From 1f702d70a143c859ded1432d85bd9513d9dc3398 Mon Sep 17 00:00:00 2001
From: Bolun <36321182+BolunDai0216@users.noreply.github.com>
Date: Sat, 22 Oct 2022 10:58:56 -0400
Subject: [PATCH 1/8] added docstring to a few envs

---
 gym_miniworld/envs/collecthealth.py | 34 ++++++++++++++++++++++++++--
 gym_miniworld/envs/fourrooms.py     | 32 ++++++++++++++++++++++----
 gym_miniworld/envs/hallway.py       | 31 +++++++++++++++++++++++--
 gym_miniworld/envs/maze.py          | 35 +++++++++++++++++++++++++++--
 gym_miniworld/miniworld.py          | 16 ++++++-------
 5 files changed, 130 insertions(+), 18 deletions(-)

diff --git a/gym_miniworld/envs/collecthealth.py b/gym_miniworld/envs/collecthealth.py
index 56877adb..54380ec4 100644
--- a/gym_miniworld/envs/collecthealth.py
+++ b/gym_miniworld/envs/collecthealth.py
@@ -4,15 +4,45 @@
 
 class CollectHealth(MiniWorldEnv):
     """
+    ## Description
+
     Environment where the agent has to collect health kits and stay
     alive as long as possible. This is inspired from the VizDoom
-    HealthGathering environment. Please note, however, that the rewards
+    `HealthGathering` environment. Please note, however, that the rewards
     produced by this environment are not directly comparable to those
     of the VizDoom environment.
 
-    reward:
+    ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+    | 3   | move back                   |
+    | 4   | pick up                     |
+    | 5   | drop                        |
+    | 6   | toggle / activate an object |
+    | 7   | complete task               |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
     +2 for each time step
     -100 for dying
+
+    ## Arguments
+
+    ```python
+    CollectHealth(size=16)
+    ```
+
+    `size`: size of the room
+
     """
 
     def __init__(self, size=16, **kwargs):
diff --git a/gym_miniworld/envs/fourrooms.py b/gym_miniworld/envs/fourrooms.py
index 7737fabf..569878dd 100644
--- a/gym_miniworld/envs/fourrooms.py
+++ b/gym_miniworld/envs/fourrooms.py
@@ -1,13 +1,37 @@
-from gymnasium import spaces
-
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
+from gymnasium import spaces
 
 
 class FourRooms(MiniWorldEnv):
     """
-    Classic four rooms environment.
-    The agent must reach the red box to get a reward.
+    ## Description
+
+    Classic four rooms environment. The agent must reach the red box to get a reward.
+
+    ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
+    +1 - 0.2 * (step_count / max_episode_steps) when red box reached
+
+    ## Arguments
+
+    ```python
+    FourRooms()
+    ```
+
     """
 
     def __init__(self, **kwargs):
diff --git a/gym_miniworld/envs/hallway.py b/gym_miniworld/envs/hallway.py
index 711e2da3..e75c61b2 100644
--- a/gym_miniworld/envs/hallway.py
+++ b/gym_miniworld/envs/hallway.py
@@ -1,15 +1,42 @@
 import math
 
-from gymnasium import spaces
-
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
+from gymnasium import spaces
 
 
 class Hallway(MiniWorldEnv):
     """
+    ## Description
+
     Environment in which the goal is to go to a red box
     at the end of a hallway
+
+    ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
+    +1 - 0.2 * (step_count / max_episode_steps) when red box reached
+
+    ## Arguments
+
+    ```python
+    FourRooms(length=12)
+    ```
+
+    `length`: length of the entire space
+
     """
 
     def __init__(self, length=12, **kwargs):
diff --git a/gym_miniworld/envs/maze.py b/gym_miniworld/envs/maze.py
index 9c23ff5b..8997d9fc 100644
--- a/gym_miniworld/envs/maze.py
+++ b/gym_miniworld/envs/maze.py
@@ -1,13 +1,44 @@
-from gymnasium import spaces
-
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
 from gym_miniworld.params import DEFAULT_PARAMS
+from gymnasium import spaces
 
 
 class Maze(MiniWorldEnv):
     """
+    ## Description
+
     Maze environment in which the agent has to reach a red box
+
+     ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
+    +1 - 0.2 * (step_count / max_episode_steps) when red box reached
+
+    ## Arguments
+
+    ```python
+    Maze(num_rows=8, num_cols=8, room_size=3)
+    ```
+
+    `num_rows`: number of rows
+
+    `num_cols`: number of columns
+
+    `room_size`: size of rooms
+
     """
 
     def __init__(
diff --git a/gym_miniworld/miniworld.py b/gym_miniworld/miniworld.py
index c8d9bb1b..ce413eb0 100644
--- a/gym_miniworld/miniworld.py
+++ b/gym_miniworld/miniworld.py
@@ -470,15 +470,15 @@ class Actions(IntEnum):
 
     def __init__(
         self,
-        max_episode_steps=1500,
-        obs_width=80,
-        obs_height=60,
-        window_width=800,
-        window_height=600,
+        max_episode_steps: int = 1500,
+        obs_width: int = 80,
+        obs_height: int = 60,
+        window_width: int = 800,
+        window_height: int = 600,
         params=DEFAULT_PARAMS,
-        domain_rand=False,
-        render_mode=None,
-        view="agent",
+        domain_rand: bool = False,
+        render_mode: Optional[str] = None,
+        view: str = "agent",
     ):
         # Action enumeration for this environment
         self.actions = MiniWorldEnv.Actions

