Rename to stable-baselines3

This commit is contained in:
Antonin RAFFIN 2020-05-05 15:02:35 +02:00
parent 4a2c247592
commit d542732c8d
72 changed files with 164 additions and 164 deletions

View file

@ -4,7 +4,7 @@ omit =
tests/* tests/*
setup.py setup.py
# Require graphical interface # Require graphical interface
torchy_baselines/common/results_plotter.py stable_baselines3/common/results_plotter.py
[report] [report]
exclude_lines = exclude_lines =

View file

@ -13,7 +13,7 @@ If you are submitting a bug report, please fill in the following details.
If your issue is related to a custom gym environment, please check it first using: If your issue is related to a custom gym environment, please check it first using:
```python ```python
from torchy_baselines.common.env_checker import check_env from stable_baselines3.common.env_checker import check_env
env = CustomEnv(arg1, ...) env = CustomEnv(arg1, ...)
# It will check your custom environment and output additional warnings if needed # It will check your custom environment and output additional warnings if needed
@ -30,7 +30,7 @@ Please use the [markdown code blocks](https://help.github.com/en/articles/creati
for both code and stack traces. for both code and stack traces.
```python ```python
from torchy_baselines import ... from stable_baselines3 import ...
``` ```

2
NOTICE
View file

@ -1,4 +1,4 @@
Large portion of the code of Torchy-Baselines (in `common/`) were ported from Stable-Baselines, a fork of OpenAI Baselines, Large portion of the code of Stable-Baselines3 (in `common/`) were ported from Stable-Baselines, a fork of OpenAI Baselines,
both licensed under the MIT License: both licensed under the MIT License:
before the fork (June 2018): before the fork (June 2018):

View file

@ -2,7 +2,7 @@
[![Build Status](https://travis-ci.com/hill-a/stable-baselines.svg?branch=master)](https://travis-ci.com/hill-a/stable-baselines) [![Documentation Status](https://readthedocs.org/projects/stable-baselines/badge/?version=master)](https://stable-baselines.readthedocs.io/en/master/?badge=master) [![Build Status](https://travis-ci.com/hill-a/stable-baselines.svg?branch=master)](https://travis-ci.com/hill-a/stable-baselines) [![Documentation Status](https://readthedocs.org/projects/stable-baselines/badge/?version=master)](https://stable-baselines.readthedocs.io/en/master/?badge=master)
# Torchy Baselines # Stable Baselines3
PyTorch version of [Stable Baselines](https://github.com/hill-a/stable-baselines), a set of improved implementations of reinforcement learning algorithms. PyTorch version of [Stable Baselines](https://github.com/hill-a/stable-baselines), a set of improved implementations of reinforcement learning algorithms.
@ -58,7 +58,7 @@ To cite this repository in publications:
``` ```
@misc{torchy-baselines, @misc{torchy-baselines,
author = {Raffin, Antonin and Hill, Ashley and Ernestus, Maximilian and Gleave, Adam and Kanervisto, Anssi and Dormann, Noah}, author = {Raffin, Antonin and Hill, Ashley and Ernestus, Maximilian and Gleave, Adam and Kanervisto, Anssi and Dormann, Noah},
title = {Torchy Baselines}, title = {Stable Baselines3},
year = {2019}, year = {2019},
publisher = {GitHub}, publisher = {GitHub},
journal = {GitHub repository}, journal = {GitHub repository},

View file

@ -44,19 +44,19 @@ MOCK_MODULES = []
sys.modules.update((mod_name, Mock()) for mod_name in MOCK_MODULES) sys.modules.update((mod_name, Mock()) for mod_name in MOCK_MODULES)
import torchy_baselines import stable_baselines3
# -- Project information ----------------------------------------------------- # -- Project information -----------------------------------------------------
project = 'Torchy Baselines' project = 'Stable Baselines3'
copyright = '2020, Torchy Baselines' copyright = '2020, Stable Baselines3'
author = 'Torchy Baselines Contributors' author = 'Stable Baselines3 Contributors'
# The short X.Y version # The short X.Y version
version = 'master (' + torchy_baselines.__version__ + ' )' version = 'master (' + stable_baselines3.__version__ + ' )'
# The full version, including alpha/beta/rc tags # The full version, including alpha/beta/rc tags
release = torchy_baselines.__version__ release = stable_baselines3.__version__
# -- General configuration --------------------------------------------------- # -- General configuration ---------------------------------------------------
@ -179,8 +179,8 @@ latex_elements = {
# (source start file, target name, title, # (source start file, target name, title,
# author, documentclass [howto, manual, or own class]). # author, documentclass [howto, manual, or own class]).
latex_documents = [ latex_documents = [
(master_doc, 'TorchyBaselines.tex', 'Torchy Baselines Documentation', (master_doc, 'TorchyBaselines.tex', 'Stable Baselines3 Documentation',
'Torchy Baselines Contributors', 'manual'), 'Stable Baselines3 Contributors', 'manual'),
] ]
@ -189,7 +189,7 @@ latex_documents = [
# One entry per manual page. List of tuples # One entry per manual page. List of tuples
# (source start file, name, description, authors, manual section). # (source start file, name, description, authors, manual section).
man_pages = [ man_pages = [
(master_doc, 'torchybaselines', 'Torchy Baselines Documentation', (master_doc, 'torchybaselines', 'Stable Baselines3 Documentation',
[author], 1) [author], 1)
] ]
@ -200,7 +200,7 @@ man_pages = [
# (source start file, target name, title, author, # (source start file, target name, title, author,
# dir menu entry, description, category) # dir menu entry, description, category)
texinfo_documents = [ texinfo_documents = [
(master_doc, 'TorchyBaselines', 'Torchy Baselines Documentation', (master_doc, 'TorchyBaselines', 'Stable Baselines3 Documentation',
author, 'TorchyBaselines', 'One line description of project.', author, 'TorchyBaselines', 'One line description of project.',
'Miscellaneous'), 'Miscellaneous'),
] ]

