From 9a999621766c7d8ccfbe55948542c3ea6fa8bf08 Mon Sep 17 00:00:00 2001 From: Dominik Roth Date: Thu, 16 Jun 2022 10:30:35 +0200 Subject: [PATCH] Split PG/SAC code --- __init__.py | 2 - trl_pg/__init__.py | 2 + trl_pg/policies.py | 7 + trl_pg/trl_pg.py | 325 +++++++++++++++++++++++++++++ trl_sac/__init__.py | 2 + policies.py => trl_sac/policies.py | 0 dtrl.py => trl_sac/trl_sac.py | 8 +- 7 files changed, 342 insertions(+), 4 deletions(-) delete mode 100644 __init__.py create mode 100644 trl_pg/__init__.py create mode 100644 trl_pg/policies.py create mode 100644 trl_pg/trl_pg.py create mode 100644 trl_sac/__init__.py rename policies.py => trl_sac/policies.py (100%) rename dtrl.py => trl_sac/trl_sac.py (98%) diff --git a/__init__.py b/__init__.py deleted file mode 100644 index c8d01cd..0000000 --- a/__init__.py +++ /dev/null @@ -1,2 +0,0 @@ -from sb3_trl.policies import CnnPolicy, MlpPolicy, MultiInputPolicy -from sb3_trl.trl import TRL diff --git a/trl_pg/__init__.py b/trl_pg/__init__.py new file mode 100644 index 0000000..df30cfc --- /dev/null +++ b/trl_pg/__init__.py @@ -0,0 +1,2 @@ +from sb3.trl_pg.policies import CnnPolicy, MlpPolicy, MultiInputPolicy +from sb3.trl_pg.trl_pg import TRL_PG diff --git a/trl_pg/policies.py b/trl_pg/policies.py new file mode 100644 index 0000000..8b784eb --- /dev/null +++ b/trl_pg/policies.py @@ -0,0 +1,7 @@ +# This file is here just to define MlpPolicy/CnnPolicy +# that work for TRL_PG +from stable_baselines3.common.policies import ActorCriticCnnPolicy, ActorCriticPolicy, MultiInputActorCriticPolicy + +MlpPolicy = ActorCriticPolicy +CnnPolicy = ActorCriticCnnPolicy +MultiInputPolicy = MultiInputActorCriticPolicy diff --git a/trl_pg/trl_pg.py b/trl_pg/trl_pg.py new file mode 100644 index 0000000..d544091 --- /dev/null +++ b/trl_pg/trl_pg.py @@ -0,0 +1,325 @@ +import warnings +from typing import Any, Dict, Optional, Type, Union + +import numpy as np +import torch as th +from gym import spaces +from torch.nn import functional as F + +from stable_baselines3.common.on_policy_algorithm import OnPolicyAlgorithm +from stable_baselines3.common.policies import ActorCriticCnnPolicy, ActorCriticPolicy, BasePolicy, MultiInputActorCriticPolicy +from stable_baselines3.common.type_aliases import GymEnv, MaybeCallback, Schedule +from stable_baselines3.common.utils import explained_variance, get_schedule_fn + + +class TRL_PG(OnPolicyAlgorithm): + """ + Differential Trust Region Layer (TRL) for Policy Gradient (PG) + + Paper: https://arxiv.org/abs/2101.09207 + Code: This implementation borrows (/steals most) code from SB3's PPO implementation https://github.com/DLR-RM/stable-baselines3/blob/master/stable_baselines3/ppo/ppo.py + The implementation of the TRL-specific parts borrows from https://github.com/boschresearch/trust-region-layers/blob/main/trust_region_projections/algorithms/pg/pg.py + + :param policy: The policy model to use (MlpPolicy, CnnPolicy, ...) + :param env: The environment to learn from (if registered in Gym, can be str) + :param learning_rate: The learning rate, it can be a function + of the current progress remaining (from 1 to 0) + :param n_steps: The number of steps to run for each environment per update + (i.e. rollout buffer size is n_steps * n_envs where n_envs is number of environment copies running in parallel) + NOTE: n_steps * n_envs must be greater than 1 (because of the advantage normalization) + See https://github.com/pytorch/pytorch/issues/29372 + :param batch_size: Minibatch size + :param n_epochs: Number of epoch when optimizing the surrogate loss + :param gamma: Discount factor + :param gae_lambda: Factor for trade-off of bias vs variance for Generalized Advantage Estimator + :param clip_range: Clipping parameter, it can be a function of the current progress + remaining (from 1 to 0). + :param clip_range_vf: Clipping parameter for the value function, + it can be a function of the current progress remaining (from 1 to 0). + This is a parameter specific