mirror of
https://github.com/labmlai/annotated_deep_learning_paper_implementations.git
synced 2025-10-31 18:58:43 +08:00
248 lines
9.0 KiB
Python
248 lines
9.0 KiB
Python
"""
|
|
# DQN Experiment with Atari Breakout
|
|
|
|
This experiment trains a Deep Q Network (DQN) to play Atari Breakout game on OpenAI Gym.
|
|
It runs the [game environments on multiple processes](../game.html) to sample efficiently.
|
|
"""
|
|
|
|
import numpy as np
|
|
import torch
|
|
|
|
from labml import tracker, experiment, logger, monit
|
|
from labml_helpers.schedule import Piecewise
|
|
from labml_nn.rl.dqn import QFuncLoss
|
|
from labml_nn.rl.dqn.model import Model
|
|
from labml_nn.rl.dqn.replay_buffer import ReplayBuffer
|
|
from labml_nn.rl.game import Worker
|
|
|
|
# Select device
|
|
if torch.cuda.is_available():
|
|
device = torch.device("cuda:0")
|
|
else:
|
|
device = torch.device("cpu")
|
|
|
|
|
|
def obs_to_torch(obs: np.ndarray) -> torch.Tensor:
|
|
"""Scale observations from `[0, 255]` to `[0, 1]`"""
|
|
return torch.tensor(obs, dtype=torch.float32, device=device) / 255.
|
|
|
|
|
|
class Trainer:
|
|
"""
|
|
## Trainer
|
|
"""
|
|
|
|
def __init__(self):
|
|
# #### Configurations
|
|
|
|
# number of workers
|
|
self.n_workers = 8
|
|
# steps sampled on each update
|
|
self.worker_steps = 4
|
|
# number of training iterations
|
|
self.train_epochs = 8
|
|
|
|
# number of updates
|
|
self.updates = 1_000_000
|
|
# size of mini batch for training
|
|
self.mini_batch_size = 32
|
|
|
|
# exploration as a function of updates
|
|
self.exploration_coefficient = Piecewise(
|
|
[
|
|
(0, 1.0),
|
|
(25_000, 0.1),
|
|
(self.updates / 2, 0.01)
|
|
], outside_value=0.01)
|
|
|
|
# update target network every 250 update
|
|
self.update_target_model = 250
|
|
|
|
# $\beta$ for replay buffer as a function of updates
|
|
self.prioritized_replay_beta = Piecewise(
|
|
[
|
|
(0, 0.4),
|
|
(self.updates, 1)
|
|
], outside_value=1)
|
|
|
|
# Replay buffer with $\alpha = 0.6$. Capacity of the replay buffer must be a power of 2.
|
|
self.replay_buffer = ReplayBuffer(2 ** 14, 0.6)
|
|
|
|
# Model for sampling and training
|
|
self.model = Model().to(device)
|
|
# target model to get $\color{orange}Q(s';\color{orange}{\theta_i^{-}})$
|
|
self.target_model = Model().to(device)
|
|
|
|
# create workers
|
|
self.workers = [Worker(47 + i) for i in range(self.n_workers)]
|
|
|
|
# initialize tensors for observations
|
|
self.obs = np.zeros((self.n_workers, 4, 84, 84), dtype=np.uint8)
|
|
for worker in self.workers:
|
|
worker.child.send(("reset", None))
|
|
for i, worker in enumerate(self.workers):
|
|
self.obs[i] = worker.child.recv()
|
|
|
|
# loss function
|
|
self.loss_func = QFuncLoss(0.99)
|
|
# optimizer
|
|
self.optimizer = torch.optim.Adam(self.model.parameters(), lr=2.5e-4)
|
|
|
|
def _sample_action(self, q_value: torch.Tensor, exploration_coefficient: float):
|
|
"""
|
|
#### $\epsilon$-greedy Sampling
|
|
When sampling actions we use a $\epsilon$-greedy strategy, where we
|
|
take a greedy action with probabiliy $1 - \epsilon$ and
|
|
take a random action with probability $\epsilon$.
|
|
We refer to $\epsilon$ as `exploration_coefficient`.
|
|
"""
|
|
|
|
# Sampling doesn't need gradients
|
|
with torch.no_grad():