View file

@ -12,9 +12,9 @@ Here is a quick example of how to train and run SAC on a Pendulum environment:
import gym import gym
from torchy_baselines.sac.policies import MlpPolicy from stable_baselines3.sac.policies import MlpPolicy
from torchy_baselines.common.vec_env import DummyVecEnv from stable_baselines3.common.vec_env import DummyVecEnv
from torchy_baselines import SAC from stable_baselines3 import SAC
env = gym.make('Pendulum-v0') env = gym.make('Pendulum-v0')
@ -34,6 +34,6 @@ the policy is registered:
.. code-block:: python .. code-block:: python
from torchy_baselines import SAC from stable_baselines3 import SAC
model = SAC('MlpPolicy', 'Pendulum-v0').learn(10000) model = SAC('MlpPolicy', 'Pendulum-v0').learn(10000)

View file

@ -1,6 +1,6 @@
.. _vec_env: .. _vec_env:
.. automodule:: torchy_baselines.common.vec_env .. automodule:: stable_baselines3.common.vec_env
Vectorized Environments Vectorized Environments
======================= =======================

View file

@ -3,10 +3,10 @@
You can adapt this file completely to your liking, but it should at least You can adapt this file completely to your liking, but it should at least
contain the root `toctree` directive. contain the root `toctree` directive.
Welcome to Torchy Baselines docs! - Pytorch RL Baselines Welcome to Stable Baselines3 docs! - Pytorch RL Baselines
======================================================== ========================================================
`Torchy Baselines <https://github.com/hill-a/stable-baselines>`_ is the PyTorch version of `Stable Baselines <https://github.com/hill-a/stable-baselines>`_, `Stable Baselines3 <https://github.com/hill-a/stable-baselines>`_ is the PyTorch version of `Stable Baselines <https://github.com/hill-a/stable-baselines>`_,
a set of improved implementations of reinforcement learning algorithms. a set of improved implementations of reinforcement learning algorithms.
RL Baselines Zoo (collection of pre-trained agents): https://github.com/araffin/rl-baselines-zoo RL Baselines Zoo (collection of pre-trained agents): https://github.com/araffin/rl-baselines-zoo
@ -41,7 +41,7 @@ RL Baselines zoo also offers a simple interface to train, evaluate agents and do
misc/changelog misc/changelog
Citing Torchy Baselines Citing Stable Baselines3
----------------------- -----------------------
To cite this project in publications: To cite this project in publications:
@ -49,7 +49,7 @@ To cite this project in publications:
@misc{torchy-baselines, @misc{torchy-baselines,
author = {Raffin, Antonin and Hill, Ashley and Ernestus, Maximilian and Gleave, Adam and Kanervisto, Anssi and Dormann, Noah}, author = {Raffin, Antonin and Hill, Ashley and Ernestus, Maximilian and Gleave, Adam and Kanervisto, Anssi and Dormann, Noah},
title = {Torchy Baselines}, title = {Stable Baselines3},
year = {2019}, year = {2019},
publisher = {GitHub}, publisher = {GitHub},
journal = {GitHub repository}, journal = {GitHub repository},

View file

@ -111,7 +111,7 @@ Pre-Release 0.2.0 (2020-02-14)
Breaking Changes: Breaking Changes:
^^^^^^^^^^^^^^^^^ ^^^^^^^^^^^^^^^^^
- Python 2 support was dropped, Torchy Baselines now requires Python 3.6 or above - Python 2 support was dropped, Stable Baselines3 now requires Python 3.6 or above
- Return type of ``evaluation.evaluate_policy()`` has been changed - Return type of ``evaluation.evaluate_policy()`` has been changed
- Refactored the replay buffer to avoid transformation between PyTorch and NumPy - Refactored the replay buffer to avoid transformation between PyTorch and NumPy
- Created `OffPolicyRLModel` base class - Created `OffPolicyRLModel` base class
@ -160,7 +160,7 @@ New Features:
Maintainers Maintainers
----------- -----------
Torchy-Baselines is currently maintained by `Antonin Raffin`_ (aka `@araffin`_). Stable-Baselines3 is currently maintained by `Antonin Raffin`_ (aka `@araffin`_).
.. _Antonin Raffin: https://araffin.github.io/ .. _Antonin Raffin: https://araffin.github.io/
.. _@araffin: https://github.com/araffin .. _@araffin: https://github.com/araffin