From b64013cf5a5e7d64baa04af21fac7ea15c0e40a4 Mon Sep 17 00:00:00 2001
From: Bolun <36321182+BolunDai0216@users.noreply.github.com>
Date: Sat, 22 Oct 2022 11:00:31 -0400
Subject: [PATCH 2/8] passed pre-commit

---
 gym_miniworld/envs/fourrooms.py | 3 ++-
 gym_miniworld/envs/hallway.py   | 3 ++-
 gym_miniworld/envs/maze.py      | 3 ++-
 3 files changed, 6 insertions(+), 3 deletions(-)

diff --git a/gym_miniworld/envs/fourrooms.py b/gym_miniworld/envs/fourrooms.py
index 569878dd..7aefe9f9 100644
--- a/gym_miniworld/envs/fourrooms.py
+++ b/gym_miniworld/envs/fourrooms.py
@@ -1,6 +1,7 @@
+from gymnasium import spaces
+
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
-from gymnasium import spaces
 
 
 class FourRooms(MiniWorldEnv):
diff --git a/gym_miniworld/envs/hallway.py b/gym_miniworld/envs/hallway.py
index e75c61b2..23060646 100644
--- a/gym_miniworld/envs/hallway.py
+++ b/gym_miniworld/envs/hallway.py
@@ -1,8 +1,9 @@
 import math
 
+from gymnasium import spaces
+
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
-from gymnasium import spaces
 
 
 class Hallway(MiniWorldEnv):
diff --git a/gym_miniworld/envs/maze.py b/gym_miniworld/envs/maze.py
index 8997d9fc..653e75a3 100644
--- a/gym_miniworld/envs/maze.py
+++ b/gym_miniworld/envs/maze.py
@@ -1,7 +1,8 @@
+from gymnasium import spaces
+
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
 from gym_miniworld.params import DEFAULT_PARAMS
-from gymnasium import spaces
 
 
 class Maze(MiniWorldEnv):

From be10adbdb4388c8e652e14116c6a60c0b8229936 Mon Sep 17 00:00:00 2001
From: Bolun <36321182+BolunDai0216@users.noreply.github.com>
Date: Sat, 29 Oct 2022 14:40:22 -0400
Subject: [PATCH 3/8] updated manual_control to conform with the new api

---
 scripts/benchmark.py      |  2 +-
 scripts/manual_control.py | 22 +++++++++++-----------
 2 files changed, 12 insertions(+), 12 deletions(-)

diff --git a/scripts/benchmark.py b/scripts/benchmark.py
index b8bb43cf..c25f0066 100755
--- a/scripts/benchmark.py
+++ b/scripts/benchmark.py
@@ -2,7 +2,7 @@
 
 import time
 
-import gym
+import gymnasium as gym
 
 # Benchmark loading time
 st = time.time()
diff --git a/scripts/manual_control.py b/scripts/manual_control.py
index c2559563..c7303d71 100755
--- a/scripts/manual_control.py
+++ b/scripts/manual_control.py
@@ -9,12 +9,13 @@
 import math
 import sys
 
-import gym
+import gym_miniworld
+import gymnasium as gym
 import pyglet
 from pyglet.window import key
 
 parser = argparse.ArgumentParser()
-parser.add_argument("--env-name", default="MiniWorld-Hallway-v0")
+parser.add_argument("--env-name", default=gym_miniworld.envs.env_ids[0])
 parser.add_argument(
     "--domain-rand", action="store_true", help="enable domain randomization"
 )
@@ -27,20 +28,19 @@
     help="show the top view instead of the agent view",
 )
 args = parser.parse_args()
+view_mode = "top" if args.top_view else "agent"
 
-env = gym.make(args.env_name)
+env = gym.make(args.env_name, view=view_mode, render_mode="human")
 
 if args.no_time_limit:
     env.max_episode_steps = math.inf
 if args.domain_rand:
     env.domain_rand = True
 
-view_mode = "top" if args.top_view else "agent"
-
 env.reset()
 
 # Create the display window
-env.render("pyglet", view=view_mode)
+env.render()
 
 
 def step(action):
@@ -50,16 +50,16 @@ def step(action):
         )
     )
 
-    obs, reward, done, info = env.step(action)
+    obs, reward, termination, truncation, info = env.step(action)
 
     if reward > 0:
         print(f"reward={reward:.2f}")
 
-    if done:
+    if termination:
         print("done!")
         env.reset()
 
-    env.render("pyglet", view=view_mode)
+    env.render()
 
 
 @env.unwrapped.window.event
@@ -72,7 +72,7 @@ def on_key_press(symbol, modifiers):
     if symbol == key.BACKSPACE or symbol == key.SLASH:
         print("RESET")
         env.reset()
-        env.render("pyglet", view=view_mode)
+        env.render()
         return
 
     if symbol == key.ESCAPE:
@@ -105,7 +105,7 @@ def on_key_release(symbol, modifiers):
 
 @env.unwrapped.window.event
 def on_draw():
-    env.render("pyglet", view=view_mode)
+    env.render()
 
 
 @env.unwrapped.window.event

From de32af49904af0fa1d21276415fc9d19e3d598e0 Mon Sep 17 00:00:00 2001
From: Bolun <36321182+BolunDai0216@users.noreply.github.com>
Date: Sat, 29 Oct 2022 14:40:43 -0400
Subject: [PATCH 4/8] updated manual_control to conform with the new api

