ray/rllib/agents/a3c/a3c_torch_policy.py

import gym
from typing import Dict, List, Optional

import ray
from ray.rllib.evaluation.episode import Episode
from ray.rllib.evaluation.postprocessing import (
    compute_gae_for_sample_batch,
    Postprocessing,
)
from ray.rllib.models.action_dist import ActionDistribution
from ray.rllib.models.modelv2 import ModelV2
from ray.rllib.models.torch.torch_action_dist import TorchDistributionWrapper
from ray.rllib.policy.policy import Policy
from ray.rllib.policy.policy_template import build_policy_class
from ray.rllib.policy.sample_batch import SampleBatch
from ray.rllib.policy.torch_policy import LearningRateSchedule, EntropyCoeffSchedule
from ray.rllib.utils.deprecation import Deprecated
from ray.rllib.utils.framework import try_import_torch
from ray.rllib.utils.torch_utils import apply_grad_clipping, sequence_mask
from ray.rllib.utils.typing import (
    TrainerConfigDict,
    TensorType,
    PolicyID,
    LocalOptimizer,
)

torch, nn = try_import_torch()


@Deprecated(
    old="rllib.agents.a3c.a3c_torch_policy.add_advantages",
    new="rllib.evaluation.postprocessing.compute_gae_for_sample_batch",
    error=False,
)
def add_advantages(
    policy: Policy,
    sample_batch: SampleBatch,
    other_agent_batches: Optional[Dict[PolicyID, SampleBatch]] = None,
    episode: Optional[Episode] = None,
) -> SampleBatch:

    return compute_gae_for_sample_batch(
        policy, sample_batch, other_agent_batches, episode
    )


def actor_critic_loss(
    policy: Policy,
    model: ModelV2,
    dist_class: ActionDistribution,
    train_batch: SampleBatch,
) -> TensorType:
    logits, _ = model(train_batch)
    values = model.value_function()

    if policy.is_recurrent():
        B = len(train_batch[SampleBatch.SEQ_LENS])
        max_seq_len = logits.shape[0] // B
        mask_orig = sequence_mask(train_batch[SampleBatch.SEQ_LENS], max_seq_len)
        valid_mask = torch.reshape(mask_orig, [-1])
    else:
        valid_mask = torch.ones_like(values, dtype=torch.bool)

    dist = dist_class(logits, model)
    log_probs = dist.logp(train_batch[SampleBatch.ACTIONS]).reshape(-1)
    pi_err = -torch.sum(
        torch.masked_select(
            log_probs * train_batch[Postprocessing.ADVANTAGES], valid_mask
        )
    )

    # Compute a value function loss.
    if policy.config["use_critic"]:
        value_err = 0.5 * torch.sum(
            torch.pow(
                torch.masked_select(
                    values.reshape(-1) - train_batch[Postprocessing.VALUE_TARGETS],
                    valid_mask,
                ),
                2.0,
            )
        )
    # Ignore the value function.
    else:
        value_err = 0.0

    entropy = torch.sum(torch.masked_select(dist.entropy(), valid_mask))

    total_loss = (
        pi_err
        + value_err * policy.config["vf_loss_coeff"]
        - entropy * policy.entropy_coeff
    )

    # Store values for stats function in model (tower), such that for
    # multi-GPU, we do not override them during the parallel loss phase.
    model.tower_stats["entropy"] = entropy
    model.tower_stats["pi_err"] = pi_err
    model.tower_stats["value_err"] = value_err

    return total_loss


def stats(policy: Policy, train_batch: SampleBatch) -> Dict[str, TensorType]:

    return {
        "cur_lr": policy.cur_lr,
        "entropy_coeff": policy.entropy_coeff,
        "policy_entropy": torch.mean(torch.stack(policy.get_tower_stats("entropy"))),
        "policy_loss": torch.mean(torch.stack(policy.get_tower_stats("pi_err"))),
        "vf_loss": torch.mean(torch.stack(policy.get_tower_stats("value_err"))),
    }


def vf_preds_fetches(
    policy: Policy,
    input_dict: Dict[str, TensorType],
    state_batches: List[TensorType],
    model: ModelV2,
    action_dist: TorchDistributionWrapper,
) -> Dict[str, TensorType]:
    """Defines extra fetches per action computation.