View file

@ -1,6 +1,6 @@
.. _a2c: .. _a2c:
.. automodule:: torchy_baselines.a2c .. automodule:: stable_baselines3.a2c
A2C A2C
@ -44,9 +44,9 @@ Train a A2C agent on `CartPole-v1` using 4 processes.
import gym import gym
from torchy_baselines.common.policies import MlpPolicy from stable_baselines3.common.policies import MlpPolicy
from torchy_baselines.common import make_vec_env from stable_baselines3.common import make_vec_env
from torchy_baselines import A2C from stable_baselines3 import A2C
# Parallel environments # Parallel environments
env = make_vec_env('CartPole-v1', n_envs=4) env = make_vec_env('CartPole-v1', n_envs=4)

View file

@ -1,6 +1,6 @@
.. _base_algo: .. _base_algo:
.. automodule:: torchy_baselines.common.base_class .. automodule:: stable_baselines3.common.base_class
Base RL Class Base RL Class

View file

@ -1,6 +1,6 @@
.. _ppo2: .. _ppo2:
.. automodule:: torchy_baselines.ppo .. automodule:: stable_baselines3.ppo
PPO PPO
=== ===
@ -53,9 +53,9 @@ Train a PPO agent on `Pendulum-v0` using 4 processes.
import gym import gym
from torchy_baselines.ppo.policies import MlpPolicy from stable_baselines3.ppo.policies import MlpPolicy
from torchy_baselines.common.vec_env import SubprocVecEnv from stable_baselines3.common.vec_env import SubprocVecEnv
from torchy_baselines import PPO from stable_baselines3 import PPO
# multiprocess environment # multiprocess environment
n_cpu = 4 n_cpu = 4

View file

@ -1,6 +1,6 @@
.. _sac: .. _sac:
.. automodule:: torchy_baselines.sac .. automodule:: stable_baselines3.sac
SAC SAC
@ -14,7 +14,7 @@ A key feature of SAC, and a major difference with common RL algorithms, is that
.. warning:: .. warning::
The SAC model does not support ``torchy_baselines.common.policies`` because it uses double q-values The SAC model does not support ``stable_baselines3.common.policies`` because it uses double q-values
and value estimation, as a result it must use its own policy models (see :ref:`sac_policies`). and value estimation, as a result it must use its own policy models (see :ref:`sac_policies`).
@ -72,9 +72,9 @@ Example
import gym import gym
import numpy as np import numpy as np
from torchy_baselines.sac.policies import MlpPolicy from stable_baselines3.sac.policies import MlpPolicy
from torchy_baselines.common.vec_env import DummyVecEnv from stable_baselines3.common.vec_env import DummyVecEnv
from torchy_baselines import SAC from stable_baselines3 import SAC
env = gym.make('Pendulum-v0') env = gym.make('Pendulum-v0')
env = DummyVecEnv([lambda: env]) env = DummyVecEnv([lambda: env])

View file

@ -1,6 +1,6 @@
.. _td3: .. _td3:
.. automodule:: torchy_baselines.td3 .. automodule:: stable_baselines3.td3
TD3 TD3
@ -14,7 +14,7 @@ We recommend reading `OpenAI Spinning guide on TD3 <https://spinningup.openai.co
.. warning:: .. warning::
The TD3 model does not support ``torchy_baselines.common.policies`` because it uses double q-values The TD3 model does not support ``stable_baselines3.common.policies`` because it uses double q-values
estimation, as a result it must use its own policy models (see :ref:`td3_policies`). estimation, as a result it must use its own policy models (see :ref:`td3_policies`).
@ -64,9 +64,9 @@ Example
import numpy as np import numpy as np
from torchy_baselines import TD3 from stable_baselines3 import TD3
from torchy_baselines.td3.policies import MlpPolicy from stable_baselines3.td3.policies import MlpPolicy
from torchy_baselines.common.noise import NormalActionNoise, OrnsteinUhlenbeckActionNoise from stable_baselines3.common.noise import NormalActionNoise, OrnsteinUhlenbeckActionNoise
# The noise objects for TD3 # The noise objects for TD3
n_actions = env.action_space.shape[-1] n_actions = env.action_space.shape[-1]

View file

@ -18,4 +18,4 @@ filterwarnings =
ignore::UserWarning:gym ignore::UserWarning:gym
[pytype] [pytype]
inputs = torchy_baselines inputs = stable_baselines3

View file

@ -3,13 +3,13 @@ import sys
import subprocess import subprocess
from setuptools import setup, find_packages from setuptools import setup, find_packages
with open(os.path.join('torchy_baselines', 'version.txt'), 'r') as file_handler: with open(os.path.join('stable_baselines3', 'version.txt'), 'r') as file_handler:
__version__ = file_handler.read() __version__ = file_handler.read()
setup(name='torchy_baselines', setup(name='stable_baselines3',
packages=[package for package in find_packages() packages=[package for package in find_packages()
if package.startswith('torchy_baselines')], if package.startswith('stable_baselines3')],
install_requires=[ install_requires=[
'gym[classic_control]>=0.11', 'gym[classic_control]>=0.11',
'numpy', 'numpy',

