saymrwulf/stable-baselines3
1
0
Fork 0
mirror of https://github.com/saymrwulf/stable-baselines3.git synced 2026-07-27 20:02:30 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Multiprocessing support for HerReplayBuffer (#704)

Browse source 
* IM compat. modif from old fork

* mp her working, without offline sampling

* update readme and doc

* fix discrete action/obs space case

* handle offline sampling

* fix pos to be consistent with the old version

* improve typing and docstring

* fix discrete obs special case

* new her, using episode uid

* deal with full buffer

* offline not implemented

* info storage; compute_reward as arg; offline sampling error

* offline sampling; timeout_termination; fix last_trans detection

* rm max_episode_length from tests

* fix loading and loading test

* Fix episode sampling strategy

* Episode interrupted not valid

* Typo

* Fix infos sampling, next_obs desired goals, offline sampling

* update tests for multienvs

* speed up code

* handle timeout sampling when samping

* give up ep_uid for ep_start and ep_lenght

* speed up sampling

* Improve docstring

* Typos and renaming

* Fix typing

* Fix linter warnings

* Renaming + add note

* fix reward type

* Fix future sampling strategy

* Fix future goal selection strategy

* env_fn as lambda

* Re-fix linter warnings

* Formatting

* Fix offline sampling

* restore the initial performance budget

* Remove max_episode_length for HerReplayBuffer kwargs

* SubprcVecEnv compat test

* Dedicated SubrocVecEnv test rm n_envs from parametrization

* Back to using the env arg instead of compute_reward

* Up VecEnv import

* fix lint warnings

* fix docstring

* Fix device issue

* actor_loss_modifier in SAV and TD3

* Merge RewardModifier and ActorLossModifier into Surgeon

* update surgeon for rnd

* fix uninteded merge

* fix uninteded merge

* fix unintended merge

* Rm unintended merge

* Fix KeyError

* Remove useless `all_inds`

* Minor docstring format

* Fix hint

* speedup!

* Speedup again

* speedup

* np.nonzero

* fix env normalization

* flat sampling for speedup

* typo

* drop online

* format

* remove observation from env_cheker (see #1335)

* update changelog

* default device to "auto"

* add comment for info storage

* add comment for ep_start and ep_length attributes

* a[b][c] to a[b, c]

* comment flatnonzero and unravel_index

* update _sample_goals docstring

* Fix future gaol sampling for split episode

* add informative error message for learning_starts too small

* use keyword arg for env

* try fix pytye

* Update stable_baselines3/common/off_policy_algorithm.py

Co-authored-by: Antonin RAFFIN <antonin.raffin@ensta.org>

* Add `copy_info_dict` option

* Ignore pytype

* Update changelog

* Rename variables and improve documentation

* Ignore new bug bear rule

* Add note about future strategy

* Add deprecation warning

* Fix bug trying to pickle buffer kwargs

---------

Co-authored-by: Antonin RAFFIN <antonin.raffin@ensta.org>
This commit is contained in:
Quentin Gallouédec 2023-03-20 12:03:57 +01:00 committed by GitHub
parent e5deeed16e
commit c5adad82b2
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
14 changed files with 426 additions and 627 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
README.md
View file

@ -173,7 +173,7 @@ All the following examples can be executed online using Google Colab notebooks:
| A2C | :x: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| DDPG | :x: | :heavy_check_mark: | :x: | :x: | :x: | :heavy_check_mark: |
| DQN | :x: | :x: | :heavy_check_mark: | :x: | :x: | :heavy_check_mark: |
| HER | :x: | :heavy_check_mark: | :heavy_check_mark: | :x: | :x: | :x: |
| HER | :x: | :heavy_check_mark: | :heavy_check_mark: | :x: | :x: | :heavy_check_mark: |
| PPO | :x: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |
| QR-DQN<sup>[1](#f1)</sup> | :x: | :x: | :heavy_check_mark: | :x: | :x: | :heavy_check_mark: |
| RecurrentPPO<sup>[1](#f1)</sup> | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: |

2
docs/guide/algos.rst
View file

@ -12,7 +12,7 @@ ARS [#f1]_ ✔️ ✔️ ❌ ❌
A2C ✔️ ✔️ ✔️ ✔️ ✔️
DDPG ✔️ ❌ ❌ ❌ ✔️
DQN ❌ ✔️ ❌ ❌ ✔️
HER ✔️ ✔️ ❌ ❌ ❌
HER ✔️ ✔️ ❌ ❌ ✔️
PPO ✔️ ✔️ ✔️ ✔️ ✔️
QR-DQN [#f1]_ ❌ ️ ✔️ ❌ ❌ ✔️
RecurrentPPO [#f1]_ ✔️ ✔️ ✔️ ✔️ ✔️

4
docs/guide/examples.rst
View file

@ -450,10 +450,6 @@ The parking env is a goal-conditioned continuous control task, in which the vehi
replay_buffer_kwargs=dict(
n_sampled_goal=n_sampled_goal,
goal_selection_strategy="future",
# IMPORTANT: because the env is not wrapped with a TimeLimit wrapper
# we have to manually specify the max number of steps per episode
max_episode_length=100,
online_sampling=True,
),
verbose=1,
buffer_size=int(1e6),

4
docs/guide/migration.rst
View file

@ -177,10 +177,8 @@ Despite this change, no change in performance should be expected.
HER
^^^
The ``HER`` implementation now also supports online sampling of the new goals. This is done in a vectorized version.
The ``HER`` implementation now only supports online sampling of the new goals. This is done in a vectorized version.
The goal selection strategy ``RANDOM`` is no longer supported.
``HER`` now supports ``VecNormalize`` wrapper but only when ``online_sampling=True``.
For performance reasons, the maximum number of steps per episodes must be specified (see :ref:`HER <her>` documentation).
New logger API

8
docs/misc/changelog.rst
View file

@ -4,7 +4,7 @@ Changelog
==========
Release 1.8.0a9 (WIP)
Release 1.8.0a10 (WIP)
--------------------------
.. warning::
@ -20,12 +20,18 @@ Breaking Changes:
- Removed shared layers in ``mlp_extractor`` (@AlexPasqua)
- Refactored ``StackedObservations`` (it now handles dict obs, ``StackedDictObservations`` was removed)
- You must now explicitely pass a ``features_extractor`` parameter when calling ``extract_features()``
- Dropped offline sampling for ``HerReplayBuffer``
- As ``HerReplayBuffer`` was refactored to support multiprocessing, previous replay buffer are incompatible with this new version
- ``HerReplayBuffer`` doesn't require a ``max_episode_length`` anymore
New Features:
^^^^^^^^^^^^^
- Added ``repeat_action_probability`` argument in ``AtariWrapper``.
- Only use ``NoopResetEnv`` and ``MaxAndSkipEnv`` when needed in ``AtariWrapper``
- Added support for dict/tuple observations spaces for ``VecCheckNan``, the check is now active in the ``env_checker()`` (@DavyMorgan)
- Added multiprocessing support for ``HerReplayBuffer``
- ``HerReplayBuffer`` now supports all datatypes supported by ``ReplayBuffer``
`SB3-Contrib`_
^^^^^^^^^^^^^^