---
 scripts/manual_control.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/scripts/manual_control.py b/scripts/manual_control.py
index c7303d71..6b62d153 100755
--- a/scripts/manual_control.py
+++ b/scripts/manual_control.py
@@ -9,11 +9,12 @@
 import math
 import sys
 
-import gym_miniworld
 import gymnasium as gym
 import pyglet
 from pyglet.window import key
 
+import gym_miniworld
+
 parser = argparse.ArgumentParser()
 parser.add_argument("--env-name", default=gym_miniworld.envs.env_ids[0])
 parser.add_argument(

From 68a938b08db2f3da96beaad975537bc21a459001 Mon Sep 17 00:00:00 2001
From: Bolun <36321182+BolunDai0216@users.noreply.github.com>
Date: Sun, 30 Oct 2022 15:09:22 -0400
Subject: [PATCH 5/8] added docstrings to all envs

---
 gym_miniworld/envs/fourrooms.py     |  5 ++-
 gym_miniworld/envs/hallway.py       |  5 ++-
 gym_miniworld/envs/maze.py          | 19 +++++-----
 gym_miniworld/envs/oneroom.py       | 31 ++++++++++++++--
 gym_miniworld/envs/pickupobjs.py    | 35 ++++++++++++++++--
 gym_miniworld/envs/putnext.py       | 33 +++++++++++++++++
 gym_miniworld/envs/remotebot.py     |  6 +---
 gym_miniworld/envs/roomobjs.py      | 33 +++++++++++++++++
 gym_miniworld/envs/sidewalk.py      | 29 +++++++++++++--
 gym_miniworld/envs/sign.py          | 56 ++++++++++++++++++++---------
 gym_miniworld/envs/simtorealgoto.py | 29 +++++++++++++--
 gym_miniworld/envs/simtorealpush.py | 30 ++++++++++++++--
 gym_miniworld/envs/threerooms.py    | 29 +++++++++++++--
 gym_miniworld/envs/tmaze.py         | 32 +++++++++++++++--
 gym_miniworld/envs/wallgap.py       | 28 +++++++++++++--
 gym_miniworld/envs/ymaze.py         | 31 ++++++++++++++--
 scripts/benchmark.py                |  5 +--
 scripts/manual_control.py           |  7 ++--
 18 files changed, 381 insertions(+), 62 deletions(-)

diff --git a/gym_miniworld/envs/fourrooms.py b/gym_miniworld/envs/fourrooms.py
index 7aefe9f9..407156c8 100644
--- a/gym_miniworld/envs/fourrooms.py
+++ b/gym_miniworld/envs/fourrooms.py
@@ -1,7 +1,6 @@
-from gymnasium import spaces
-
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
+from gymnasium import spaces
 
 
 class FourRooms(MiniWorldEnv):
@@ -25,7 +24,7 @@ class FourRooms(MiniWorldEnv):
 
     ## Rewards:
 
-    +1 - 0.2 * (step_count / max_episode_steps) when red box reached
+    +(1 - 0.2 * (step_count / max_episode_steps)) when red box reached
 
     ## Arguments
 
diff --git a/gym_miniworld/envs/hallway.py b/gym_miniworld/envs/hallway.py
index 23060646..4e625188 100644
--- a/gym_miniworld/envs/hallway.py
+++ b/gym_miniworld/envs/hallway.py
@@ -1,9 +1,8 @@
 import math
 
-from gymnasium import spaces
-
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
+from gymnasium import spaces
 
 
 class Hallway(MiniWorldEnv):
@@ -28,7 +27,7 @@ class Hallway(MiniWorldEnv):
 
     ## Rewards:
 
-    +1 - 0.2 * (step_count / max_episode_steps) when red box reached
+    +(1 - 0.2 * (step_count / max_episode_steps)) when red box reached
 
     ## Arguments
 
diff --git a/gym_miniworld/envs/maze.py b/gym_miniworld/envs/maze.py
index 653e75a3..d73313f4 100644
--- a/gym_miniworld/envs/maze.py
+++ b/gym_miniworld/envs/maze.py
@@ -1,8 +1,7 @@
-from gymnasium import spaces
-
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
 from gym_miniworld.params import DEFAULT_PARAMS
+from gymnasium import spaces
 
 
 class Maze(MiniWorldEnv):
@@ -11,7 +10,7 @@ class Maze(MiniWorldEnv):
 
     Maze environment in which the agent has to reach a red box
 
-     ## Action Space
+    ## Action Space
 
     | Num | Action                      |
     |-----|-----------------------------|
@@ -26,20 +25,18 @@ class Maze(MiniWorldEnv):
 
     ## Rewards:
 
-    +1 - 0.2 * (step_count / max_episode_steps) when red box reached
+    +(1 - 0.2 * (step_count / max_episode_steps)) when red box reached
 
     ## Arguments
 
     ```python
-    Maze(num_rows=8, num_cols=8, room_size=3)
+    MazeS2()
+    # or
+    MazeS3()
+    # or
+    MazeS3Fast()
     ```
 
