 00:00:1705004711.393610  544566 device_compiler.h:186] Compiled cluster using XLA!  This line is logged at most once for the lifetime of the process.
  4%|▎         | 367/10000 [01:30<39:29,  4.07it/s, episode_reward=500, running_reward=476]
Solved at episode 367: average reward: 476.04!
CPU times: user 3min 8s, sys: 32 s, total: 3min 40s
Wall time: 1min 30s

# Render an episode and save as a GIF file

from IPython import display as ipythondisplay
from PIL import Image

render_env = gym.make("CartPole-v1", render_mode='rgb_array')

def render_episode(env: gym.Env, model: tf.keras.Model, max_steps: int): 
  state, info = env.reset()
  state = tf.constant(state, dtype=tf.float32)
  screen = env.render()
  images = [Image.fromarray(screen)]

  for i in range(1, max_steps + 1):
    state = tf.expand_dims(state, 0)
    action_probs, _ = model(state)
    action = np.argmax(np.squeeze(action_probs))

    state, reward, done, truncated, info = env.step(action)
    state = tf.constant(state, dtype=tf.float32)

    # Render screen every 10 steps
    if i % 10 == 0:
      screen = env.render()
      images.append(Image.fromarray(screen))

    if done:
      break

  return images


# Save GIF image
images = render_episode(render_env, model, max_steps_per_episode)
image_file = 'cartpole-v1.gif'
# loop=0: loop forever, duration=1: play each frame for 1ms
images[0].save(
    image_file, save_all=True, append_images=images[1:], loop=0, duration=1)

import tensorflow_docs.vis.embed as embed
embed.embed_file(image_file)