View file

@ -1,9 +1,9 @@
import os import os
from torchy_baselines.a2c import A2C from stable_baselines3.a2c import A2C
from torchy_baselines.ppo import PPO from stable_baselines3.ppo import PPO
from torchy_baselines.sac import SAC from stable_baselines3.sac import SAC
from torchy_baselines.td3 import TD3 from stable_baselines3.td3 import TD3
# Read version from file # Read version from file
version_file = os.path.join(os.path.dirname(__file__), 'version.txt') version_file = os.path.join(os.path.dirname(__file__), 'version.txt')

View file

@ -0,0 +1,2 @@
from stable_baselines3.a2c.a2c import A2C
from stable_baselines3.ppo.policies import MlpPolicy

View file

@ -3,11 +3,11 @@ import torch.nn.functional as F
from gym import spaces from gym import spaces
from typing import Type, Union, Callable, Optional, Dict, Any from typing import Type, Union, Callable, Optional, Dict, Any
from torchy_baselines.common import logger from stable_baselines3.common import logger
from torchy_baselines.common.type_aliases import GymEnv, MaybeCallback from stable_baselines3.common.type_aliases import GymEnv, MaybeCallback
from torchy_baselines.common.utils import explained_variance from stable_baselines3.common.utils import explained_variance
from torchy_baselines.ppo.policies import PPOPolicy from stable_baselines3.ppo.policies import PPOPolicy
from torchy_baselines.ppo.ppo import PPO from stable_baselines3.ppo.ppo import PPO
class A2C(PPO): class A2C(PPO):

View file

@ -11,17 +11,17 @@ import gym
import torch as th import torch as th
import numpy as np import numpy as np
from torchy_baselines.common import logger from stable_baselines3.common import logger
from torchy_baselines.common.policies import BasePolicy, get_policy_from_name from stable_baselines3.common.policies import BasePolicy, get_policy_from_name
from torchy_baselines.common.utils import set_random_seed, get_schedule_fn, update_learning_rate, get_device from stable_baselines3.common.utils import set_random_seed, get_schedule_fn, update_learning_rate, get_device
from torchy_baselines.common.vec_env import DummyVecEnv, VecEnv, unwrap_vec_normalize, VecNormalize, VecTransposeImage from stable_baselines3.common.vec_env import DummyVecEnv, VecEnv, unwrap_vec_normalize, VecNormalize, VecTransposeImage
from torchy_baselines.common.preprocessing import is_image_space from stable_baselines3.common.preprocessing import is_image_space
from torchy_baselines.common.save_util import data_to_json, json_to_data, recursive_getattr, recursive_setattr from stable_baselines3.common.save_util import data_to_json, json_to_data, recursive_getattr, recursive_setattr
from torchy_baselines.common.type_aliases import GymEnv, TensorDict, RolloutReturn, MaybeCallback from stable_baselines3.common.type_aliases import GymEnv, TensorDict, RolloutReturn, MaybeCallback
from torchy_baselines.common.callbacks import BaseCallback, CallbackList, ConvertCallback, EvalCallback from stable_baselines3.common.callbacks import BaseCallback, CallbackList, ConvertCallback, EvalCallback
from torchy_baselines.common.monitor import Monitor from stable_baselines3.common.monitor import Monitor
from torchy_baselines.common.noise import ActionNoise from stable_baselines3.common.noise import ActionNoise
from torchy_baselines.common.buffers import ReplayBuffer from stable_baselines3.common.buffers import ReplayBuffer
class BaseRLModel(ABC): class BaseRLModel(ABC):

View file

@ -4,9 +4,9 @@ import numpy as np
import torch as th import torch as th
from gym import spaces from gym import spaces
from torchy_baselines.common.vec_env import VecNormalize from stable_baselines3.common.vec_env import VecNormalize
from torchy_baselines.common.type_aliases import RolloutBufferSamples, ReplayBufferSamples from stable_baselines3.common.type_aliases import RolloutBufferSamples, ReplayBufferSamples
from torchy_baselines.common.preprocessing import get_action_dim, get_obs_shape from stable_baselines3.common.preprocessing import get_action_dim, get_obs_shape
class BaseBuffer(object): class BaseBuffer(object):

View file

@ -7,12 +7,12 @@ from typing import Union, List, Dict, Any, Optional
import gym import gym
import numpy as np import numpy as np
from torchy_baselines.common.vec_env import DummyVecEnv, VecEnv, sync_envs_normalization from stable_baselines3.common.vec_env import DummyVecEnv, VecEnv, sync_envs_normalization
from torchy_baselines.common.evaluation import evaluate_policy from stable_baselines3.common.evaluation import evaluate_policy
from torchy_baselines.common.logger import Logger from stable_baselines3.common.logger import Logger
if typing.TYPE_CHECKING: if typing.TYPE_CHECKING:
from torchy_baselines.common.base_class import BaseRLModel # pytype: disable=pyi-error from stable_baselines3.common.base_class import BaseRLModel # pytype: disable=pyi-error
class BaseCallback(ABC): class BaseCallback(ABC):

View file

@ -6,7 +6,7 @@ import torch.nn as nn
from torch.distributions import Normal, Categorical from torch.distributions import Normal, Categorical
from gym import spaces from gym import spaces
from torchy_baselines.common.preprocessing import get_action_dim from stable_baselines3.common.preprocessing import get_action_dim
class Distribution(object): class Distribution(object):