to the OpenAI implementation. If None is passed (default), + no clipping will be done on the value function. + IMPORTANT: this clipping depends on the reward scaling. + :param normalize_advantage: Whether to normalize or not the advantage + :param ent_coef: Entropy coefficient for the loss calculation + :param vf_coef: Value function coefficient for the loss calculation + :param max_grad_norm: The maximum value for the gradient clipping + :param use_sde: Whether to use generalized State Dependent Exploration (gSDE) + instead of action noise exploration (default: False) + :param sde_sample_freq: Sample a new noise matrix every n steps when using gSDE + Default: -1 (only sample at the beginning of the rollout) + :param target_kl: Limit the KL divergence between updates, + because the clipping is not enough to prevent large update + see issue #213 (cf https://github.com/hill-a/stable-baselines/issues/213) + By default, there is no limit on the kl div. + :param tensorboard_log: the log location for tensorboard (if None, no logging) + :param create_eval_env: Whether to create a second environment that will be + used for evaluating the agent periodically. (Only available when passing string for the environment) + :param policy_kwargs: additional arguments to be passed to the policy on creation + :param verbose: the verbosity level: 0 no output, 1 info, 2 debug + :param seed: Seed for the pseudo random generators + :param device: Device (cpu, cuda, ...) on which the code should be run. + Setting it to auto, the code will be run on the GPU if possible. + :param _init_setup_model: Whether or not to build the network at the creation of the instance + """ + + policy_aliases: Dict[str, Type[BasePolicy]] = { + "MlpPolicy": ActorCriticPolicy, + "CnnPolicy": ActorCriticCnnPolicy, + "MultiInputPolicy": MultiInputActorCriticPolicy, + } + + def __init__( + self, + policy: Union[str, Type[ActorCriticPolicy]], + env: Union[GymEnv, str], + learning_rate: Union[float, Schedule] = 3e-4, + n_steps: int = 2048, + batch_size: int = 64, + n_epochs: int = 10, + gamma: float = 0.99, + gae_lambda: float = 0.95, + clip_range: Union[float, Schedule] = 0.2, + clip_range_vf: Union[None, float, Schedule] = None, + normalize_advantage: bool = True, + ent_coef: float = 0.0, + vf_coef: float = 0.5, + max_grad_norm: float = 0.5, + use_sde: bool = False, + sde_sample_freq: int = -1, + target_kl: Optional[float] = None, + tensorboard_log: Optional[str] = None, + create_eval_env: bool = False, + policy_kwargs: Optional[Dict[str, Any]] = None, + verbose: int = 0, + seed: Optional[int] = None, + device: Union[th.device, str] = "auto", + + # Different from PPO: + importance_ratio_clip: Union[float, None] = 0.2 + #TODO: projection: BaseProjectionLayer = None, + + _init_setup_model: bool = True, + ): + + super().__init__( + policy, + env, + learning_rate=learning_rate, + n_steps=n_steps, + gamma=gamma, + gae_lambda=gae_lambda, + ent_coef=ent_coef, + vf_coef=vf_coef, + max_grad_norm=max_grad_norm, + use_sde=use_sde, + sde_sample_freq=sde_sample_freq, + tensorboard_log=tensorboard_log, + policy_kwargs=policy_kwargs, + verbose=verbose, + device=device, + create_eval_env=create_eval_env, + seed=seed, + _init_setup_model=False, + supported_action_spaces=( + spaces.Box, + spaces.Discrete, + spaces.MultiDiscrete, + spaces.MultiBinary, + ), + ) + + # Sanity check, otherwise it will lead to noisy gradient and NaN + # because of the advantage normalization + if normalize_advantage: + assert ( + batch_size > 1 + ), "`batch_size` must be greater than 1. See https://github.com/DLR-RM/stable-baselines3/issues/440" + + if self.env is not None: + # Check that `n_steps * n_envs > 1` to avoid NaN + # when doing advantage normalization + buffer_size = self.env.num_envs * self.n_steps + assert ( + buffer_size > 1 + ), f"`n_steps * n_envs` must be greater than 1. Currently n_steps={self.n_steps} and n_envs={self.env.num_envs}" + # Check that the rollout buffer size is a multiple of the mini-batch size + untruncated_batches = buffer_size // batch_size + if buffer_size % batch_size > 0: + warnings.warn( + f"You have specified a mini-batch size of {batch_size}," + f" but because the `RolloutBuffer` is of size `n_steps * n_envs = {buffer_size}`," + f" after every {untruncated_batches} untruncated mini-batches," + f" there will be a truncated mini-batch of size {buffer_size % batch_size}\n" + f"We recommend using a `batch_size` that is a factor of `n_steps * n_envs`.