    Args:
        policy (Policy): The Policy to perform the extra action fetch on.
        input_dict (Dict[str, TensorType]): The input dict used for the action
            computing forward pass.
        state_batches (List[TensorType]): List of state tensors (empty for
            non-RNNs).
        model (ModelV2): The Model object of the Policy.
        action_dist (TorchDistributionWrapper): The instantiated distribution
            object, resulting from the model's outputs and the given
            distribution class.

    Returns:
        Dict[str, TensorType]: Dict with extra tf fetches to perform per
            action computation.
    """
    # Return value function outputs. VF estimates will hence be added to the
    # SampleBatches produced by the sampler(s) to generate the train batches
    # going into the loss function.
    return {
        SampleBatch.VF_PREDS: model.value_function(),
    }


def torch_optimizer(policy: Policy, config: TrainerConfigDict) -> LocalOptimizer:
    return torch.optim.Adam(policy.model.parameters(), lr=config["lr"])


class ValueNetworkMixin:
    """Assigns the `_value()` method to the PPOPolicy.

    This way, Policy can call `_value()` to get the current VF estimate on a
    single(!) observation (as done in `postprocess_trajectory_fn`).
    Note: When doing this, an actual forward pass is being performed.
    This is different from only calling `model.value_function()`, where
    the result of the most recent forward pass is being used to return an
    already calculated tensor.
    """

    def __init__(self, obs_space, action_space, config):
        # When doing GAE, we need the value function estimate on the
        # observation.
        if config["use_gae"]:
            # Input dict is provided to us automatically via the Model's
            # requirements. It's a single-timestep (last one in trajectory)
            # input_dict.

            def value(**input_dict):
                input_dict = SampleBatch(input_dict)
                input_dict = self._lazy_tensor_dict(input_dict)
                model_out, _ = self.model(input_dict)
                # [0] = remove the batch dim.
                return self.model.value_function()[0].item()

        # When not doing GAE, we do not require the value function's output.
        else:

            def value(*args, **kwargs):
                return 0.0

        self._value = value


def setup_mixins(
    policy: Policy,
    obs_space: gym.spaces.Space,
    action_space: gym.spaces.Space,
    config: TrainerConfigDict,
) -> None:
    """Call all mixin classes' constructors before PPOPolicy initialization.

    Args:
        policy (Policy): The Policy object.
        obs_space (gym.spaces.Space): The Policy's observation space.
        action_space (gym.spaces.Space): The Policy's action space.
        config (TrainerConfigDict): The Policy's config.
    """
    EntropyCoeffSchedule.__init__(
        policy, config["entropy_coeff"], config["entropy_coeff_schedule"]
    )
    LearningRateSchedule.__init__(policy, config["lr"], config["lr_schedule"])
    ValueNetworkMixin.__init__(policy, obs_space, action_space, config)


A3CTorchPolicy = build_policy_class(
    name="A3CTorchPolicy",
    framework="torch",
    get_default_config=lambda: ray.rllib.agents.a3c.a3c.DEFAULT_CONFIG,
    loss_fn=actor_critic_loss,
    stats_fn=stats,
    postprocess_fn=compute_gae_for_sample_batch,
    extra_action_out_fn=vf_preds_fetches,
    extra_grad_process_fn=apply_grad_clipping,
    optimizer_fn=torch_optimizer,
    before_loss_init=setup_mixins,
    mixins=[ValueNetworkMixin, LearningRateSchedule, EntropyCoeffSchedule],
)