View file

@ -1,7 +1,7 @@
# Copied from stable_baselines # Copied from stable_baselines
import numpy as np import numpy as np
from torchy_baselines.common.vec_env import VecEnv from stable_baselines3.common.vec_env import VecEnv
def evaluate_policy(model, env, n_eval_episodes=10, deterministic=True, def evaluate_policy(model, env, n_eval_episodes=10, deterministic=True,

View file

@ -5,7 +5,7 @@ from gym import Env
from gym.spaces import Discrete, MultiDiscrete, MultiBinary, Box from gym.spaces import Discrete, MultiDiscrete, MultiBinary, Box
from torchy_baselines.common.type_aliases import GymStepReturn, GymObs from stable_baselines3.common.type_aliases import GymStepReturn, GymObs
class IdentityEnv(Env): class IdentityEnv(Env):

View file

@ -7,9 +7,9 @@ import torch as th
import torch.nn as nn import torch.nn as nn
import numpy as np import numpy as np
from torchy_baselines.common.preprocessing import preprocess_obs, get_flattened_obs_dim, is_image_space from stable_baselines3.common.preprocessing import preprocess_obs, get_flattened_obs_dim, is_image_space
from torchy_baselines.common.utils import get_device from stable_baselines3.common.utils import get_device
from torchy_baselines.common.vec_env import VecTransposeImage from stable_baselines3.common.vec_env import VecTransposeImage
class BaseFeaturesExtractor(nn.Module): class BaseFeaturesExtractor(nn.Module):

View file

@ -6,7 +6,7 @@ import pandas as pd
# matplotlib.use('TkAgg') # Can change to 'Agg' for non-interactive mode # matplotlib.use('TkAgg') # Can change to 'Agg' for non-interactive mode
import matplotlib.pyplot as plt import matplotlib.pyplot as plt
from torchy_baselines.common.monitor import load_results from stable_baselines3.common.monitor import load_results
X_TIMESTEPS = 'timesteps' X_TIMESTEPS = 'timesteps'

View file

@ -7,8 +7,8 @@ import numpy as np
import torch as th import torch as th
import gym import gym
from torchy_baselines.common.vec_env import VecEnv from stable_baselines3.common.vec_env import VecEnv
from torchy_baselines.common.callbacks import BaseCallback from stable_baselines3.common.callbacks import BaseCallback
GymEnv = Union[gym.Env, VecEnv] GymEnv = Union[gym.Env, VecEnv]

View file

@ -3,17 +3,17 @@ import typing
from typing import Optional, Union from typing import Optional, Union
from copy import deepcopy from copy import deepcopy
from torchy_baselines.common.vec_env.base_vec_env import (AlreadySteppingError, NotSteppingError, from stable_baselines3.common.vec_env.base_vec_env import (AlreadySteppingError, NotSteppingError,
VecEnv, VecEnvWrapper, CloudpickleWrapper) VecEnv, VecEnvWrapper, CloudpickleWrapper)
from torchy_baselines.common.vec_env.dummy_vec_env import DummyVecEnv from stable_baselines3.common.vec_env.dummy_vec_env import DummyVecEnv
from torchy_baselines.common.vec_env.subproc_vec_env import SubprocVecEnv from stable_baselines3.common.vec_env.subproc_vec_env import SubprocVecEnv
from torchy_baselines.common.vec_env.vec_frame_stack import VecFrameStack from stable_baselines3.common.vec_env.vec_frame_stack import VecFrameStack
from torchy_baselines.common.vec_env.vec_normalize import VecNormalize from stable_baselines3.common.vec_env.vec_normalize import VecNormalize
from torchy_baselines.common.vec_env.vec_transpose import VecTransposeImage from stable_baselines3.common.vec_env.vec_transpose import VecTransposeImage
# Avoid circular import # Avoid circular import
if typing.TYPE_CHECKING: if typing.TYPE_CHECKING:
from torchy_baselines.common.type_aliases import GymEnv from stable_baselines3.common.type_aliases import GymEnv
def unwrap_vec_normalize(env: Union['GymEnv', VecEnv]) -> Optional[VecNormalize]: def unwrap_vec_normalize(env: Union['GymEnv', VecEnv]) -> Optional[VecNormalize]:

View file

@ -3,8 +3,8 @@ from copy import deepcopy
import numpy as np import numpy as np
from torchy_baselines.common.vec_env.base_vec_env import VecEnv from stable_baselines3.common.vec_env.base_vec_env import VecEnv
from torchy_baselines.common.vec_env.util import copy_obs_dict, dict_to_obs, obs_space_info from stable_baselines3.common.vec_env.util import copy_obs_dict, dict_to_obs, obs_space_info
class DummyVecEnv(VecEnv): class DummyVecEnv(VecEnv):

View file

@ -4,7 +4,7 @@ from collections import OrderedDict
import gym import gym
import numpy as np import numpy as np
from torchy_baselines.common.vec_env.base_vec_env import VecEnv, CloudpickleWrapper from stable_baselines3.common.vec_env.base_vec_env import VecEnv, CloudpickleWrapper
def _worker(remote, parent_remote, env_fn_wrapper): def _worker(remote, parent_remote, env_fn_wrapper):

