Remove all references to gym and all extra code for it.

rl_zoo3/__init__.py

@@ -1,6 +1,5 @@
 import os
 
-# Important: import gym patches before everything
 # isort: off
 
 import rl_zoo3.gym_patches  # noqa: F401

rl_zoo3/exp_manager.py

@@ -9,7 +9,6 @@
 from pprint import pprint
 from typing import Any, Callable, Dict, List, Optional, Tuple, Union
 
-import gym as gym26
 import gymnasium as gym
 import numpy as np
 import optuna
@@ -514,10 +513,7 @@ def create_callbacks(self):
 
     @staticmethod
     def entry_point(env_id: str) -> str:
-        try:
-            return str(gym.envs.registry[env_id].entry_point)  # pytype: disable=module-attr
-        except KeyError:
-            return str(gym26.envs.registry[env_id].entry_point)  # pytype: disable=module-attr
+        return str(gym.envs.registry[env_id].entry_point)  # pytype: disable=module-attr
 
     @staticmethod
     def is_atari(env_id: str) -> bool:
@@ -600,23 +596,13 @@ def create_envs(self, n_envs: int, eval_env: bool = False, no_log: bool = False)
         ):
             self.monitor_kwargs = dict(info_keywords=("is_success",))
 
-        # Make Pybullet compatible with gym 0.26
-        if self.is_bullet(self.env_name.gym_id):
-            spec = gym26.spec(self.env_name.gym_id)
-            self.env_kwargs.update(dict(apply_api_compatibility=True))
-        else:
-            # Define make_env here so it works with subprocesses
-            # when the registry was modified with `--gym-packages`
-            # See https://github.com/HumanCompatibleAI/imitation/pull/160
-            try:
-                spec = gym.spec(self.env_name.gym_id)
-            except gym.error.NameNotFound:
-                # Registered with gym 0.26
-                spec = gym26.spec(self.env_name.gym_id)
+        spec = gym.spec(self.env_name.gym_id)
 
+        # Define make_env here, so it works with subprocesses
+        # when the registry was modified with `--gym-packages`
+        # See https://github.com/HumanCompatibleAI/imitation/pull/160
         def make_env(**kwargs) -> gym.Env:
-            env = spec.make(**kwargs)
-            return env
+            return spec.make(**kwargs)
 
         # On most env, SubprocVecEnv does not help and is quite memory hungry,
         # therefore, we use DummyVecEnv by default

rl_zoo3/gym_patches.py

@@ -9,7 +9,6 @@
 # Deprecation warning with gym 0.26 and numpy 1.24
 np.bool8 = np.bool_  # type: ignore[attr-defined]
 
-import gym  # noqa: E402
 import gymnasium  # noqa: E402
 
 
@@ -50,42 +49,6 @@ def step(self, action):
         return observation, reward, terminated, truncated, info
 
 
-# Use gym as base class otherwise the patch_env won't work
-class PatchedGymTimeLimit(gym.wrappers.TimeLimit):
-    """
-    See https://github.com/openai/gym/issues/3102
-    and https://github.com/Farama-Foundation/Gymnasium/pull/101:
-    keep the behavior as before and provide additionnal info
-    that the episode reached a timeout, but only
-    when the episode is over because of that.
-    """
-
-    def step(self, action):
-        observation, reward, terminated, truncated, info = self.env.step(action)
-        self._elapsed_steps += 1
-
-        if self._elapsed_steps >= self._max_episode_steps:
-            done = truncated or terminated
-            # TimeLimit.truncated key may have been already set by the environment
-            # do not overwrite it
-            # only set it when the episode is not over for other reasons
-            episode_truncated = not done or info.get("TimeLimit.truncated", False)
-            info["TimeLimit.truncated"] = episode_truncated
-            # truncated may have been set by the env too
-            truncated = truncated or episode_truncated
-
-        return observation, reward, terminated, truncated, info
-
-
-# Patch Gym registry (for Pybullet)
-patched_registry = PatchedRegistry()
-patched_registry.update(gym.envs.registration.registry)
-gym.envs.registry = patched_registry
-gym.envs.registration.registry = patched_registry
-# Patch gym TimeLimit
-gym.wrappers.TimeLimit = PatchedGymTimeLimit  # type: ignore[misc]
-gym.wrappers.time_limit.TimeLimit = PatchedGymTimeLimit  # type: ignore[misc]
-gym.envs.registration.TimeLimit = PatchedGymTimeLimit  # type: ignore[misc]
 # Patch Gymnasium TimeLimit
 gymnasium.wrappers.TimeLimit = PatchedTimeLimit  # type: ignore[misc]
 gymnasium.wrappers.time_limit.TimeLimit = PatchedTimeLimit  # type: ignore[misc]

