AnnaMats's picture
Second Push
05c9ac2
from typing import cast
import numpy as np
from mlagents_envs.logging_util import get_logger
from mlagents.trainers.buffer import BufferKey
from mlagents.trainers.policy.torch_policy import TorchPolicy
from mlagents.trainers.trainer.off_policy_trainer import OffPolicyTrainer
from mlagents.trainers.optimizer.torch_optimizer import TorchOptimizer
from mlagents.trainers.trajectory import Trajectory, ObsUtil
from mlagents.trainers.behavior_id_utils import BehaviorIdentifiers
from mlagents_envs.base_env import BehaviorSpec
from mlagents.trainers.settings import TrainerSettings
from .dqn_optimizer import DQNOptimizer, DQNSettings, QNetwork
logger = get_logger(__name__)
TRAINER_NAME = "dqn"
class DQNTrainer(OffPolicyTrainer):
"""The DQNTrainer is an implementation of"""
def __init__(
self,
behavior_name: str,
reward_buff_cap: int,
trainer_settings: TrainerSettings,
training: bool,
load: bool,
seed: int,
artifact_path: str,
):
"""
Responsible for collecting experiences and training SAC model.
:param behavior_name: The name of the behavior associated with trainer config
:param reward_buff_cap: Max reward history to track in the reward buffer
:param trainer_settings: The parameters for the trainer.
:param training: Whether the trainer is set for training.
:param load: Whether the model should be loaded.
:param seed: The seed the model will be initialized with
:param artifact_path: The directory within which to store artifacts from this trainer.
"""
super().__init__(
behavior_name,
reward_buff_cap,
trainer_settings,
training,
load,
seed,
artifact_path,
)
self.policy: TorchPolicy = None # type: ignore
self.optimizer: DQNOptimizer = None # type: ignore
def _process_trajectory(self, trajectory: Trajectory) -> None:
"""
Takes a trajectory and processes it, putting it into the replay buffer.
"""
super()._process_trajectory(trajectory)
last_step = trajectory.steps[-1]
agent_id = trajectory.agent_id # All the agents should have the same ID
agent_buffer_trajectory = trajectory.to_agentbuffer()
# Check if we used group rewards, warn if so.
self._warn_if_group_reward(agent_buffer_trajectory)
# Update the normalization
if self.is_training:
self.policy.actor.update_normalization(agent_buffer_trajectory)
self.optimizer.critic.update_normalization(agent_buffer_trajectory)
# Evaluate all reward functions for reporting purposes
self.collected_rewards["environment"][agent_id] += np.sum(
agent_buffer_trajectory[BufferKey.ENVIRONMENT_REWARDS]
)
for name, reward_signal in self.optimizer.reward_signals.items():
evaluate_result = (
reward_signal.evaluate(agent_buffer_trajectory) * reward_signal.strength
)
# Report the reward signals
self.collected_rewards[name][agent_id] += np.sum(evaluate_result)
# Get all value estimates for reporting purposes
(
value_estimates,
_,
value_memories,
) = self.optimizer.get_trajectory_value_estimates(
agent_buffer_trajectory, trajectory.next_obs, trajectory.done_reached
)
if value_memories is not None:
agent_buffer_trajectory[BufferKey.CRITIC_MEMORY].set(value_memories)
for name, v in value_estimates.items():
self._stats_reporter.add_stat(
f"Policy/{self.optimizer.reward_signals[name].name.capitalize()} Value",
np.mean(v),
)
# Bootstrap using the last step rather than the bootstrap step if max step is reached.
# Set last element to duplicate obs and remove dones.
if last_step.interrupted:
last_step_obs = last_step.obs
for i, obs in enumerate(last_step_obs):
agent_buffer_trajectory[ObsUtil.get_name_at_next(i)][-1] = obs
agent_buffer_trajectory[BufferKey.DONE][-1] = False
self._append_to_update_buffer(agent_buffer_trajectory)
if trajectory.done_reached:
self._update_end_episode_stats(agent_id, self.optimizer)
def create_optimizer(self) -> TorchOptimizer:
"""
Creates an Optimizer object
"""
return DQNOptimizer( # type: ignore
cast(TorchPolicy, self.policy), self.trainer_settings # type: ignore
) # type: ignore
def create_policy(
self, parsed_behavior_id: BehaviorIdentifiers, behavior_spec: BehaviorSpec
) -> TorchPolicy:
"""
Creates a policy with a PyTorch backend and give DQN hyperparameters
:param parsed_behavior_id:
:param behavior_spec: specifications for policy construction
:return policy
"""
# initialize online Q-network which works as actor
exploration_initial_eps = cast(
DQNSettings, self.trainer_settings.hyperparameters
).exploration_initial_eps
actor_kwargs = {
"exploration_initial_eps": exploration_initial_eps,
"stream_names": [
signal.value for signal in self.trainer_settings.reward_signals.keys()
],
}
policy = TorchPolicy(
self.seed,
behavior_spec,
self.trainer_settings.network_settings,
actor_cls=QNetwork,
actor_kwargs=actor_kwargs,
)
self.maybe_load_replay_buffer()
return policy
@staticmethod
def get_settings_type():
return DQNSettings
@staticmethod
def get_trainer_name() -> str:
return TRAINER_NAME
def get_type_and_setting():
return {DQNTrainer.get_trainer_name(): DQNTrainer}, {
DQNTrainer.get_trainer_name(): DQNTrainer.get_settings_type()
}