|
|
# Sample the action with highest Q-value. This is the greedy action.
|
|
greedy_action = torch.argmax(q_value, dim=-1)
|
|
# Uniformly sample and action
|
|
random_action = torch.randint(q_value.shape[-1], greedy_action.shape, device=q_value.device)
|
|
# Whether to chose greedy action or the random action
|
|
is_choose_rand = torch.rand(greedy_action.shape, device=q_value.device) < exploration_coefficient
|
|
# Pick the action based on `is_choose_rand`
|
|
return torch.where(is_choose_rand, random_action, greedy_action).cpu().numpy()
|
|
|
|
def sample(self, exploration_coefficient: float):
|
|
"""### Sample data"""
|
|
|
|
# This doesn't need gradients
|
|
with torch.no_grad():
|
|
# Sample `worker_steps`
|
|
for t in range(self.worker_steps):
|
|
# Get Q_values for the current observation
|
|
q_value = self.model(obs_to_torch(self.obs))
|
|
# Sample actions
|
|
actions = self._sample_action(q_value, exploration_coefficient)
|
|
|
|
# Run sampled actions on each worker
|
|
for w, worker in enumerate(self.workers):
|
|
worker.child.send(("step", actions[w]))
|
|
|
|
# Collect information from each worker
|
|
for w, worker in enumerate(self.workers):
|
|
# Get results after executing the actions
|
|
next_obs, reward, done, info = worker.child.recv()
|
|
|
|
# Add transition to replay buffer
|
|
self.replay_buffer.add(self.obs[w], actions[w], reward, next_obs, done)
|
|
|
|
# update episode information
|
|
# collect episode info, which is available if an episode finished;
|
|
# this includes total reward and length of the episode -
|
|
# look at `Game` to see how it works.
|
|
if info:
|
|
tracker.add('reward', info['reward'])
|
|
tracker.add('length', info['length'])
|
|
|
|
# update current observation
|
|
self.obs[w] = next_obs
|
|
|
|
def train(self, beta: float):
|
|
"""
|
|
### Train the model
|
|
"""
|
|
for _ in range(self.train_epochs):
|
|
# Sample from priority replay buffer
|
|
samples = self.replay_buffer.sample(self.mini_batch_size, beta)
|
|
# Get the predicted Q-value
|
|
q_value = self.model(obs_to_torch(samples['obs']))
|
|
|
|
# Get the Q-values of the next state for [Double Q-learning](index.html).
|
|
# Gradients shouldn't propagate for these
|
|
with torch.no_grad():
|
|
# Get $\color{cyan}Q(s';\color{cyan}{\theta_i})$
|
|
double_q_value = self.model(obs_to_torch(samples['next_obs']))
|
|
# Get $\color{orange}Q(s';\color{orange}{\theta_i^{-}})$
|
|
target_q_value = self.target_model(obs_to_torch(samples['next_obs']))
|
|
|
|
# Compute Temporal Difference (TD) errors, $\delta$, and the loss, $\mathcal{L}(\theta)$.
|
|
td_errors, loss = self.loss_func(q_value,
|
|
q_value.new_tensor(samples['action']),
|
|
double_q_value, target_q_value,
|
|
q_value.new_tensor(samples['done']),
|
|
q_value.new_tensor(samples['reward']),
|
|
q_value.new_tensor(samples['weights']))
|
|
|
|
# Calculate priorities for replay buffer $p_i = |\delta_i| + \epsilon$
|
|
new_priorities = np.abs(td_errors.cpu().numpy()) + 1e-6
|
|
# Update replay buffer priorities
|
|
self.replay_buffer.update_priorities(samples['indexes'], new_priorities)
|
|
|
|
# Zero out the previously calculated gradients
|
|
self.optimizer.zero_grad()
|
|
# Calculate gradients
|
|
loss.backward()
|
|
# Clip gradients
|
|
torch.nn.utils.clip_grad_norm_(self.model.parameters(), max_norm=0.5)
|
|
# Update parameters based on gradients
|
|
self.optimizer.step()
|
|
|
|
def run_training_loop(self):
|
|
"""
|
|
### Run training loop
|
|
"""
|
|
|
|
# Last 100 episode information
|
|
tracker.set_queue('reward', 100, True)
|
|
tracker.set_queue('length', 100, True)
|
|
|
|
# Copy to target network initially
|
|
self.target_model.load_state_dict(self.model.state_dict())
|
|
|
|
for update in monit.loop(self.updates):
|
|
# $\epsilon$, exploration fraction
|
|
exploration = self.exploration_coefficient(update)
|
|
tracker.add('exploration', exploration)
|
|
# $\beta$ for prioritized replay
|
|
beta = self.prioritized_replay_beta(update)
|
|
tracker.add('beta', beta)
|
|
|
|
# Sample with current policy
|
|
self.sample(exploration)
|
|
|
|
# Start training after the buffer is full
|
|
if self.replay_buffer.is_full():
|
|
# Train the model
|
|
self.train(beta)
|
|
|
|
# Periodically update target network
|
|
if update % self.update_target_model == 0:
|
|
self.target_model.load_state_dict(self.model.state_dict())
|
|
|
|
# Save tracked indicators.
|
|
tracker.save()
|
|
# Add a new line to the screen periodically
|
|
if (update + 1) % 1_000 == 0:
|
|
logger.log()
|
|
|
|
def destroy(self):
|
|
"""
|
|
### Destroy
|
|
Stop the workers
|
|
"""
|
|
for worker in self.workers:
|
|
worker.child.send(("close", None))
|
|
|
|
|
|
def main():
|
|
# Create the experiment
|
|
experiment.create(name='dqn')
|
|
# Initialize the trainer
|
|
m = Trainer()
|
|
# Run and monitor the experiment
|
|
with experiment.start():
|
|
m.run_training_loop()
|
|
# Stop the workers
|
|
m.destroy()
|
|
|
|
|
|
# ## Run it
|
|
if __name__ == "__main__":
|
|
main()
|