View file

@ -3,7 +3,7 @@ import warnings
import numpy as np import numpy as np
from gym import spaces from gym import spaces
from torchy_baselines.common.vec_env.base_vec_env import VecEnv, VecEnvWrapper from stable_baselines3.common.vec_env.base_vec_env import VecEnv, VecEnvWrapper
class VecFrameStack(VecEnvWrapper): class VecFrameStack(VecEnvWrapper):

View file

@ -2,8 +2,8 @@ import pickle
import numpy as np import numpy as np
from torchy_baselines.common.vec_env.base_vec_env import VecEnvWrapper from stable_baselines3.common.vec_env.base_vec_env import VecEnvWrapper
from torchy_baselines.common.running_mean_std import RunningMeanStd from stable_baselines3.common.running_mean_std import RunningMeanStd
class VecNormalize(VecEnvWrapper): class VecNormalize(VecEnvWrapper):

View file

@ -2,11 +2,11 @@ import typing
import numpy as np import numpy as np
from gym import spaces from gym import spaces
from torchy_baselines.common.vec_env.base_vec_env import VecEnv, VecEnvWrapper from stable_baselines3.common.vec_env.base_vec_env import VecEnv, VecEnvWrapper
from torchy_baselines.common.preprocessing import is_image_space from stable_baselines3.common.preprocessing import is_image_space
if typing.TYPE_CHECKING: if typing.TYPE_CHECKING:
from torchy_baselines.common.type_aliases import GymStepReturn from stable_baselines3.common.type_aliases import GymStepReturn
class VecTransposeImage(VecEnvWrapper): class VecTransposeImage(VecEnvWrapper):

View file

@ -0,0 +1,2 @@
from stable_baselines3.ppo.ppo import PPO
from stable_baselines3.ppo.policies import MlpPolicy

View file

@ -6,10 +6,10 @@ import torch as th
import torch.nn as nn import torch.nn as nn
import numpy as np import numpy as np
from torchy_baselines.common.policies import (BasePolicy, register_policy, MlpExtractor, from stable_baselines3.common.policies import (BasePolicy, register_policy, MlpExtractor,
create_sde_features_extractor, NatureCNN, create_sde_features_extractor, NatureCNN,
BaseFeaturesExtractor, FlattenExtractor) BaseFeaturesExtractor, FlattenExtractor)
from torchy_baselines.common.distributions import (make_proba_distribution, Distribution, from stable_baselines3.common.distributions import (make_proba_distribution, Distribution,
DiagGaussianDistribution, CategoricalDistribution, DiagGaussianDistribution, CategoricalDistribution,
StateDependentNoiseDistribution) StateDependentNoiseDistribution)

View file

@ -14,14 +14,14 @@ import torch.nn.functional as F
# SummaryWriter = None # SummaryWriter = None
import numpy as np import numpy as np
from torchy_baselines.common import logger from stable_baselines3.common import logger
from torchy_baselines.common.base_class import BaseRLModel from stable_baselines3.common.base_class import BaseRLModel
from torchy_baselines.common.type_aliases import GymEnv, MaybeCallback from stable_baselines3.common.type_aliases import GymEnv, MaybeCallback
from torchy_baselines.common.buffers import RolloutBuffer from stable_baselines3.common.buffers import RolloutBuffer
from torchy_baselines.common.utils import explained_variance, get_schedule_fn from stable_baselines3.common.utils import explained_variance, get_schedule_fn
from torchy_baselines.common.vec_env import VecEnv from stable_baselines3.common.vec_env import VecEnv
from torchy_baselines.common.callbacks import BaseCallback from stable_baselines3.common.callbacks import BaseCallback
from torchy_baselines.ppo.policies import PPOPolicy from stable_baselines3.ppo.policies import PPOPolicy
class PPO(BaseRLModel): class PPO(BaseRLModel):

View file

@ -0,0 +1,2 @@
from stable_baselines3.sac.sac import SAC
from stable_baselines3.sac.policies import MlpPolicy

View file

@ -4,11 +4,11 @@ import gym
import torch as th import torch as th
import torch.nn as nn import torch.nn as nn
from torchy_baselines.common.preprocessing import get_action_dim from stable_baselines3.common.preprocessing import get_action_dim
from torchy_baselines.common.policies import (BasePolicy, register_policy, create_mlp, from stable_baselines3.common.policies import (BasePolicy, register_policy, create_mlp,
create_sde_features_extractor, NatureCNN, create_sde_features_extractor, NatureCNN,
BaseFeaturesExtractor, FlattenExtractor) BaseFeaturesExtractor, FlattenExtractor)
from torchy_baselines.common.distributions import SquashedDiagGaussianDistribution, StateDependentNoiseDistribution from stable_baselines3.common.distributions import SquashedDiagGaussianDistribution, StateDependentNoiseDistribution
# CAP the standard deviation of the actor # CAP the standard deviation of the actor
LOG_STD_MAX = 2 LOG_STD_MAX = 2