\n" + f"Info: (n_steps={self.n_steps} and n_envs={self.env.num_envs})" + ) + self.batch_size = batch_size + self.n_epochs = n_epochs + self.clip_range = clip_range + self.clip_range_vf = clip_range_vf + self.normalize_advantage = normalize_advantage + self.target_kl = target_kl + + # Different from PPO: + self.importance_ratio_clip = importance_ratio_clip or 0.0 + + if _init_setup_model: + self._setup_model() + + def _setup_model(self) -> None: + super()._setup_model() + + # Initialize schedules for policy/value clipping + self.clip_range = get_schedule_fn(self.clip_range) + if self.clip_range_vf is not None: + if isinstance(self.clip_range_vf, (float, int)): + assert self.clip_range_vf > 0, "`clip_range_vf` must be positive, " "pass `None` to deactivate vf clipping" + + self.clip_range_vf = get_schedule_fn(self.clip_range_vf) + + def train(self) -> None: + """ + Update policy using the currently gathered rollout buffer. + """ + # Switch to train mode (this affects batch norm / dropout) + self.policy.set_training_mode(True) + # Update optimizer learning rate + self._update_learning_rate(self.policy.optimizer) + # Compute current clip range + clip_range = self.clip_range(self._current_progress_remaining) + # Optional: clip range for the value function + if self.clip_range_vf is not None: + clip_range_vf = self.clip_range_vf(self._current_progress_remaining) + + entropy_losses = [] + pg_losses, value_losses = [], [] + clip_fractions = [] + + continue_training = True + + # train for n_epochs epochs + for epoch in range(self.n_epochs): + approx_kl_divs = [] + # Do a complete pass on the rollout buffer + for rollout_data in self.rollout_buffer.get(self.batch_size): + actions = rollout_data.actions + if isinstance(self.action_space, spaces.Discrete): + # Convert discrete action from float to long + actions = rollout_data.actions.long().flatten() + + # Re-sample the noise matrix because the log_std has changed + if self.use_sde: + self.policy.reset_noise(self.batch_size) + + values, log_prob, entropy = self.policy.evaluate_actions(rollout_data.observations, actions) + values = values.flatten() + # Normalize advantage + advantages = rollout_data.advantages + if self.normalize_advantage: + advantages = (advantages - advantages.mean()) / (advantages.std() + 1e-8) + + # ratio between old and new policy, should be one at the first iteration + ratio = th.exp(log_prob - rollout_data.old_log_prob) + + # clipped surrogate loss + policy_loss_1 = advantages * ratio + policy_loss_2 = advantages * th.clamp(ratio, 1 - clip_range, 1 + clip_range) + policy_loss = -th.min(policy_loss_1, policy_loss_2).mean() + + # Logging + pg_losses.append(policy_loss.item()) + clip_fraction = th.mean((th.abs(ratio - 1) > clip_range).float()).item() + clip_fractions.append(clip_fraction) + + if self.clip_range_vf is None: + # No clipping + values_pred = values + else: + # Clip the different between old and new value + # NOTE: this depends on the reward scaling + values_pred = rollout_data.old_values + th.clamp( + values - rollout_data.old_values, -clip_range_vf, clip_range_vf + ) + # Value loss using the TD(gae_lambda) target + value_loss = F.mse_loss(rollout_data.returns, values_pred) + value_losses.append(value_loss.item()) + + # Entropy loss favor exploration + if entropy is None: + # Approximate entropy when no analytical form + entropy_loss = -th.mean(-log_prob) + else: + entropy_loss = -th.mean(entropy) + + entropy_losses.append(entropy_loss.item()) + + loss = policy_loss + self.ent_coef * entropy_loss + self.vf_coef * value_loss + + # Calculate approximate form of reverse KL Divergence for early stopping + # see issue #417: https://github.com/DLR-RM/stable-baselines3/issues/417 + # and discussion in PR #419: https://github.com/DLR-RM/stable-baselines3/pull/419 + # and Schulman blog: http://joschu.net/blog/kl-approx.html + with th.no_grad(): + log_ratio = log_prob - rollout_data.old_log_prob + approx_kl_div = th.mean((th.exp(log_ratio) - 1) - log_ratio).cpu().numpy() + approx_kl_divs.append(approx_kl_div) + + if self.target_kl is not None and approx_kl_div > 1.5 * self.target_kl: + continue_training = False + if self.verbose >= 1: + print(f"Early stopping at step {epoch} due to reaching max kl: {approx_kl_div:.2f}") + break + + # Optimization step + self.policy.optimizer.zero_grad() + loss.backward() + # Clip grad norm + th.nn.utils.clip_grad_norm_(self.policy.parameters(), self.max_grad_norm) + self.policy.optimizer.step() + + if not continue_training: + break + + self._n_updates += self.n_epochs + explained_var = explained_variance(self.rollout_buffer.values.flatten(), self.rollout_buffer.returns.flatten()) + + # Logs + self.logger.record("train/entropy_loss", np.mean(entropy_losses)) + self.logger.record("train/policy_gradient_loss", np.mean(pg_losses)) + self.logger.record("train/value_loss", np.mean(value_losses)) + self.logger.record("train/approx_kl", np.mean(approx_kl_divs)) + self.logger.record("train/clip_fraction", np.mean(clip_fractions)) + self.logger.record("train/loss", loss.item()) + self.logger.record("train/explained_variance", explained_var) + if hasattr(self.policy, "log_std"): + self.logger.record("train/std", th.exp(self.policy.log_std).mean().item()) + + self.logger.record("train/n_updates", self._n_updates, exclude="tensorboard") + self.logger.record("train/clip_range", clip_range) + if self.clip_range_vf is not None: + self.logger.record("train/clip_range_vf", clip_range_vf) + + def learn( + self, + total_timesteps: int, + callback: MaybeCallback = None, + log_interval: int = 1, + eval_env: Optional[GymEnv] = None, + eval_freq: int = -1, + n_eval_episodes: int = 5, + tb_log_name: str = "PPO", + eval_log_path: Optional[str] = None, + reset_num_timesteps: bool = True, + ) -> "PPO": + + return super().learn( + total_timesteps=total_timesteps, + callback=callback, + log_interval=log_interval, + eval_env=eval_env, + eval_freq=eval_freq, + n_eval_episodes=n_eval_episodes, + tb_log_name=tb_log_name, + eval_log_path=eval_log_path, + reset_num_timesteps=reset_num_timesteps, + ) diff --git a/trl_sac/__init__.py b/trl_sac/__init__.py new file mode 100644 index 0000000..c0e01b7 --- /dev/null +++ b/trl_sac/__init__.py @@ -0,0 +1,2 @@ +from sb3_trl.trl_sac.policies import CnnPolicy, MlpPolicy, MultiInputPolicy +from sb3_trl.trl_sac.trl_sac import TRL_SAC diff --git a/policies.py b/trl_sac/policies.py similarity index 100% rename from policies.py rename to trl_sac/policies.py diff --git a/dtrl.py b/trl_sac/trl_sac.py similarity index 98% rename from dtrl.py rename to trl_sac/trl_sac.py index 07f88d9..2e884b9 100644 --- a/dtrl.py +++ b/trl_sac/trl_sac.py @@ -14,9 +14,13 @@ from stable_baselines3.common.utils import polyak_update from stable_baselines3.sac.policies import CnnPolicy, MlpPolicy, MultiInputPolicy, SACPolicy -class SAC(OffPolicyAlgorithm): +class TRL_SAC(OffPolicyAlgorithm): """ - Soft Actor-Critic (SAC) + Trust Region Layers (TRL) based on SAC (Soft Actor Critic) + This implementation is almost a 1:1-copy of the sb3-code for SAC. + Only minor changes have been made to implement Differential Trust Region Layers + + Description from original SAC implementation: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor, This implementation borrows code from original implementation (https://github.com/haarnoja/sac) from OpenAI Spinning Up (https://github.com/openai/spinningup), from the softlearning repo