-    `num_rows`: number of rows
-
-    `num_cols`: number of columns
-
-    `room_size`: size of rooms
-
     """
 
     def __init__(
diff --git a/gym_miniworld/envs/oneroom.py b/gym_miniworld/envs/oneroom.py
index df8e0d18..260e92a3 100644
--- a/gym_miniworld/envs/oneroom.py
+++ b/gym_miniworld/envs/oneroom.py
@@ -1,14 +1,41 @@
-from gymnasium import spaces
-
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
 from gym_miniworld.params import DEFAULT_PARAMS
+from gymnasium import spaces
 
 
 class OneRoom(MiniWorldEnv):
     """
+    ## Description
+
     Environment in which the goal is to go to a red box
     placed randomly in one big room.
+
+    ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
+    +(1 - 0.2 * (step_count / max_episode_steps)) when red box reached
+
+    ## Arguments
+
+    ```python
+    OneRoomS6()
+    # or
+    OneRoomS6Fast()
+    ```
+
     """
 
     def __init__(self, size=10, max_episode_steps=180, **kwargs):
diff --git a/gym_miniworld/envs/pickupobjs.py b/gym_miniworld/envs/pickupobjs.py
index 6cc3d9e8..ad294f27 100644
--- a/gym_miniworld/envs/pickupobjs.py
+++ b/gym_miniworld/envs/pickupobjs.py
@@ -1,13 +1,44 @@
-from gymnasium import spaces
-
 from gym_miniworld.entity import Ball, Box, Key
 from gym_miniworld.miniworld import MiniWorldEnv
+from gymnasium import spaces
 
 
 class PickupObjs(MiniWorldEnv):
     """
+    ## Description
+
     Room with multiple objects. The agent collects +1 reward for picking up
     each object. Objects disappear when picked up.
+
+    ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+    | 3   | move_back                   |
+    | 4   | pickup                      |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
+    +1 when agent picked up object
+
+    ## Arguments
+
+    ```python
+    PickupObjs(size=12, num_objs=5)
+    ```
+
+    `size`: size of world
+
+    `num_objs`: number of objects
+
     """
 
     def __init__(self, size=12, num_objs=5, **kwargs):
diff --git a/gym_miniworld/envs/putnext.py b/gym_miniworld/envs/putnext.py
index ed3ff2eb..7df931fb 100644
--- a/gym_miniworld/envs/putnext.py
+++ b/gym_miniworld/envs/putnext.py
@@ -4,8 +4,41 @@
 
 class PutNext(MiniWorldEnv):
     """
+    ## Description
+
     Single-room environment where a red box must be placed next
     to a yellow box.
+
+    ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+    | 3   | move back                   |
+    | 4   | pick up                     |
+    | 5   | drop                        |
+    | 6   | toggle / activate an object |
+    | 7   | complete task               |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
+    +(1 - 0.2 * (step_count / max_episode_steps)) when red box is next to yellow box
+
+    ## Arguments
+
+    ```python
+    PutNext(size=12)
+    ```
+
+    `size`: size of world
+
     """
 
     def __init__(self, size=12, **kwargs):
diff --git a/gym_miniworld/envs/remotebot.py b/gym_miniworld/envs/remotebot.py
index fd0a3585..bc48e505 100644
--- a/gym_miniworld/envs/remotebot.py
+++ b/gym_miniworld/envs/remotebot.py
@@ -7,6 +7,7 @@
 import numpy
 import numpy as np
 import pyglet
+from gym_miniworld.miniworld import MiniWorldEnv
 from gymnasium import spaces
 
 # Try importing ZMQ
@@ -21,11 +22,6 @@
     glViewport,
 )
 
-from gym_miniworld.miniworld import MiniWorldEnv
-
-# from gym.utils import seeding
-
-
 try:
     import zmq
 except ImportError:
diff --git a/gym_miniworld/envs/roomobjs.py b/gym_miniworld/envs/roomobjs.py
index 3ef8c4ae..9f195dfa 100644
--- a/gym_miniworld/envs/roomobjs.py
+++ b/gym_miniworld/envs/roomobjs.py
@@ -6,10 +6,43 @@
 
 class RoomObjs(MiniWorldEnv):
     """
+    ## Description
+
     Single room with multiple objects
     Inspired by the single room environment of
     the Generative Query Networks paper:
     https://deepmind.com/blog/neural-scene-representation-and-rendering/
+
+    ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+    | 3   | move back                   |
+    | 4   | pick up                     |
+    | 5   | drop                        |
+    | 6   | toggle / activate an object |
+    | 7   | complete task               |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
+    +0
+
+    ## Arguments
+
+    ```python
+    RoomObjs(size=16)
+    ```
+
+    `size`: size of world
+
     """
 
     def __init__(self, size=10, **kwargs):
diff --git a/gym_miniworld/envs/sidewalk.py b/gym_miniworld/envs/sidewalk.py
index 5ea93359..332b004f 100644
--- a/gym_miniworld/envs/sidewalk.py
+++ b/gym_miniworld/envs/sidewalk.py
@@ -1,16 +1,41 @@
 import math
 
 import numpy as np
-from gymnasium import spaces
-
 from gym_miniworld.entity import Box, MeshEnt
 from gym_miniworld.miniworld import MiniWorldEnv
+from gymnasium import spaces
 
 
 class Sidewalk(MiniWorldEnv):
     """
+    ## Description
+
     Walk on a sidewalk up to an object to be collected.
     Don't walk into the street.
+
+    ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
+    +(1 - 0.2 * (step_count / max_episode_steps)) when object reached
+
+    ## Arguments
+
+    ```python
+    Sidewalk()
+    ```
+
     """
 
     def __init__(self, **kwargs):
diff --git a/gym_miniworld/envs/sign.py b/gym_miniworld/envs/sign.py
index 0a64098c..b028b511 100644
--- a/gym_miniworld/envs/sign.py
+++ b/gym_miniworld/envs/sign.py
@@ -2,12 +2,11 @@
 from typing import Optional, Tuple
 
 import gymnasium as gym
-from gymnasium.core import ObsType
-from gymnasium.spaces import Dict, Discrete
-
 from gym_miniworld.entity import COLOR_NAMES, Box, Key, MeshEnt, TextFrame
 from gym_miniworld.miniworld import MiniWorldEnv
 from gym_miniworld.params import DEFAULT_PARAMS
+from gymnasium.core import ObsType
+from gymnasium.spaces import Dict, Discrete
 
 
 class BigKey(Key):
@@ -19,7 +18,10 @@ def __init__(self, color, size=0.6):
 
 
 class Sign(MiniWorldEnv):
-    """Sign environment from https://arxiv.org/abs/2008.02790.
+    """
+    ## Description
+
+    Sign environment from https://arxiv.org/abs/2008.02790.
 