View file

@ -4,11 +4,11 @@ import torch as th
import torch.nn.functional as F import torch.nn.functional as F
import numpy as np import numpy as np
from torchy_baselines.common import logger from stable_baselines3.common import logger
from torchy_baselines.common.base_class import OffPolicyRLModel from stable_baselines3.common.base_class import OffPolicyRLModel
from torchy_baselines.common.type_aliases import GymEnv, MaybeCallback from stable_baselines3.common.type_aliases import GymEnv, MaybeCallback
from torchy_baselines.common.noise import ActionNoise from stable_baselines3.common.noise import ActionNoise
from torchy_baselines.sac.policies import SACPolicy from stable_baselines3.sac.policies import SACPolicy
class SAC(OffPolicyRLModel): class SAC(OffPolicyRLModel):

View file

@ -0,0 +1,2 @@
from stable_baselines3.td3.td3 import TD3
from stable_baselines3.td3.policies import MlpPolicy

View file

@ -4,11 +4,11 @@ import gym
import torch as th import torch as th
import torch.nn as nn import torch.nn as nn
from torchy_baselines.common.preprocessing import get_action_dim from stable_baselines3.common.preprocessing import get_action_dim
from torchy_baselines.common.policies import (BasePolicy, register_policy, create_mlp, from stable_baselines3.common.policies import (BasePolicy, register_policy, create_mlp,
create_sde_features_extractor, NatureCNN, create_sde_features_extractor, NatureCNN,
BaseFeaturesExtractor, FlattenExtractor) BaseFeaturesExtractor, FlattenExtractor)
from torchy_baselines.common.distributions import StateDependentNoiseDistribution from stable_baselines3.common.distributions import StateDependentNoiseDistribution
class Actor(BasePolicy): class Actor(BasePolicy):

View file

@ -2,11 +2,11 @@ import torch as th
import torch.nn.functional as F import torch.nn.functional as F
from typing import List, Tuple, Type, Union, Callable, Optional, Dict, Any from typing import List, Tuple, Type, Union, Callable, Optional, Dict, Any
from torchy_baselines.common import logger from stable_baselines3.common import logger
from torchy_baselines.common.base_class import OffPolicyRLModel from stable_baselines3.common.base_class import OffPolicyRLModel
from torchy_baselines.common.noise import ActionNoise from stable_baselines3.common.noise import ActionNoise
from torchy_baselines.common.type_aliases import GymEnv, MaybeCallback from stable_baselines3.common.type_aliases import GymEnv, MaybeCallback
from torchy_baselines.td3.policies import TD3Policy from stable_baselines3.td3.policies import TD3Policy
class TD3(OffPolicyRLModel): class TD3(OffPolicyRLModel):

View file

@ -4,8 +4,8 @@ import shutil
import pytest import pytest
import gym import gym
from torchy_baselines import A2C, PPO, SAC, TD3 from stable_baselines3 import A2C, PPO, SAC, TD3
from torchy_baselines.common.callbacks import (CallbackList, CheckpointCallback, EvalCallback, from stable_baselines3.common.callbacks import (CallbackList, CheckpointCallback, EvalCallback,
EveryNTimesteps, StopTrainingOnRewardThreshold) EveryNTimesteps, StopTrainingOnRewardThreshold)

View file

@ -3,8 +3,8 @@ import os
import numpy as np import numpy as np
import pytest import pytest
from torchy_baselines import A2C, PPO, SAC, TD3 from stable_baselines3 import A2C, PPO, SAC, TD3
from torchy_baselines.common.identity_env import FakeImageEnv from stable_baselines3.common.identity_env import FakeImageEnv
SAVE_PATH = './cnn_model.zip' SAVE_PATH = './cnn_model.zip'

View file

@ -1,7 +1,7 @@
import pytest import pytest
import torch as th import torch as th
from torchy_baselines import A2C, PPO, SAC, TD3 from stable_baselines3 import A2C, PPO, SAC, TD3
@pytest.mark.parametrize('net_arch', [ @pytest.mark.parametrize('net_arch', [

View file

@ -1,11 +1,11 @@
import pytest import pytest
import torch as th import torch as th
from torchy_baselines import A2C, PPO from stable_baselines3 import A2C, PPO
from torchy_baselines.common.distributions import (DiagGaussianDistribution, TanhBijector, from stable_baselines3.common.distributions import (DiagGaussianDistribution, TanhBijector,
StateDependentNoiseDistribution, StateDependentNoiseDistribution,
CategoricalDistribution, SquashedDiagGaussianDistribution) CategoricalDistribution, SquashedDiagGaussianDistribution)
from torchy_baselines.common.utils import set_random_seed from stable_baselines3.common.utils import set_random_seed
N_ACTIONS = 2 N_ACTIONS = 2

View file

@ -1,10 +1,10 @@
import numpy as np import numpy as np
import pytest import pytest
from torchy_baselines import A2C, PPO, SAC, TD3 from stable_baselines3 import A2C, PPO, SAC, TD3
from torchy_baselines.common.identity_env import IdentityEnvBox, IdentityEnv from stable_baselines3.common.identity_env import IdentityEnvBox, IdentityEnv
from torchy_baselines.common.evaluation import evaluate_policy from stable_baselines3.common.evaluation import evaluate_policy
from torchy_baselines.common.noise import NormalActionNoise from stable_baselines3.common.noise import NormalActionNoise
@pytest.mark.parametrize("model_class", [A2C, PPO]) @pytest.mark.parametrize("model_class", [A2C, PPO])