21
docs/modules/her.rst
View file

@ -27,14 +27,6 @@ It creates "virtual" transitions by relabeling transitions (changing the desired
- a dictionary observation space with three keys: ``observation``, ``achieved_goal`` and ``desired_goal``
.. warning::
For performance reasons, the maximum number of steps per episodes must be specified.
In most cases, it will be inferred if you specify ``max_episode_steps`` when registering the environment
or if you use a ``gym.wrappers.TimeLimit`` (and ``env.spec`` is not None).
Otherwise, you can directly pass ``max_episode_length`` to the model constructor
.. warning::
Because it needs access to ``env.compute_reward()``
@ -42,6 +34,12 @@ It creates "virtual" transitions by relabeling transitions (changing the desired
without instantiating the environment, we recommend saving the policy only.
.. note::
Compared to other implementations, the ``future`` goal sampling strategy is inclusive:
the current transition can be used when re-sampling.
Notes
-----
@ -77,11 +75,6 @@ This example is only to demonstrate the use of the library and its functions, an
# Available strategies (cf paper): future, final, episode
goal_selection_strategy = "future" # equivalent to GoalSelectionStrategy.FUTURE
# If True the HER transitions will get sampled online
online_sampling = True
# Time limit for the episodes
max_episode_length = N_BITS
# Initialize the model
model = model_class(
"MultiInputPolicy",
@ -91,8 +84,6 @@ This example is only to demonstrate the use of the library and its functions, an
replay_buffer_kwargs=dict(
n_sampled_goal=4,
goal_selection_strategy=goal_selection_strategy,
online_sampling=online_sampling,
max_episode_length=max_episode_length,
),
verbose=1,
)

3
pyproject.toml
View file

@ -5,7 +5,8 @@ line-length = 127
target-version = "py37"
# See https://beta.ruff.rs/docs/rules/
select = ["E", "F", "B", "UP", "C90", "RUF"]
ignore = []
# Ignore explicit stacklevel`
ignore = ["B028"]
[tool.ruff.per-file-ignores]
# Default implementation in abstract methods

2
stable_baselines3/common/env_checker.py
View file

@ -117,7 +117,7 @@ def _check_goal_env_obs(obs: dict, observation_space: spaces.Dict, method_name:
f"The current observation contains {len(observation_space.spaces)} keys: {list(observation_space.spaces.keys())}"
)
for key in ["observation", "achieved_goal", "desired_goal"]:
for key in ["achieved_goal", "desired_goal"]:
if key not in observation_space.spaces:
raise AssertionError(
f"The observation returned by the `{method_name}()` method of a goal-conditioned env requires the '{key}' "

58
stable_baselines3/common/off_policy_algorithm.py
View file

@ -125,10 +125,14 @@ class OffPolicyAlgorithm(BaseAlgorithm):
self.gradient_steps = gradient_steps
self.action_noise = action_noise
self.optimize_memory_usage = optimize_memory_usage
self.replay_buffer_class = replay_buffer_class
if replay_buffer_kwargs is None:
replay_buffer_kwargs = {}
self.replay_buffer_kwargs = replay_buffer_kwargs
if replay_buffer_class is None:
if isinstance(self.observation_space, spaces.Dict):
self.replay_buffer_class = DictReplayBuffer
else:
self.replay_buffer_class = ReplayBuffer
else:
self.replay_buffer_class = replay_buffer_class
self.replay_buffer_kwargs = replay_buffer_kwargs or {}
self._episode_storage = None
# Save train freq parameter, will be converted later to TrainFreq object
@ -170,37 +174,13 @@ class OffPolicyAlgorithm(BaseAlgorithm):
self._setup_lr_schedule()
self.set_random_seed(self.seed)
# Use DictReplayBuffer if needed
if self.replay_buffer_class is None:
if isinstance(self.observation_space, spaces.Dict):
self.replay_buffer_class = DictReplayBuffer
else:
self.replay_buffer_class = ReplayBuffer
elif self.replay_buffer_class == HerReplayBuffer:
assert self.env is not None, "You must pass an environment when using `HerReplayBuffer`"
# If using offline sampling, we need a classic replay buffer too
if self.replay_buffer_kwargs.get("online_sampling", True):
replay_buffer = None
else:
replay_buffer = DictReplayBuffer(
self.buffer_size,
self.observation_space,
self.action_space,
device=self.device,
optimize_memory_usage=self.optimize_memory_usage,
)
self.replay_buffer = HerReplayBuffer(
self.env,
self.buffer_size,
device=self.device,
replay_buffer=replay_buffer,
**self.replay_buffer_kwargs,
)
if self.replay_buffer is None:
# Make a local copy as we should not pickle
# the environment when using HerReplayBuffer
replay_buffer_kwargs = self.replay_buffer_kwargs.copy()
if issubclass(self.replay_buffer_class, HerReplayBuffer):
assert self.env is not None, "You must pass an environment when using `HerReplayBuffer`"
replay_buffer_kwargs["env"] = self.env
self.replay_buffer = self.replay_buffer_class(
self.buffer_size,
self.observation_space,
@ -208,7 +188,7 @@ class OffPolicyAlgorithm(BaseAlgorithm):
device=self.device,
n_envs=self.n_envs,
optimize_memory_usage=self.optimize_memory_usage,
**self.replay_buffer_kwargs,
**replay_buffer_kwargs, # pytype:disable=wrong-keyword-args
)
self.policy = self.policy_class( # pytype:disable=not-instantiable
@ -276,12 +256,7 @@ class OffPolicyAlgorithm(BaseAlgorithm):
# when using memory efficient replay buffer
# see https://github.com/DLR-RM/stable-baselines3/issues/46
# Special case when using HerReplayBuffer,
# the classic replay buffer is inside it when using offline sampling
if isinstance(self.replay_buffer, HerReplayBuffer):
replay_buffer = self.replay_buffer.replay_buffer
else:
replay_buffer = self.replay_buffer
replay_buffer = self.replay_buffer
truncate_last_traj = (
self.optimize_memory_usage
@ -552,7 +527,6 @@ class OffPolicyAlgorithm(BaseAlgorithm):
callback.on_rollout_start()
continue_training = True
while should_collect_more_steps(train_freq, num_collected_steps, num_collected_episodes):
if self.use_sde and self.sde_sample_freq > 0 and num_collected_steps % self.sde_sample_freq == 0:
# Sample a new noise matrix