     If you use this environment, please cite the above paper (Liu et al., 2020).
 
@@ -30,26 +32,48 @@ class Sign(MiniWorldEnv):
     includes a goal under state["goal"] that specifies box or key.
 
     The episode ends when any object is touched.
-    Touching the object where the color matches the sign and the shape matches the
-    goal yields +1 reward.
-    Touching any other object yields -1 reward.
 
     The sign and goal can be configured via the color_index and goal arguments to
     the constructor respectively.
 
     Includes an action to end the episode.
+
+    ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+    | 3   | end episode                 |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
+    +1 for touching the object where the color matches the sign and the shape matches the goal
+    -1 for touching any other object
+
+    ## Arguments
+
+    ```python
+    Sign(size=10, max_episode_steps=20, color_index=0, goal=0)
+    ```
+
+    `size`:  size of the square room.
+
+    `max_episode_steps`: number of steps before the episode ends.
+
+    `color_index`: specifies whether the sign says blue (0), green (1), or red (2).
+
+    `goal`: specifies box (0) or key (1).
+
     """
 
     def __init__(self, size=10, max_episode_steps=20, color_index=0, goal=0):
-        """Constructs.
-
-        Args:
-          size (int): size of the square room.
-          max_episode_steps (int): number of steps before the episode ends.
-          color_index (int): specifies whether the sign says blue (0), green (1), or
-            red (2).
-          goal (int): specifies box (0) or key (1).
-        """
         if color_index not in [0, 1, 2]:
             raise ValueError("Only supported values for color_index are 0, 1, 2.")
 
diff --git a/gym_miniworld/envs/simtorealgoto.py b/gym_miniworld/envs/simtorealgoto.py
index e87b4e4f..22f3915e 100644
--- a/gym_miniworld/envs/simtorealgoto.py
+++ b/gym_miniworld/envs/simtorealgoto.py
@@ -1,8 +1,7 @@
-from gymnasium import spaces
-
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
 from gym_miniworld.params import DEFAULT_PARAMS
+from gymnasium import spaces
 
 # Simulation parameters
 # These assume a robot about 15cm tall with a pi camera module v2
@@ -21,8 +20,34 @@
 
 class SimToRealGoTo(MiniWorldEnv):
     """
+    ## Description
+
     Environment designed for sim-to-real transfer.
     In this environment, the robot has to go to the red box.
+
+    ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
+    +(1 - 0.2 * (step_count / max_episode_steps)) when red box reached
+
+    ## Arguments
+
+    ```python
+    SimToRealGoTo()
+    ```
+
     """
 
     def __init__(self, **kwargs):
diff --git a/gym_miniworld/envs/simtorealpush.py b/gym_miniworld/envs/simtorealpush.py
index 6598183e..d4fa3c56 100644
--- a/gym_miniworld/envs/simtorealpush.py
+++ b/gym_miniworld/envs/simtorealpush.py
@@ -1,11 +1,10 @@
 import math
 
 import numpy as np
-from gymnasium import spaces
-
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
 from gym_miniworld.params import DEFAULT_PARAMS
+from gymnasium import spaces
 
 # Simulation parameters
 # These assume a robot about 15cm tall with a pi camera module v2
@@ -23,9 +22,36 @@
 
 class SimToRealPush(MiniWorldEnv):
     """
+    ## Description
+
     Environment designed for sim-to-real transfer.
     In this environment, the robot has to push the
     red box towards the yellow box.
+
+    ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+    | 3   | move back                   |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
+    +1 when the red box is close to the yellow box
+
+    ## Arguments
+
+    ```python
+    SimToRealPush()
+    ```
+
     """
 
     def __init__(self, **kwargs):
diff --git a/gym_miniworld/envs/threerooms.py b/gym_miniworld/envs/threerooms.py
index 8d8487bc..edd59cb9 100644
--- a/gym_miniworld/envs/threerooms.py
+++ b/gym_miniworld/envs/threerooms.py
@@ -1,14 +1,39 @@
 import math
 
-from gymnasium import spaces
-
 from gym_miniworld.entity import Ball, Box, ImageFrame, Key, MeshEnt
 from gym_miniworld.miniworld import MiniWorldEnv
+from gymnasium import spaces
 
 
 class ThreeRooms(MiniWorldEnv):
     """
+    ## Description
+
     Two small rooms connected to one large room
+
+    ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
+    None
+
+    ## Arguments
+
+    ```python
+    ThreeRooms()
+    ```
+
     """
 
     def __init__(self, **kwargs):
diff --git a/gym_miniworld/envs/tmaze.py b/gym_miniworld/envs/tmaze.py
index b0cc0dca..3741da6e 100644
--- a/gym_miniworld/envs/tmaze.py
+++ b/gym_miniworld/envs/tmaze.py
@@ -1,14 +1,42 @@
 import math
 