rl_zoo3/train.py

@@ -5,7 +5,6 @@
 import time
 import uuid
 
-import gym as gym26
 import gymnasium as gym
 import numpy as np
 import stable_baselines3 as sb3
@@ -159,8 +158,6 @@ def train() -> None:
 
     env_id = args.env
     registered_envs = set(gym.envs.registry.keys())  # pytype: disable=module-attr
-    # Add gym 0.26 envs
-    registered_envs.update(gym26.envs.registry.keys())  # pytype: disable=module-attr
 
     # If the environment is not found, suggest the closest match
     if env_id not in registered_envs:

rl_zoo3/utils.py

@@ -5,7 +5,6 @@
 from copy import deepcopy
 from typing import Any, Callable, Dict, List, Optional, Tuple, Type, Union
 
-import gym as gym26
 import gymnasium as gym
 import stable_baselines3 as sb3  # noqa: F401
 import torch as th  # noqa: F401
@@ -238,23 +237,13 @@ def create_test_env(
     if "render_mode" not in env_kwargs and should_render:
         env_kwargs.update(render_mode="human")
 
-    # Make Pybullet compatible with gym 0.26
-    if ExperimentManager.is_bullet(env_id):
-        spec = gym26.spec(env_id)
-        env_kwargs.update(dict(apply_api_compatibility=True))
-    else:
-        # Define make_env here so it works with subprocesses
-        # when the registry was modified with `--gym-packages`
-        # See https://github.com/HumanCompatibleAI/imitation/pull/160
-        try:
-            spec = gym.spec(env_id)  # type: ignore[assignment]
-        except gym.error.NameNotFound:
-            # Registered with gym 0.26
-            spec = gym26.spec(env_id)
+    spec = gym.spec(env_id)
 
+    # Define make_env here, so it works with subprocesses
+    # when the registry was modified with `--gym-packages`
+    # See https://github.com/HumanCompatibleAI/imitation/pull/160
     def make_env(**kwargs) -> gym.Env:
-        env = spec.make(**kwargs)
-        return env  # type: ignore[return-value]
+        return spec.make(**kwargs)
 
     env = make_vec_env(
         make_env,

setup.py

@@ -28,7 +28,7 @@
     entry_points={"console_scripts": ["rl_zoo3=rl_zoo3.cli:main"]},
     install_requires=[
         "sb3_contrib>=2.1.0",
-        "gym==0.26.2",  # for patches to make gym backward compat
+        "gymnasium~=0.29.1",
         "huggingface_sb3>=2.3",
         "tqdm",
         "rich",
@@ -45,7 +45,7 @@
     url="https://github.com/DLR-RM/rl-baselines3-zoo",
     author_email="[email protected]",
     keywords="reinforcement-learning-algorithms reinforcement-learning machine-learning "
-    "gym gymnasium openai stable baselines sb3 toolbox python data-science",
+    "gymnasium openai stable baselines sb3 toolbox python data-science",
     license="MIT",
     long_description=long_description,
     long_description_content_type="text/markdown",