686
stable_baselines3/her/her_replay_buffer.py
View file

@ -1,87 +1,84 @@
import copy
import warnings
from collections import deque
from typing import Any, Dict, List, Optional, Tuple, Union
from typing import Any, Dict, List, Optional, Union
import numpy as np
import torch as th
from gym import spaces
from stable_baselines3.common.buffers import DictReplayBuffer
from stable_baselines3.common.preprocessing import get_obs_shape
from stable_baselines3.common.type_aliases import DictReplayBufferSamples
from stable_baselines3.common.type_aliases import DictReplayBufferSamples, TensorDict
from stable_baselines3.common.vec_env import VecEnv, VecNormalize
from stable_baselines3.her.goal_selection_strategy import KEY_TO_GOAL_STRATEGY, GoalSelectionStrategy
def get_time_limit(env: VecEnv, current_max_episode_length: Optional[int]) -> int:
"""
Get time limit from environment.
:param env: Environment from which we want to get the time limit.
:param current_max_episode_length: Current value for max_episode_length.
:return: max episode length
"""
# try to get the attribute from environment
if current_max_episode_length is None:
try:
current_max_episode_length = env.get_attr("spec")[0].max_episode_steps
# Raise the error because the attribute is present but is None
if current_max_episode_length is None:
raise AttributeError
# if not available check if a valid value was passed as an argument
except AttributeError as e:
raise ValueError(
"The max episode length could not be inferred.\n"
"You must specify a `max_episode_steps` when registering the environment,\n"
"use a `gym.wrappers.TimeLimit` wrapper "
"or pass `max_episode_length` to the model constructor"
) from e
return current_max_episode_length
class HerReplayBuffer(DictReplayBuffer):
"""
Hindsight Experience Replay (HER) buffer.
Paper: https://arxiv.org/abs/1707.01495
.. warning::
For performance reasons, the maximum number of steps per episodes must be specified.
In most cases, it will be inferred if you specify ``max_episode_steps`` when registering the environment
or if you use a ``gym.wrappers.TimeLimit`` (and ``env.spec`` is not None).
Otherwise, you can directly pass ``max_episode_length`` to the replay buffer constructor.
Replay buffer for sampling HER (Hindsight Experience Replay) transitions.
In the online sampling case, these new transitions will not be saved in the replay buffer
and will only be created at sampling time.
.. note::
Compared to other implementations, the ``future`` goal sampling strategy is inclusive:
the current transition can be used when re-sampling.
:param buffer_size: Max number of element in the buffer
:param observation_space: Observation space
:param action_space: Action space
:param env: The training environment
:param buffer_size: The size of the buffer measured in transitions.
:param max_episode_length: The maximum length of an episode. If not specified,
it will be automatically inferred if the environment uses a ``gym.wrappers.TimeLimit`` wrapper.
:param goal_selection_strategy: Strategy for sampling goals for replay.
One of ['episode', 'final', 'future']
:param device: PyTorch device
:param n_sampled_goal: Number of virtual transitions to create per real transition,
by sampling new goals.
:param n_envs: Number of parallel environments
:param optimize_memory_usage: Enable a memory efficient variant
Disabled for now (see https://github.com/DLR-RM/stable-baselines3/pull/243#discussion_r531535702)
:param handle_timeout_termination: Handle timeout termination (due to timelimit)
separately and treat the task as infinite horizon task.
https://github.com/DLR-RM/stable-baselines3/issues/284
:param n_sampled_goal: Number of virtual transitions to create per real transition,
by sampling new goals.
:param goal_selection_strategy: Strategy for sampling goals for replay.
One of ['episode', 'final', 'future']
:param copy_info_dict: Whether to copy the info dictionary and pass it to
``compute_reward()`` method.
Please note that the copy may cause a slowdown.
False by default.
"""
def __init__(
self,
env: VecEnv,
buffer_size: int,
observation_space: spaces.Space,
action_space: spaces.Space,
env: VecEnv,
device: Union[th.device, str] = "auto",
replay_buffer: Optional[DictReplayBuffer] = None,
max_episode_length: Optional[int] = None,
n_envs: int = 1,
optimize_memory_usage: bool = False,
handle_timeout_termination: bool = True,
n_sampled_goal: int = 4,
goal_selection_strategy: Union[GoalSelectionStrategy, str] = "future",
online_sampling: bool = True,
handle_timeout_termination: bool = True,
copy_info_dict: bool = False,
online_sampling: Optional[bool] = None,
):
super().__init__(buffer_size, env.observation_space, env.action_space, device, env.num_envs)
super().__init__(
buffer_size,
observation_space,
action_space,
device=device,
n_envs=n_envs,
optimize_memory_usage=optimize_memory_usage,
handle_timeout_termination=handle_timeout_termination,
)
self.env = env
self.copy_info_dict = copy_info_dict
if online_sampling is not None:
assert online_sampling is True, "Since v1.8.0, SB3 only supports online sampling with HerReplayBuffer."
warnings.warn(
"Since v1.8.0, the `online_sampling` argument is deprecated "
"as SB3 only supports online sampling with HerReplayBuffer. It will be removed in v2.0",
stacklevel=1,
)
# convert goal_selection_strategy into GoalSelectionStrategy if string
if isinstance(goal_selection_strategy, str):
@ -95,67 +92,24 @@ class HerReplayBuffer(DictReplayBuffer):
), f"Invalid goal selection strategy, please use one of {list(GoalSelectionStrategy)}"
self.n_sampled_goal = n_sampled_goal
# if we sample her transitions online use custom replay buffer
self.online_sampling = online_sampling
# compute ratio between HER replays and regular replays in percent for online HER sampling
# Compute ratio between HER replays and regular replays in percent
self.her_ratio = 1 - (1.0 / (self.n_sampled_goal + 1))
# maximum steps in episode
self.max_episode_length = get_time_limit(env, max_episode_length)
# storage for transitions of current episode for offline sampling
# for online sampling, it replaces the "classic" replay buffer completely
her_buffer_size = buffer_size if online_sampling else self.max_episode_length
self.env = env
self.buffer_size = her_buffer_size
if online_sampling:
replay_buffer = None
self.replay_buffer = replay_buffer
self.online_sampling = online_sampling
# Handle timeouts termination properly if needed
# see https://github.com/DLR-RM/stable-baselines3/issues/284
self.handle_timeout_termination = handle_timeout_termination
# buffer with episodes
# number of episodes which can be stored until buffer size is reached
self.max_episode_stored = self.buffer_size // self.max_episode_length
self.current_idx = 0
# Counter to prevent overflow
self.episode_steps = 0
# Get shape of observation and goal (usually the same)
self.obs_shape = get_obs_shape(self.env.observation_space.spaces["observation"])
self.goal_shape = get_obs_shape(self.env.observation_space.spaces["achieved_goal"])
# input dimensions for buffer initialization
input_shape = {
"observation": (self.env.num_envs, *self.obs_shape),
"achieved_goal": (self.env.num_envs, *self.goal_shape),
"desired_goal": (self.env.num_envs, *self.goal_shape),
"action": (self.action_dim,),
"reward": (1,),
"next_obs": (self.env.num_envs, *self.obs_shape),
"next_achieved_goal": (self.env.num_envs, *self.goal_shape),
"next_desired_goal": (self.env.num_envs, *self.goal_shape),
"done": (1,),
}
self._observation_keys = ["observation", "achieved_goal", "desired_goal"]
self._buffer = {
key: np.zeros((self.max_episode_stored, self.max_episode_length, *dim), dtype=np.float32)
for key, dim in input_shape.items()
}