-from gymnasium import spaces
-
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
+from gymnasium import spaces
 
 
 class TMaze(MiniWorldEnv):
     """
+    ## Description
+
     Two hallways connected in a T-junction
+
+    ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
+    +(1 - 0.2 * (step_count / max_episode_steps)) when box reached
+
+    ## Arguments
+
+    ```python
+    TMazeLeft()
+    # or
+    TMazeRight()
+    ```
+
+
     """
 
     def __init__(self, goal_pos=None, **kwargs):
diff --git a/gym_miniworld/envs/wallgap.py b/gym_miniworld/envs/wallgap.py
index 95750f2a..a95c373d 100644
--- a/gym_miniworld/envs/wallgap.py
+++ b/gym_miniworld/envs/wallgap.py
@@ -1,15 +1,39 @@
 import math
 
 import numpy as np
-from gymnasium import spaces
-
 from gym_miniworld.entity import Box, MeshEnt
 from gym_miniworld.miniworld import MiniWorldEnv
+from gymnasium import spaces
 
 
 class WallGap(MiniWorldEnv):
     """
+    ## Description
+
     Outside environment with two rooms connected by a gap in a wall
+
+    ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
+    +(1 - 0.2 * (step_count / max_episode_steps)) when box reached
+
+    ## Arguments
+
+    ```python
+    WallGap()
+    ```
     """
 
     def __init__(self, **kwargs):
diff --git a/gym_miniworld/envs/ymaze.py b/gym_miniworld/envs/ymaze.py
index e9dd9f3d..01eb9278 100644
--- a/gym_miniworld/envs/ymaze.py
+++ b/gym_miniworld/envs/ymaze.py
@@ -1,16 +1,43 @@
 import math
 
 import numpy as np
-from gymnasium import spaces
-
 from gym_miniworld.entity import Box
 from gym_miniworld.math import gen_rot_matrix
 from gym_miniworld.miniworld import MiniWorldEnv
+from gymnasium import spaces
 
 
 class YMaze(MiniWorldEnv):
     """
+    ## Description
+
     Two hallways connected in a Y-junction
+
+    ## Action Space
+
+    | Num | Action                      |
+    |-----|-----------------------------|
+    | 0   | turn left                   |
+    | 1   | turn right                  |
+    | 2   | move forward                |
+
+    ## Observation Space
+
+    The observation space is an `ndarray` with shape `(obs_height, obs_width, 3)`
+    representing the view the agents sees.
+
+    ## Rewards:
+
+    +(1 - 0.2 * (step_count / max_episode_steps)) when box reached
+
+    ## Arguments
+
+    ```python
+    YMazeLeft()
+    # or
+    YMazeRight()
+    ```
+
     """
 
     def __init__(self, goal_pos=None, **kwargs):
diff --git a/scripts/benchmark.py b/scripts/benchmark.py
index c25f0066..5c63c497 100755
--- a/scripts/benchmark.py
+++ b/scripts/benchmark.py
@@ -2,6 +2,7 @@
 
 import time
 
+import gym_miniworld
 import gymnasium as gym
 
 # Benchmark loading time
@@ -29,9 +30,9 @@
 
     # Slow movement speed to minimize resets
     action = 0
-    obs, reward, done, info = env.step(action)
+    obs, reward, termination, truncation, info = env.step(action)
 
-    if done:
+    if termination or truncation:
         env.reset()
 
     num_frames += 1
diff --git a/scripts/manual_control.py b/scripts/manual_control.py
index 6b62d153..712eaf15 100755
--- a/scripts/manual_control.py
+++ b/scripts/manual_control.py
@@ -9,12 +9,11 @@
 import math
 import sys
 
+import gym_miniworld
 import gymnasium as gym
 import pyglet
 from pyglet.window import key
 
-import gym_miniworld
-
 parser = argparse.ArgumentParser()
 parser.add_argument("--env-name", default=gym_miniworld.envs.env_ids[0])
 parser.add_argument(
@@ -56,7 +55,7 @@ def step(action):
     if reward > 0:
         print(f"reward={reward:.2f}")
 
-    if termination:
+    if termination or truncation:
         print("done!")
         env.reset()
 
@@ -78,7 +77,7 @@ def on_key_press(symbol, modifiers):
 
     if symbol == key.ESCAPE:
         env.close()
-        sys.exit(0)
+        # sys.exit(0)
 
     if symbol == key.UP:
         step(env.actions.move_forward)

From 0e47e5acd5cd6e5a5b6948f4ac27a2d625ef6ee5 Mon Sep 17 00:00:00 2001
From: Bolun <36321182+BolunDai0216@users.noreply.github.com>
Date: Sun, 30 Oct 2022 15:13:50 -0400
Subject: [PATCH 6/8] pass pre-commit