View file

@ -4,7 +4,7 @@ import shutil
import pytest import pytest
import numpy as np import numpy as np
from torchy_baselines.common.logger import (make_output_format, read_csv, read_json, DEBUG, ScopedConfigure, from stable_baselines3.common.logger import (make_output_format, read_csv, read_json, DEBUG, ScopedConfigure,
info, debug, set_level, configure, logkv, logkvs, info, debug, set_level, configure, logkv, logkvs,
dumpkvs, logkv_mean, warn, error, reset) dumpkvs, logkv_mean, warn, error, reset)
@ -18,7 +18,7 @@ KEY_VALUES = {
"g": np.array([[[1]]]), "g": np.array([[[1]]]),
} }
LOG_DIR = '/tmp/torchy_baselines/' LOG_DIR = '/tmp/stable_baselines3/'
def test_main(): def test_main():

View file

@ -5,7 +5,7 @@ import os
import pandas import pandas
import gym import gym
from torchy_baselines.common.monitor import Monitor, get_monitor_files, load_results from stable_baselines3.common.monitor import Monitor, get_monitor_files, load_results
def test_monitor(tmp_path): def test_monitor(tmp_path):

View file

@ -1,8 +1,8 @@
import gym import gym
import pytest import pytest
from torchy_baselines import A2C, PPO, SAC, TD3 from stable_baselines3 import A2C, PPO, SAC, TD3
from torchy_baselines.common.vec_env import DummyVecEnv from stable_baselines3.common.vec_env import DummyVecEnv
MODEL_LIST = [ MODEL_LIST = [
PPO, PPO,

View file

@ -1,8 +1,8 @@
import numpy as np import numpy as np
import pytest import pytest
from torchy_baselines import A2C, PPO, SAC, TD3 from stable_baselines3 import A2C, PPO, SAC, TD3
from torchy_baselines.common.noise import NormalActionNoise, OrnsteinUhlenbeckActionNoise from stable_baselines3.common.noise import NormalActionNoise, OrnsteinUhlenbeckActionNoise
normal_action_noise = NormalActionNoise(np.zeros(1), 0.1 * np.ones(1)) normal_action_noise = NormalActionNoise(np.zeros(1), 0.1 * np.ones(1))

View file

@ -5,10 +5,10 @@ import pytest
import numpy as np import numpy as np
import torch as th import torch as th
from torchy_baselines import A2C, PPO, SAC, TD3 from stable_baselines3 import A2C, PPO, SAC, TD3
from torchy_baselines.common.identity_env import IdentityEnvBox from stable_baselines3.common.identity_env import IdentityEnvBox
from torchy_baselines.common.vec_env import DummyVecEnv from stable_baselines3.common.vec_env import DummyVecEnv
from torchy_baselines.common.identity_env import FakeImageEnv from stable_baselines3.common.identity_env import FakeImageEnv
MODEL_LIST = [ MODEL_LIST = [

View file

@ -2,7 +2,7 @@ import pytest
import torch as th import torch as th
from torch.distributions import Normal from torch.distributions import Normal
from torchy_baselines import A2C, TD3, SAC, PPO from stable_baselines3 import A2C, TD3, SAC, PPO
def test_state_dependent_exploration_grad(): def test_state_dependent_exploration_grad():

View file

@ -7,7 +7,7 @@ import pytest
import gym import gym
import numpy as np import numpy as np
from torchy_baselines.common.vec_env import DummyVecEnv, SubprocVecEnv, VecNormalize, VecFrameStack from stable_baselines3.common.vec_env import DummyVecEnv, SubprocVecEnv, VecNormalize, VecFrameStack
N_ENVS = 3 N_ENVS = 3
VEC_ENV_CLASSES = [DummyVecEnv, SubprocVecEnv] VEC_ENV_CLASSES = [DummyVecEnv, SubprocVecEnv]

View file

@ -2,9 +2,9 @@ import gym
import pytest import pytest
import numpy as np import numpy as np
from torchy_baselines.common.running_mean_std import RunningMeanStd from stable_baselines3.common.running_mean_std import RunningMeanStd
from torchy_baselines.common.vec_env import DummyVecEnv, VecNormalize, VecFrameStack, sync_envs_normalization, unwrap_vec_normalize from stable_baselines3.common.vec_env import DummyVecEnv, VecNormalize, VecFrameStack, sync_envs_normalization, unwrap_vec_normalize
from torchy_baselines import SAC, TD3 from stable_baselines3 import SAC, TD3
ENV_ID = 'Pendulum-v0' ENV_ID = 'Pendulum-v0'

View file

@ -1,2 +0,0 @@
from torchy_baselines.a2c.a2c import A2C
from torchy_baselines.ppo.policies import MlpPolicy

View file

@ -1,2 +0,0 @@
from torchy_baselines.ppo.ppo import PPO
from torchy_baselines.ppo.policies import MlpPolicy

View file

@ -1,2 +0,0 @@
from torchy_baselines.sac.sac import SAC
from torchy_baselines.sac.policies import MlpPolicy

View file

@ -1,2 +0,0 @@
from torchy_baselines.td3.td3 import TD3
from torchy_baselines.td3.policies import MlpPolicy