# Store info dicts are it can be used to compute the reward (e.g. continuity cost)
self.info_buffer = [deque(maxlen=self.max_episode_length) for _ in range(self.max_episode_stored)]
# episode length storage, needed for episodes which has less steps than the maximum length
self.episode_lengths = np.zeros(self.max_episode_stored, dtype=np.int64)
# In some environments, the info dict is used to compute the reward. Then, we need to store it.
self.infos = np.array([[{} for _ in range(self.n_envs)] for _ in range(self.buffer_size)])
# To create virtual transitions, we need to know for each transition
# when an episode starts and ends.
# We use the following arrays to store the indices,
# and update them when an episode ends.
self.ep_start = np.zeros((self.buffer_size, self.n_envs), dtype=np.int64)
self.ep_length = np.zeros((self.buffer_size, self.n_envs), dtype=np.int64)
self._current_ep_start = np.zeros(self.n_envs, dtype=np.int64)
def __getstate__(self) -> Dict[str, Any]:
"""
Gets state for pickling.
Excludes self.env, as in general Env's may not be pickleable.
Note: when using offline sampling, this will also save the offline replay buffer.
"""
state = self.__dict__.copy()
# these attributes are not pickleable
@ -185,347 +139,257 @@ class HerReplayBuffer(DictReplayBuffer):
self.env = env
def _get_samples(self, batch_inds: np.ndarray, env: Optional[VecNormalize] = None) -> DictReplayBufferSamples:
"""
Abstract method from base class.
"""
raise NotImplementedError()
def sample(self, batch_size: int, env: Optional[VecNormalize] = None) -> DictReplayBufferSamples:
"""
Sample function for online sampling of HER transition,
this replaces the "regular" replay buffer ``sample()``
method in the ``train()`` function.
:param batch_size: Number of element to sample
:param env: Associated gym VecEnv
to normalize the observations/rewards when sampling
:return: Samples.
"""
if self.replay_buffer is not None:
return self.replay_buffer.sample(batch_size, env)
return self._sample_transitions(batch_size, maybe_vec_env=env, online_sampling=True) # pytype: disable=bad-return-type
def _sample_offline(
self,
n_sampled_goal: Optional[int] = None,
) -> Tuple[Dict[str, np.ndarray], Dict[str, np.ndarray], np.ndarray, np.ndarray]:
"""
Sample function for offline sampling of HER transition,
in that case, only one episode is used and transitions
are added to the regular replay buffer.
:param n_sampled_goal: Number of sampled goals for replay
:return: at most(n_sampled_goal * episode_length) HER transitions.
"""
# `maybe_vec_env=None` as we should store unnormalized transitions,
# they will be normalized at sampling time
return self._sample_transitions(
batch_size=None,
maybe_vec_env=None,
online_sampling=False,
n_sampled_goal=n_sampled_goal,
) # pytype: disable=bad-return-type
def sample_goals(
self,
episode_indices: np.ndarray,
her_indices: np.ndarray,
transitions_indices: np.ndarray,
) -> np.ndarray:
"""
Sample goals based on goal_selection_strategy.
This is a vectorized (fast) version.
:param episode_indices: Episode indices to use.
:param her_indices: HER indices.
:param transitions_indices: Transition indices to use.
:return: Return sampled goals.
"""
her_episode_indices = episode_indices[her_indices]
if self.goal_selection_strategy == GoalSelectionStrategy.FINAL:
# replay with final state of current episode
transitions_indices = self.episode_lengths[her_episode_indices] - 1
elif self.goal_selection_strategy == GoalSelectionStrategy.FUTURE:
# replay with random state which comes from the same episode and was observed after current transition
transitions_indices = np.random.randint(
transitions_indices[her_indices], self.episode_lengths[her_episode_indices]
)
elif self.goal_selection_strategy == GoalSelectionStrategy.EPISODE:
# replay with random state which comes from the same episode as current transition
transitions_indices = np.random.randint(self.episode_lengths[her_episode_indices])
else:
raise ValueError(f"Strategy {self.goal_selection_strategy} for sampling goals not supported!")
return self._buffer["next_achieved_goal"][her_episode_indices, transitions_indices]
def _sample_transitions(
self,
batch_size: Optional[int],
maybe_vec_env: Optional[VecNormalize],
online_sampling: bool,
n_sampled_goal: Optional[int] = None,
) -> Union[DictReplayBufferSamples, Tuple[Dict[str, np.ndarray], Dict[str, np.ndarray], np.ndarray, np.ndarray]]:
"""
:param batch_size: Number of element to sample (only used for online sampling)
:param env: associated gym VecEnv to normalize the observations/rewards
Only valid when using online sampling
:param online_sampling: Using online_sampling for HER or not.
:param n_sampled_goal: Number of sampled goals for replay. (offline sampling)
:return: Samples.
"""
# Select which episodes to use
if online_sampling:
assert batch_size is not None, "No batch_size specified for online sampling of HER transitions"
# Do not sample the episode with index `self.pos` as the episode is invalid
if self.full:
episode_indices = (
np.random.randint(1, self.n_episodes_stored, batch_size) + self.pos
) % self.n_episodes_stored
else:
episode_indices = np.random.randint(0, self.n_episodes_stored, batch_size)
# A subset of the transitions will be relabeled using HER algorithm
her_indices = np.arange(batch_size)[: int(self.her_ratio * batch_size)]
else:
assert maybe_vec_env is None, "Transitions must be stored unnormalized in the replay buffer"
assert n_sampled_goal is not None, "No n_sampled_goal specified for offline sampling of HER transitions"
# Offline sampling: there is only one episode stored
episode_length = self.episode_lengths[0]
# we sample n_sampled_goal per timestep in the episode (only one is stored).
episode_indices = np.tile(0, (episode_length * n_sampled_goal))
# we only sample virtual transitions
# as real transitions are already stored in the replay buffer
her_indices = np.arange(len(episode_indices))
ep_lengths = self.episode_lengths[episode_indices]
if online_sampling:
# Select which transitions to use
transitions_indices = np.random.randint(ep_lengths)
else:
if her_indices.size == 0:
# Episode of one timestep, not enough for using the "future" strategy
# no virtual transitions are created in that case
return {}, {}, np.zeros(0), np.zeros(0)
else:
# Repeat every transition index n_sampled_goals times
# to sample n_sampled_goal per timestep in the episode (only one is stored).
# Now with the corrected episode length when using "future" strategy
transitions_indices = np.tile(np.arange(ep_lengths[0]), n_sampled_goal)
episode_indices = episode_indices[transitions_indices]
her_indices = np.arange(len(episode_indices))
# get selected transitions
transitions = {key: self._buffer[key][episode_indices, transitions_indices].copy() for key in self._buffer.keys()}
# sample new desired goals and relabel the transitions
new_goals = self.sample_goals(episode_indices, her_indices, transitions_indices)
transitions["desired_goal"][her_indices] = new_goals
# Convert info buffer to numpy array
transitions["info"] = np.array(
[
self.info_buffer[episode_idx][transition_idx]
for episode_idx, transition_idx in zip(episode_indices, transitions_indices)
]
)
# Edge case: episode of one timesteps with the future strategy
# no virtual transition can be created
if len(her_indices) > 0:
# Vectorized computation of the new reward
transitions["reward"][her_indices, 0] = self.env.env_method(
"compute_reward",
# the new state depends on the previous state and action
# s_{t+1} = f(s_t, a_t)
# so the next_achieved_goal depends also on the previous state and action
# because we are in a GoalEnv:
# r_t = reward(s_t, a_t) = reward(next_achieved_goal, desired_goal)
# therefore we have to use "next_achieved_goal" and not "achieved_goal"
transitions["next_achieved_goal"][her_indices, 0],