---
 gym_miniworld/envs/fourrooms.py     | 3 ++-
 gym_miniworld/envs/hallway.py       | 3 ++-
 gym_miniworld/envs/maze.py          | 3 ++-
 gym_miniworld/envs/oneroom.py       | 3 ++-
 gym_miniworld/envs/pickupobjs.py    | 3 ++-
 gym_miniworld/envs/remotebot.py     | 3 ++-
 gym_miniworld/envs/sidewalk.py      | 3 ++-
 gym_miniworld/envs/sign.py          | 5 +++--
 gym_miniworld/envs/simtorealgoto.py | 3 ++-
 gym_miniworld/envs/simtorealpush.py | 3 ++-
 gym_miniworld/envs/threerooms.py    | 3 ++-
 gym_miniworld/envs/tmaze.py         | 3 ++-
 gym_miniworld/envs/wallgap.py       | 3 ++-
 gym_miniworld/envs/ymaze.py         | 3 ++-
 scripts/benchmark.py                | 3 ++-
 scripts/manual_control.py           | 7 +++++--
 16 files changed, 36 insertions(+), 18 deletions(-)

diff --git a/gym_miniworld/envs/fourrooms.py b/gym_miniworld/envs/fourrooms.py
index 407156c8..f676de02 100644
--- a/gym_miniworld/envs/fourrooms.py
+++ b/gym_miniworld/envs/fourrooms.py
@@ -1,6 +1,7 @@
+from gymnasium import spaces
+
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
-from gymnasium import spaces
 
 
 class FourRooms(MiniWorldEnv):
diff --git a/gym_miniworld/envs/hallway.py b/gym_miniworld/envs/hallway.py
index 4e625188..4336c9ad 100644
--- a/gym_miniworld/envs/hallway.py
+++ b/gym_miniworld/envs/hallway.py
@@ -1,8 +1,9 @@
 import math
 
+from gymnasium import spaces
+
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
-from gymnasium import spaces
 
 
 class Hallway(MiniWorldEnv):
diff --git a/gym_miniworld/envs/maze.py b/gym_miniworld/envs/maze.py
index d73313f4..5363686a 100644
--- a/gym_miniworld/envs/maze.py
+++ b/gym_miniworld/envs/maze.py
@@ -1,7 +1,8 @@
+from gymnasium import spaces
+
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
 from gym_miniworld.params import DEFAULT_PARAMS
-from gymnasium import spaces
 
 
 class Maze(MiniWorldEnv):
diff --git a/gym_miniworld/envs/oneroom.py b/gym_miniworld/envs/oneroom.py
index 260e92a3..9974f43d 100644
--- a/gym_miniworld/envs/oneroom.py
+++ b/gym_miniworld/envs/oneroom.py
@@ -1,7 +1,8 @@
+from gymnasium import spaces
+
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
 from gym_miniworld.params import DEFAULT_PARAMS
-from gymnasium import spaces
 
 
 class OneRoom(MiniWorldEnv):
diff --git a/gym_miniworld/envs/pickupobjs.py b/gym_miniworld/envs/pickupobjs.py
index ad294f27..5c38e915 100644
--- a/gym_miniworld/envs/pickupobjs.py
+++ b/gym_miniworld/envs/pickupobjs.py
@@ -1,6 +1,7 @@
+from gymnasium import spaces
+
 from gym_miniworld.entity import Ball, Box, Key
 from gym_miniworld.miniworld import MiniWorldEnv
-from gymnasium import spaces
 
 
 class PickupObjs(MiniWorldEnv):
diff --git a/gym_miniworld/envs/remotebot.py b/gym_miniworld/envs/remotebot.py
index bc48e505..45bf9d82 100644
--- a/gym_miniworld/envs/remotebot.py
+++ b/gym_miniworld/envs/remotebot.py
@@ -7,7 +7,6 @@
 import numpy
 import numpy as np
 import pyglet
-from gym_miniworld.miniworld import MiniWorldEnv
 from gymnasium import spaces
 
 # Try importing ZMQ
@@ -22,6 +21,8 @@
     glViewport,
 )
 
+from gym_miniworld.miniworld import MiniWorldEnv
+
 try:
     import zmq
 except ImportError:
diff --git a/gym_miniworld/envs/sidewalk.py b/gym_miniworld/envs/sidewalk.py
index 332b004f..19bbd9e0 100644
--- a/gym_miniworld/envs/sidewalk.py
+++ b/gym_miniworld/envs/sidewalk.py
@@ -1,9 +1,10 @@
 import math
 
 import numpy as np
+from gymnasium import spaces
+
 from gym_miniworld.entity import Box, MeshEnt
 from gym_miniworld.miniworld import MiniWorldEnv
-from gymnasium import spaces
 
 
 class Sidewalk(MiniWorldEnv):
diff --git a/gym_miniworld/envs/sign.py b/gym_miniworld/envs/sign.py
index b028b511..fe944165 100644
--- a/gym_miniworld/envs/sign.py
+++ b/gym_miniworld/envs/sign.py
@@ -2,11 +2,12 @@
 from typing import Optional, Tuple
 
 import gymnasium as gym
+from gymnasium.core import ObsType
+from gymnasium.spaces import Dict, Discrete
+
 from gym_miniworld.entity import COLOR_NAMES, Box, Key, MeshEnt, TextFrame
 from gym_miniworld.miniworld import MiniWorldEnv
 from gym_miniworld.params import DEFAULT_PARAMS
-from gymnasium.core import ObsType
-from gymnasium.spaces import Dict, Discrete
 
 
 class BigKey(Key):
diff --git a/gym_miniworld/envs/simtorealgoto.py b/gym_miniworld/envs/simtorealgoto.py
index 22f3915e..d59865bc 100644
--- a/gym_miniworld/envs/simtorealgoto.py
+++ b/gym_miniworld/envs/simtorealgoto.py
@@ -1,7 +1,8 @@
+from gymnasium import spaces
+
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
 from gym_miniworld.params import DEFAULT_PARAMS
-from gymnasium import spaces
 
 # Simulation parameters
 # These assume a robot about 15cm tall with a pi camera module v2
diff --git a/gym_miniworld/envs/simtorealpush.py b/gym_miniworld/envs/simtorealpush.py
index d4fa3c56..f8bb9976 100644
--- a/gym_miniworld/envs/simtorealpush.py
+++ b/gym_miniworld/envs/simtorealpush.py
@@ -1,10 +1,11 @@
 import math
 
 import numpy as np
+from gymnasium import spaces
+
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
 from gym_miniworld.params import DEFAULT_PARAMS
-from gymnasium import spaces
 
 # Simulation parameters
 # These assume a robot about 15cm tall with a pi camera module v2
diff --git a/gym_miniworld/envs/threerooms.py b/gym_miniworld/envs/threerooms.py
index edd59cb9..8dbb5859 100644
--- a/gym_miniworld/envs/threerooms.py
+++ b/gym_miniworld/envs/threerooms.py
@@ -1,8 +1,9 @@
 import math
 
+from gymnasium import spaces
+
 from gym_miniworld.entity import Ball, Box, ImageFrame, Key, MeshEnt
 from gym_miniworld.miniworld import MiniWorldEnv
-from gymnasium import spaces
 
 
 class ThreeRooms(MiniWorldEnv):
diff --git a/gym_miniworld/envs/tmaze.py b/gym_miniworld/envs/tmaze.py
index 3741da6e..31d8f758 100644
--- a/gym_miniworld/envs/tmaze.py
+++ b/gym_miniworld/envs/tmaze.py
@@ -1,8 +1,9 @@
 import math
 
+from gymnasium import spaces
+
 from gym_miniworld.entity import Box
 from gym_miniworld.miniworld import MiniWorldEnv
-from gymnasium import spaces
 
 
 class TMaze(MiniWorldEnv):
diff --git a/gym_miniworld/envs/wallgap.py b/gym_miniworld/envs/wallgap.py
index a95c373d..6d953fba 100644
--- a/gym_miniworld/envs/wallgap.py
+++ b/gym_miniworld/envs/wallgap.py
@@ -1,9 +1,10 @@
 import math
 
 import numpy as np
+from gymnasium import spaces
+
 from gym_miniworld.entity import Box, MeshEnt
 from gym_miniworld.miniworld import MiniWorldEnv
-from gymnasium import spaces
 
 
 class WallGap(MiniWorldEnv):
diff --git a/gym_miniworld/envs/ymaze.py b/gym_miniworld/envs/ymaze.py
index 01eb9278..9b187843 100644
--- a/gym_miniworld/envs/ymaze.py
+++ b/gym_miniworld/envs/ymaze.py
@@ -1,10 +1,11 @@
 import math
 
 import numpy as np
+from gymnasium import spaces
+
 from gym_miniworld.entity import Box
 from gym_miniworld.math import gen_rot_matrix
 from gym_miniworld.miniworld import MiniWorldEnv
-from gymnasium import spaces
 
 
 class YMaze(MiniWorldEnv):
diff --git a/scripts/benchmark.py b/scripts/benchmark.py
index 5c63c497..17e96e84 100755
--- a/scripts/benchmark.py
+++ b/scripts/benchmark.py
@@ -2,9 +2,10 @@
 
 import time
 
-import gym_miniworld
 import gymnasium as gym
 
+# import gym_miniworld
+
 # Benchmark loading time
 st = time.time()
 env = gym.make("MiniWorld-Maze-v0")
diff --git a/scripts/manual_control.py b/scripts/manual_control.py
index 712eaf15..a04ed9fd 100755
--- a/scripts/manual_control.py
+++ b/scripts/manual_control.py
@@ -7,13 +7,16 @@
 
 import argparse
 import math
-import sys
 
-import gym_miniworld
 import gymnasium as gym
 import pyglet
 from pyglet.window import key
 
+import gym_miniworld
+
+# import sys
+
+
 parser = argparse.ArgumentParser()
 parser.add_argument("--env-name", default=gym_miniworld.envs.env_ids[0])
 parser.add_argument(

From 01473a885fb6cc9b033c4e0ef2c3d2387045e44f Mon Sep 17 00:00:00 2001
From: Bolun Dai <bd1555@nyu.edu>
Date: Thu, 3 Nov 2022 14:16:05 -0400
Subject: [PATCH 7/8] push test

---
 README.md | 1 +
 1 file changed, 1 insertion(+)

diff --git a/README.md b/README.md
index a8e71235..a831e10c 100644
--- a/README.md
+++ b/README.md
@@ -118,3 +118,4 @@ Alternatively, if this doesn't work, you can also try running MiniWorld with `xv
 ```
 xvfb-run -a -s "-screen 0 1024x768x24 -ac +extension GLX +render -noreset" python3 your_script.py
 ```
+

From 6d747e3ca0ee1793c241650d0a1dbe1fa04fa6c8 Mon Sep 17 00:00:00 2001
From: Bolun Dai <bd1555@nyu.edu>
Date: Thu, 3 Nov 2022 14:31:33 -0400
Subject: [PATCH 8/8] updated setup.py

---
 setup.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/setup.py b/setup.py
index edec8a4e..87df9f74 100644
--- a/setup.py
+++ b/setup.py
@@ -59,7 +59,7 @@ def get_version():
     extras_require={"testing": ["pytest==7.0.1", "torch"]},
     install_requires=[
         "numpy>=1.18.0",
-        "pyglet>=1.5.11",
+        "pyglet==1.5.27",
         "gymnasium>=0.26.2",
     ],
     # Include textures and meshes in the package