# here we use the new desired goal
transitions["desired_goal"][her_indices, 0],
transitions["info"][her_indices, 0],
)
# concatenate observation with (desired) goal
observations = self._normalize_obs(transitions, maybe_vec_env)
# HACK to make normalize obs and `add()` work with the next observation
next_observations = {
"observation": transitions["next_obs"],
"achieved_goal": transitions["next_achieved_goal"],
# The desired goal for the next observation must be the same as the previous one
"desired_goal": transitions["desired_goal"],
}
next_observations = self._normalize_obs(next_observations, maybe_vec_env)
if online_sampling:
next_obs = {key: self.to_torch(next_observations[key][:, 0, :]) for key in self._observation_keys}
normalized_obs = {key: self.to_torch(observations[key][:, 0, :]) for key in self._observation_keys}
return DictReplayBufferSamples(
observations=normalized_obs,
actions=self.to_torch(transitions["action"]),
next_observations=next_obs,
dones=self.to_torch(transitions["done"]),
rewards=self.to_torch(self._normalize_reward(transitions["reward"], maybe_vec_env)),
)
else:
return observations, next_observations, transitions["action"], transitions["reward"]
def add(
self,
obs: Dict[str, np.ndarray],
next_obs: Dict[str, np.ndarray],
obs: TensorDict,
next_obs: TensorDict,
action: np.ndarray,
reward: np.ndarray,
done: np.ndarray,
infos: List[Dict[str, Any]],
) -> None:
if self.current_idx == 0 and self.full:
# Clear info buffer
self.info_buffer[self.pos] = deque(maxlen=self.max_episode_length)
# When the buffer is full, we rewrite on old episodes. When we start to
# rewrite on an old episodes, we want the whole old episode to be deleted
# (and not only the transition on which we rewrite). To do this, we set
# the length of the old episode to 0, so it can't be sampled anymore.
for env_idx in range(self.n_envs):
episode_start = self.ep_start[self.pos, env_idx]
episode_length = self.ep_length[self.pos, env_idx]
if episode_length > 0:
episode_end = episode_start + episode_length
episode_indices = np.arange(self.pos, episode_end) % self.buffer_size
self.ep_length[episode_indices, env_idx] = 0
# Remove termination signals due to timeout
if self.handle_timeout_termination:
done_ = done * (1 - np.array([info.get("TimeLimit.truncated", False) for info in infos]))
else:
done_ = done
# Update episode start
self.ep_start[self.pos] = self._current_ep_start.copy()
self._buffer["observation"][self.pos][self.current_idx] = obs["observation"]
self._buffer["achieved_goal"][self.pos][self.current_idx] = obs["achieved_goal"]
self._buffer["desired_goal"][self.pos][self.current_idx] = obs["desired_goal"]
self._buffer["action"][self.pos][self.current_idx] = action
self._buffer["done"][self.pos][self.current_idx] = done_
self._buffer["reward"][self.pos][self.current_idx] = reward
self._buffer["next_obs"][self.pos][self.current_idx] = next_obs["observation"]
self._buffer["next_achieved_goal"][self.pos][self.current_idx] = next_obs["achieved_goal"]
self._buffer["next_desired_goal"][self.pos][self.current_idx] = next_obs["desired_goal"]
if self.copy_info_dict:
self.infos[self.pos] = infos
# Store the transition
super().add(obs, next_obs, action, reward, done, infos)
# When doing offline sampling
# Add real transition to normal replay buffer
if self.replay_buffer is not None:
self.replay_buffer.add(
obs,
next_obs,
action,
reward,
done,
infos,
# When episode ends, compute and store the episode length
for env_idx in range(self.n_envs):
if done[env_idx]:
episode_start = self._current_ep_start[env_idx]
episode_end = self.pos
if episode_end < episode_start:
# Occurs when the buffer becomes full, the storage resumes at the
# beginning of the buffer. This can happen in the middle of an episode.
episode_end += self.buffer_size
episode_indices = np.arange(episode_start, episode_end) % self.buffer_size
self.ep_length[episode_indices, env_idx] = episode_end - episode_start
# Update the current episode start
self._current_ep_start[env_idx] = self.pos
def sample(self, batch_size: int, env: Optional[VecNormalize] = None) -> DictReplayBufferSamples:
"""
Sample elements from the replay buffer.
:param batch_size: Number of element to sample
:param env: Associated VecEnv to normalize the observations/rewards when sampling
:return: Samples
"""
# When the buffer is full, we rewrite on old episodes. We don't want to
# sample incomplete episode transitions, so we have to eliminate some indexes.
is_valid = self.ep_length > 0
if not np.any(is_valid):
raise RuntimeError(
"Unable to sample before the end of the first episode. We recommend choosing a value "
"for learning_starts that is greater than the maximum number of timesteps in the environment."
)
# Get the indices of valid transitions
# Example:
# if is_valid = [[True, False, False], [True, False, True]],
# is_valid has shape (buffer_size=2, n_envs=3)
# then valid_indices = [0, 3, 5]
# they correspond to is_valid[0, 0], is_valid[1, 0] and is_valid[1, 2]
# or in numpy format ([rows], [columns]): (array([0, 1, 1]), array([0, 0, 2]))
# Those indices are obtained back using np.unravel_index(valid_indices, is_valid.shape)
valid_indices = np.flatnonzero(is_valid)
# Sample valid transitions that will constitute the minibatch of size batch_size
sampled_indices = np.random.choice(valid_indices, size=batch_size, replace=True)
# Unravel the indexes, i.e. recover the batch and env indices.
# Example: if sampled_indices = [0, 3, 5], then batch_indices = [0, 1, 1] and env_indices = [0, 0, 2]
batch_indices, env_indices = np.unravel_index(sampled_indices, is_valid.shape)
self.info_buffer[self.pos].append(infos)
# Split the indexes between real and virtual transitions.
nb_virtual = int(self.her_ratio * batch_size)
virtual_batch_indices, real_batch_indices = np.split(batch_indices, [nb_virtual])
virtual_env_indices, real_env_indices = np.split(env_indices, [nb_virtual])
# update current pointer
self.current_idx += 1
# Get real and virtual data
real_data = self._get_real_samples(real_batch_indices, real_env_indices, env)
# Create virtual transitions by sampling new desired goals and computing new rewards
virtual_data = self._get_virtual_samples(virtual_batch_indices, virtual_env_indices, env)
self.episode_steps += 1
# Concatenate real and virtual data
observations = {
key: th.cat((real_data.observations[key], virtual_data.observations[key]))
for key in virtual_data.observations.keys()
}
actions = th.cat((real_data.actions, virtual_data.actions))
next_observations = {
key: th.cat((real_data.next_observations[key], virtual_data.next_observations[key]))
for key in virtual_data.next_observations.keys()
}
dones = th.cat((real_data.dones, virtual_data.dones))
rewards = th.cat((real_data.rewards, virtual_data.rewards))
if done or self.episode_steps >= self.max_episode_length:
self.store_episode()
if not self.online_sampling:
# sample virtual transitions and store them in replay buffer
self._sample_her_transitions()
# clear storage for current episode
self.reset()
return DictReplayBufferSamples(
observations=observations,
actions=actions,
next_observations=next_observations,
dones=dones,
rewards=rewards,
)
self.episode_steps = 0
def store_episode(self) -> None:
def _get_real_samples(
self,
batch_indices: np.ndarray,
env_indices: np.ndarray,
env: Optional[VecNormalize] = None,
) -> DictReplayBufferSamples:
"""
Increment episode counter
and reset transition pointer.
"""
# add episode length to length storage
self.episode_lengths[self.pos] = self.current_idx
Get the samples corresponding to the batch and environment indices.
# update current episode pointer
# Note: in the OpenAI implementation
# when the buffer is full, the episode replaced
# is randomly chosen
self.pos += 1
if self.pos == self.max_episode_stored:
self.full = True
self.pos = 0
# reset transition pointer
self.current_idx = 0
def _sample_her_transitions(self) -> None:
"""
Sample additional goals and store new transitions in replay buffer
when using offline sampling.
:param batch_indices: Indices of the transitions
:param env_indices: Indices of the envrionments
:param env: associated gym VecEnv to normalize the
observations/rewards when sampling, defaults to None
:return: Samples
"""
# Normalize if needed and remove extra dimension (we are using only one env for now)
obs_ = self._normalize_obs({key: obs[batch_indices, env_indices, :] for key, obs in self.observations.items()}, env)
next_obs_ = self._normalize_obs(
{key: obs[batch_indices, env_indices, :] for key, obs in self.next_observations.items()}, env
)
# Sample goals to create virtual transitions for the last episode.
observations, next_observations, actions, rewards = self._sample_offline(n_sampled_goal=self.n_sampled_goal)
# Convert to torch tensor
observations = {key: self.to_torch(obs) for key, obs in obs_.items()}
next_observations = {key: self.to_torch(obs) for key, obs in next_obs_.items()}
# Store virtual transitions in the replay buffer, if available
if len(observations) > 0:
for i in range(len(observations["observation"])):
self.replay_buffer.add(
{key: obs[i] for key, obs in observations.items()},
{key: next_obs[i] for key, next_obs in next_observations.items()},
actions[i],
rewards[i],
# We consider the transition as non-terminal
done=[False],
infos=[{}],
)
return DictReplayBufferSamples(
observations=observations,
actions=self.to_torch(self.actions[batch_indices, env_indices]),
next_observations=next_observations,
# Only use dones that are not due to timeouts
# deactivated by default (timeouts is initialized as an array of False)
dones=self.to_torch(
self.dones[batch_indices, env_indices] * (1 - self.timeouts[batch_indices, env_indices])
).reshape(-1, 1),
rewards=self.to_torch(self._normalize_reward(self.rewards[batch_indices, env_indices].reshape(-1, 1), env)),
)
@property
def n_episodes_stored(self) -> int:
if self.full:
return self.max_episode_stored
return self.pos
def _get_virtual_samples(
self,
batch_indices: np.ndarray,
env_indices: np.ndarray,
env: Optional[VecNormalize] = None,
) -> DictReplayBufferSamples:
"""
Get the samples, sample new desired goals and compute new rewards.
def size(self) -> int:
:param batch_indices: Indices of the transitions
:param env_indices: Indices of the envrionments
:param env: associated gym VecEnv to normalize the
observations/rewards when sampling, defaults to None
:return: Samples, with new desired goals and new rewards
"""
:return: The current number of transitions in the buffer.
"""
return int(np.sum(self.episode_lengths))
# Get infos and obs
obs = {key: obs[batch_indices, env_indices, :] for key, obs in self.observations.items()}
next_obs = {key: obs[batch_indices, env_indices, :] for key, obs in self.next_observations.items()}
if self.copy_info_dict:
# The copy may cause a slow down
infos = copy.deepcopy(self.infos[batch_indices, env_indices])
else:
infos = [{} for _ in range(len(batch_indices))]
# Sample and set new goals
new_goals = self._sample_goals(batch_indices, env_indices)
obs["desired_goal"] = new_goals
# The desired goal for the next observation must be the same as the previous one
next_obs["desired_goal"] = new_goals
def reset(self) -> None:
# Compute new reward
rewards = self.env.env_method(
"compute_reward",
# here we use the new desired goal
obs["desired_goal"],
# the new state depends on the previous state and action
# s_{t+1} = f(s_t, a_t)
# so the next achieved_goal depends also on the previous state and action
# because we are in a GoalEnv:
# r_t = reward(s_t, a_t) = reward(next_achieved_goal, desired_goal)
# therefore we have to use next_obs["achieved_goal"] and not obs["achieved_goal"]
next_obs["achieved_goal"],
infos,
# we use the method of the first environment assuming that all environments are identical.
indices=[0],
)
rewards = rewards[0].astype(np.float32) # env_method returns a list containing one element
obs = self._normalize_obs(obs, env)
next_obs = self._normalize_obs(next_obs, env)
# Convert to torch tensor
observations = {key: self.to_torch(obs) for key, obs in obs.items()}
next_observations = {key: self.to_torch(obs) for key, obs in next_obs.items()}
return DictReplayBufferSamples(
observations=observations,
actions=self.to_torch(self.actions[batch_indices, env_indices]),
next_observations=next_observations,
# Only use dones that are not due to timeouts
# deactivated by default (timeouts is initialized as an array of False)
dones=self.to_torch(
self.dones[batch_indices, env_indices] * (1 - self.timeouts[batch_indices, env_indices])
).reshape(-1, 1),
rewards=self.to_torch(self._normalize_reward(rewards.reshape(-1, 1), env)),
)
def _sample_goals(self, batch_indices: np.ndarray, env_indices: np.ndarray) -> np.ndarray:
"""
Reset the buffer.
Sample goals based on goal_selection_strategy.
:param batch_indices: Indices of the transitions
:param env_indices: Indices of the envrionments
:return: Sampled goals
"""
self.pos = 0
self.current_idx = 0
self.full = False
self.episode_lengths = np.zeros(self.max_episode_stored, dtype=np.int64)
batch_ep_start = self.ep_start[batch_indices, env_indices]
batch_ep_length = self.ep_length[batch_indices, env_indices]
if self.goal_selection_strategy == GoalSelectionStrategy.FINAL:
# Replay with final state of current episode
transition_indices_in_episode = batch_ep_length - 1
elif self.goal_selection_strategy == GoalSelectionStrategy.FUTURE:
# Replay with random state which comes from the same episode and was observed after current transition
# Note: our implementation is inclusive: current transition can be sampled
current_indices_in_episode = (batch_indices - batch_ep_start) % self.buffer_size
transition_indices_in_episode = np.random.randint(current_indices_in_episode, batch_ep_length)
elif self.goal_selection_strategy == GoalSelectionStrategy.EPISODE:
# Replay with random state which comes from the same episode as current transition
transition_indices_in_episode = np.random.randint(0, batch_ep_length)
else:
raise ValueError(f"Strategy {self.goal_selection_strategy} for sampling goals not supported!")
transition_indices = (transition_indices_in_episode + batch_ep_start) % self.buffer_size
return self.next_observations["achieved_goal"][transition_indices, env_indices]
def truncate_last_trajectory(self) -> None:
"""
Only for online sampling, called when loading the replay buffer.
If called, we assume that the last trajectory in the replay buffer was finished
(and truncate it).
If not called, we assume that we continue the same trajectory (same episode).
"""
# If we are at the start of an episode, no need to truncate
current_idx = self.current_idx
# truncate interrupted episode
if current_idx > 0:
if (self.ep_start[self.pos] != self.pos).any():
warnings.warn(
"The last trajectory in the replay buffer will be truncated.\n"
"If you are in the same episode as when the replay buffer was saved,\n"
"you should use `truncate_last_trajectory=False` to avoid that issue."
)
# get current episode and transition index
pos = self.pos
# set episode length for current episode
self.episode_lengths[pos] = current_idx
# set done = True for current episode
# current_idx was already incremented
self._buffer["done"][pos][current_idx - 1] = np.array([True], dtype=np.float32)
# reset current transition index
self.current_idx = 0
# increment episode counter
self.pos = (self.pos + 1) % self.max_episode_stored
# update "full" indicator
self.full = self.full or self.pos == 0
self.ep_start[-1] = self.pos
# set done = True for current episodes
self.dones[self.pos - 1] = True

2
stable_baselines3/version.txt
View file

@ -1 +1 @@
1.8.0a9
1.8.0a10

1
tests/test_callbacks.py
View file

@ -140,7 +140,6 @@ def test_eval_success_logging(tmp_path):
replay_buffer_class=HerReplayBuffer,
learning_starts=100,
seed=0,
replay_buffer_kwargs=dict(max_episode_length=n_bits),
)
model.learn(500, callback=eval_callback)
assert len(eval_callback._is_success_buffer) > 0

200
tests/test_her.py
View file

@ -3,19 +3,18 @@ import pathlib
import warnings
from copy import deepcopy
import gym
import numpy as np
import pytest
import torch as th
from stable_baselines3 import DDPG, DQN, SAC, TD3, HerReplayBuffer
from stable_baselines3.common.env_util import make_vec_env
from stable_baselines3.common.envs import BitFlippingEnv
from stable_baselines3.common.evaluation import evaluate_policy
from stable_baselines3.common.monitor import Monitor
from stable_baselines3.common.noise import NormalActionNoise
from stable_baselines3.common.vec_env import DummyVecEnv
from stable_baselines3.common.vec_env import SubprocVecEnv
from stable_baselines3.her.goal_selection_strategy import GoalSelectionStrategy
from stable_baselines3.her.her_replay_buffer import get_time_limit
def test_import_error():
@ -27,18 +26,22 @@ def test_import_error():
@pytest.mark.parametrize("model_class", [SAC, TD3, DDPG, DQN])
@pytest.mark.parametrize("online_sampling", [True, False])
@pytest.mark.parametrize("image_obs_space", [True, False])
def test_her(model_class, online_sampling, image_obs_space):
def test_her(model_class, image_obs_space):
"""
Test Hindsight Experience Replay.
"""
n_envs = 1
n_bits = 4
env = BitFlippingEnv(
n_bits=n_bits,
continuous=not (model_class == DQN),
image_obs_space=image_obs_space,
)
def env_fn():
return BitFlippingEnv(
n_bits=n_bits,
continuous=not (model_class == DQN),
image_obs_space=image_obs_space,
)
env = make_vec_env(env_fn, n_envs)
model = model_class(
"MultiInputPolicy",
@ -47,18 +50,28 @@ def test_her(model_class, online_sampling, image_obs_space):
replay_buffer_kwargs=dict(
n_sampled_goal=2,
goal_selection_strategy="future",
online_sampling=online_sampling,
max_episode_length=n_bits,
copy_info_dict=True,
),
train_freq=4,
gradient_steps=1,
gradient_steps=n_envs,
policy_kwargs=dict(net_arch=[64]),
learning_starts=100,
buffer_size=int(2e4),
)
model.learn(total_timesteps=150)
evaluate_policy(model, Monitor(env))
evaluate_policy(model, Monitor(env_fn()))
@pytest.mark.parametrize("model_class", [TD3, DQN])
@pytest.mark.parametrize("image_obs_space", [True, False])
def test_multiprocessing(model_class, image_obs_space):
def env_fn():
return BitFlippingEnv(n_bits=4, continuous=not (model_class == DQN), image_obs_space=image_obs_space)
env = make_vec_env(env_fn, n_envs=2, vec_env_cls=SubprocVecEnv)
model = model_class("MultiInputPolicy", env, replay_buffer_class=HerReplayBuffer, buffer_size=int(2e4), train_freq=4)
model.learn(total_timesteps=150)
@pytest.mark.parametrize(
@ -72,12 +85,16 @@ def test_her(model_class, online_sampling, image_obs_space):
GoalSelectionStrategy.FUTURE,
],
)
@pytest.mark.parametrize("online_sampling", [True, False])
def test_goal_selection_strategy(goal_selection_strategy, online_sampling):
def test_goal_selection_strategy(goal_selection_strategy):
"""
Test different goal strategies.
"""
env = BitFlippingEnv(continuous=True)
n_envs = 2
def env_fn():
return BitFlippingEnv(continuous=True)
env = make_vec_env(env_fn, n_envs)
normal_action_noise = NormalActionNoise(np.zeros(1), 0.1 * np.ones(1))
@ -87,12 +104,10 @@ def test_goal_selection_strategy(goal_selection_strategy, online_sampling):
replay_buffer_class=HerReplayBuffer,
replay_buffer_kwargs=dict(
goal_selection_strategy=goal_selection_strategy,
online_sampling=online_sampling,
max_episode_length=10,
n_sampled_goal=2,
),
train_freq=4,
gradient_steps=1,
gradient_steps=n_envs,
policy_kwargs=dict(net_arch=[64]),
learning_starts=100,
buffer_size=int(1e5),
@ -104,16 +119,20 @@ def test_goal_selection_strategy(goal_selection_strategy, online_sampling):
@pytest.mark.parametrize("model_class", [SAC, TD3, DDPG, DQN])
@pytest.mark.parametrize("use_sde", [False, True])
@pytest.mark.parametrize("online_sampling", [False, True])
def test_save_load(tmp_path, model_class, use_sde, online_sampling):
def test_save_load(tmp_path, model_class, use_sde):
"""
Test if 'save' and 'load' saves and loads model correctly
"""
if use_sde and model_class != SAC:
pytest.skip("Only SAC has gSDE support")
n_envs = 2
n_bits = 4
env = BitFlippingEnv(n_bits=n_bits, continuous=not (model_class == DQN))
def env_fn():
return BitFlippingEnv(n_bits=n_bits, continuous=not (model_class == DQN))
env = make_vec_env(env_fn, n_envs)
kwargs = dict(use_sde=True) if use_sde else {}
@ -125,8 +144,6 @@ def test_save_load(tmp_path, model_class, use_sde, online_sampling):
replay_buffer_kwargs=dict(
n_sampled_goal=2,
goal_selection_strategy="future",
online_sampling=online_sampling,
max_episode_length=n_bits,
),
verbose=0,
tau=0.05,
@ -135,7 +152,7 @@ def test_save_load(tmp_path, model_class, use_sde, online_sampling):
policy_kwargs=dict(net_arch=[64]),
buffer_size=int(1e5),
gamma=0.98,
gradient_steps=1,
gradient_steps=n_envs,
train_freq=4,
learning_starts=100,
**kwargs
@ -143,14 +160,9 @@ def test_save_load(tmp_path, model_class, use_sde, online_sampling):
model.learn(total_timesteps=150)
obs = env.reset()
observations = {key: [] for key in obs.keys()}
for _ in range(10):
obs = env.step(env.action_space.sample())[0]
for key in obs.keys():
observations[key].append(obs[key])
observations = {key: np.array(obs) for key, obs in observations.items()}
env.reset()
action = np.array([env.action_space.sample() for _ in range(n_envs)])
observations = env.step(action)[0]
# Get dictionary of current parameters
params = deepcopy(model.policy.state_dict())
@ -210,9 +222,9 @@ def test_save_load(tmp_path, model_class, use_sde, online_sampling):
os.remove(tmp_path / "test_save.zip")
@pytest.mark.parametrize("online_sampling", [False, True])
@pytest.mark.parametrize("n_envs", [1, 2])
@pytest.mark.parametrize("truncate_last_trajectory", [False, True])
def test_save_load_replay_buffer(tmp_path, recwarn, online_sampling, truncate_last_trajectory):
def test_save_load_replay_buffer(n_envs, tmp_path, recwarn, truncate_last_trajectory):
"""
Test if 'save_replay_buffer' and 'load_replay_buffer' works correctly
"""
@ -222,7 +234,11 @@ def test_save_load_replay_buffer(tmp_path, recwarn, online_sampling, truncate_la
path = pathlib.Path(tmp_path / "replay_buffer.pkl")
path.parent.mkdir(exist_ok=True, parents=True) # to not raise a warning
env = BitFlippingEnv(n_bits=4, continuous=True)
def env_fn():
return BitFlippingEnv(n_bits=4, continuous=True)
env = make_vec_env(env_fn, n_envs)
model = SAC(
"MultiInputPolicy",
env,
@ -230,20 +246,16 @@ def test_save_load_replay_buffer(tmp_path, recwarn, online_sampling, truncate_la
replay_buffer_kwargs=dict(
n_sampled_goal=2,
goal_selection_strategy="future",
online_sampling=online_sampling,
max_episode_length=4,
),
gradient_steps=1,
gradient_steps=n_envs,
train_freq=4,
buffer_size=int(2e4),
policy_kwargs=dict(net_arch=[64]),
seed=1,
)
model.learn(200)
if online_sampling:
old_replay_buffer = deepcopy(model.replay_buffer)
else:
old_replay_buffer = deepcopy(model.replay_buffer.replay_buffer)
old_replay_buffer = deepcopy(model.replay_buffer)
model.save_replay_buffer(path)
del model.replay_buffer
@ -262,36 +274,15 @@ def test_save_load_replay_buffer(tmp_path, recwarn, online_sampling, truncate_la
else:
assert len(recwarn) == 0
if online_sampling:
n_episodes_stored = model.replay_buffer.n_episodes_stored
assert np.allclose(
old_replay_buffer._buffer["observation"][:n_episodes_stored],
model.replay_buffer._buffer["observation"][:n_episodes_stored],
)
assert np.allclose(
old_replay_buffer._buffer["next_obs"][:n_episodes_stored],
model.replay_buffer._buffer["next_obs"][:n_episodes_stored],
)
assert np.allclose(
old_replay_buffer._buffer["action"][:n_episodes_stored],
model.replay_buffer._buffer["action"][:n_episodes_stored],
)
assert np.allclose(
old_replay_buffer._buffer["reward"][:n_episodes_stored],
model.replay_buffer._buffer["reward"][:n_episodes_stored],
)
# we might change the last done of the last trajectory so we don't compare it
assert np.allclose(
old_replay_buffer._buffer["done"][: n_episodes_stored - 1],
model.replay_buffer._buffer["done"][: n_episodes_stored - 1],
)
else:
replay_buffer = model.replay_buffer.replay_buffer
assert np.allclose(old_replay_buffer.observations["observation"], replay_buffer.observations["observation"])
assert np.allclose(old_replay_buffer.observations["desired_goal"], replay_buffer.observations["desired_goal"])
assert np.allclose(old_replay_buffer.actions, replay_buffer.actions)
assert np.allclose(old_replay_buffer.rewards, replay_buffer.rewards)
assert np.allclose(old_replay_buffer.dones, replay_buffer.dones)
replay_buffer = model.replay_buffer
pos = replay_buffer.pos
for key in ["observation", "desired_goal", "achieved_goal"]:
assert np.allclose(old_replay_buffer.observations[key][:pos], replay_buffer.observations[key][:pos])
assert np.allclose(old_replay_buffer.next_observations[key][:pos], replay_buffer.next_observations[key][:pos])
assert np.allclose(old_replay_buffer.actions[:pos], replay_buffer.actions[:pos])
assert np.allclose(old_replay_buffer.rewards[:pos], replay_buffer.rewards[:pos])
# we might change the last done of the last trajectory so we don't compare it
assert np.allclose(old_replay_buffer.dones[: pos - 1], replay_buffer.dones[: pos - 1])
# test if continuing training works properly
reset_num_timesteps = False if truncate_last_trajectory is False else True
@ -304,7 +295,12 @@ def test_full_replay_buffer():
It should not sample the current episode which is not finished.
"""
n_bits = 4
env = BitFlippingEnv(n_bits=n_bits, continuous=True)
n_envs = 2
def env_fn():
return BitFlippingEnv(n_bits=n_bits, continuous=True)
env = make_vec_env(env_fn, n_envs)
# use small buffer size to get the buffer full
model = SAC(
@ -314,14 +310,12 @@ def test_full_replay_buffer():
replay_buffer_kwargs=dict(
n_sampled_goal=2,
goal_selection_strategy="future",
online_sampling=True,
max_episode_length=n_bits,
),
gradient_steps=1,
train_freq=4,
policy_kwargs=dict(net_arch=[64]),
learning_starts=1,
buffer_size=20,
learning_starts=n_bits * n_envs,
buffer_size=20 * n_envs,
verbose=1,
seed=757,
)
@ -329,49 +323,18 @@ def test_full_replay_buffer():
model.learn(total_timesteps=100)
def test_get_max_episode_length():
dict_env = DummyVecEnv([lambda: BitFlippingEnv()])
# Cannot infer max epsiode length
with pytest.raises(ValueError):
get_time_limit(dict_env, current_max_episode_length=None)
default_length = 10
assert get_time_limit(dict_env, current_max_episode_length=default_length) == default_length
env = gym.make("CartPole-v1")
vec_env = DummyVecEnv([lambda: env])
assert get_time_limit(vec_env, current_max_episode_length=None) == 500
# Overwrite max_episode_steps
assert get_time_limit(vec_env, current_max_episode_length=default_length) == default_length
# Set max_episode_steps to None
env.spec.max_episode_steps = None
vec_env = DummyVecEnv([lambda: env])
with pytest.raises(ValueError):
get_time_limit(vec_env, current_max_episode_length=None)
# Initialize HER and specify max_episode_length, should not raise an issue
DQN("MultiInputPolicy", dict_env, replay_buffer_class=HerReplayBuffer, replay_buffer_kwargs=dict(max_episode_length=5))
with pytest.raises(ValueError):
DQN("MultiInputPolicy", dict_env, replay_buffer_class=HerReplayBuffer)
# Wrapped in a timelimit, should be fine
# Note: it requires env.spec to be defined
env = DummyVecEnv([lambda: gym.wrappers.TimeLimit(BitFlippingEnv(), 10)])
DQN("MultiInputPolicy", env, replay_buffer_class=HerReplayBuffer, replay_buffer_kwargs=dict(max_episode_length=5))
@pytest.mark.parametrize("online_sampling", [False, True])
@pytest.mark.parametrize("n_bits", [10])
def test_performance_her(online_sampling, n_bits):
def test_performance_her(n_bits):
"""
That DQN+HER can solve BitFlippingEnv.
It should not work when n_sampled_goal=0 (DQN alone).
"""
env = BitFlippingEnv(n_bits=n_bits, continuous=False)
n_envs = 2
def env_fn():
return BitFlippingEnv(n_bits=n_bits, continuous=False)
env = make_vec_env(env_fn, n_envs)
model = DQN(
"MultiInputPolicy",
@ -380,12 +343,11 @@ def test_performance_her(online_sampling, n_bits):
replay_buffer_kwargs=dict(
n_sampled_goal=5,
goal_selection_strategy="future",
online_sampling=online_sampling,
max_episode_length=n_bits,
),
verbose=1,
learning_rate=5e-4,
train_freq=1,
gradient_steps=n_envs,
learning_starts=100,
exploration_final_eps=0.02,
target_update_interval=500,

60
tests/test_vec_normalize.py
View file

@ -338,36 +338,44 @@ def test_normalize_dict_selected_keys():
np.testing.assert_array_equal(obs["achieved_goal"], orig_obs["achieved_goal"])
@pytest.mark.parametrize("model_class", [SAC, TD3, HerReplayBuffer])
@pytest.mark.parametrize("online_sampling", [False, True])
def test_offpolicy_normalization(model_class, online_sampling):
if online_sampling and model_class != HerReplayBuffer:
pytest.skip()
make_env_ = make_dict_env if model_class == HerReplayBuffer else make_env
env = DummyVecEnv([make_env_])
def test_her_normalization():
env = DummyVecEnv([make_dict_env])
env = VecNormalize(env, norm_obs=True, norm_reward=True, clip_obs=10.0, clip_reward=10.0)
eval_env = DummyVecEnv([make_env_])
eval_env = DummyVecEnv([make_dict_env])
eval_env = VecNormalize(eval_env, training=False, norm_obs=True, norm_reward=False, clip_obs=10.0, clip_reward=10.0)
if model_class == HerReplayBuffer:
model = SAC(
"MultiInputPolicy",
env,
verbose=1,
learning_starts=100,
policy_kwargs=dict(net_arch=[64]),
replay_buffer_kwargs=dict(
max_episode_length=100,
online_sampling=online_sampling,
n_sampled_goal=2,
),
replay_buffer_class=HerReplayBuffer,
seed=2,
)
else:
model = model_class("MlpPolicy", env, verbose=1, learning_starts=100, policy_kwargs=dict(net_arch=[64]))
model = SAC(
"MultiInputPolicy",
env,
verbose=1,
learning_starts=100,
policy_kwargs=dict(net_arch=[64]),
replay_buffer_kwargs=dict(n_sampled_goal=2),
replay_buffer_class=HerReplayBuffer,
seed=2,
)
# Check that VecNormalize object is correctly updated
assert model.get_vec_normalize_env() is env
model.set_env(eval_env)
assert model.get_vec_normalize_env() is eval_env
model.learn(total_timesteps=10)
model.set_env(env)
model.learn(total_timesteps=150)
# Check getter
assert isinstance(model.get_vec_normalize_env(), VecNormalize)
@pytest.mark.parametrize("model_class", [SAC, TD3])
def test_offpolicy_normalization(model_class):
env = DummyVecEnv([make_env])
env = VecNormalize(env, norm_obs=True, norm_reward=True, clip_obs=10.0, clip_reward=10.0)
eval_env = DummyVecEnv([make_env])
eval_env = VecNormalize(eval_env, training=False, norm_obs=True, norm_reward=False, clip_obs=10.0, clip_reward=10.0)
model = model_class("MlpPolicy", env, verbose=1, learning_starts=100, policy_kwargs=dict(net_arch=[64]))
# Check that VecNormalize object is correctly updated
assert model.get_vec_normalize_env() is env