Add extra tasks used in E3B paper

minihack/envs/minigrid.py

@@ -155,6 +155,37 @@ def __init__(self, *args, **kwargs):
         super().__init__(*args, env_name="MiniGrid-MultiRoom-N6-v0", **kwargs)
 
 
+register(
+    id="MiniGrid-MultiRoom-N10-v0",
+    entry_point="gym_minigrid.envs:MultiRoomEnv",
+    kwargs={"minNumRooms": 10, "maxNumRooms": 10},
+)
+
+
+class MiniHackMultiRoomN10(MiniGridHack):
+    def __init__(self, *args, **kwargs):
+        kwargs["max_episode_steps"] = kwargs.pop("max_episode_steps", 360)
+        super().__init__(*args, env_name="MiniGrid-MultiRoom-N10-v0", **kwargs)
+
+
+# open door version of the above envs - methods with message-based bonuses tend to fail on these
+# because there is no message "the door opens" when visiting a new room
+
+
+class MiniHackMultiRoomN6OpenDoor(MiniGridHack):
+    def __init__(self, *args, **kwargs):
+        kwargs["max_episode_steps"] = kwargs.pop("max_episode_steps", 240)
+        kwargs["door_state"] = kwargs.pop("door_state", "open")
+        super().__init__(*args, env_name="MiniGrid-MultiRoom-N6-v0", **kwargs)
+
+
+class MiniHackMultiRoomN10OpenDoor(MiniGridHack):
+    def __init__(self, *args, **kwargs):
+        kwargs["max_episode_steps"] = kwargs.pop("max_episode_steps", 360)
+        kwargs["door_state"] = kwargs.pop("door_state", "open")
+        super().__init__(*args, env_name="MiniGrid-MultiRoom-N10-v0", **kwargs)
+
+
 register(
     id="MiniHack-MultiRoom-N2-v0",
     entry_point="minihack.envs.minigrid:MiniHackMultiRoomN2",
@@ -167,6 +198,18 @@ def __init__(self, *args, **kwargs):
     id="MiniHack-MultiRoom-N6-v0",
     entry_point="minihack.envs.minigrid:MiniHackMultiRoomN6",
 )
+register(
+    id="MiniHack-MultiRoom-N10-v0",
+    entry_point="minihack.envs.minigrid:MiniHackMultiRoomN10",
+)
+register(
+    id="MiniHack-MultiRoom-N6-OpenDoor-v0",
+    entry_point="minihack.envs.minigrid:MiniHackMultiRoomN6OpenDoor",
+)
+register(
+    id="MiniHack-MultiRoom-N10-OpenDoor-v0",
+    entry_point="minihack.envs.minigrid:MiniHackMultiRoomN10OpenDoor",
+)
 
 
 # MiniGrid: LockedMultiRoom
@@ -235,6 +278,29 @@ def __init__(self, *args, **kwargs):
         super().__init__(*args, env_name="MiniGrid-MultiRoom-N6-v0", **kwargs)
 
 
+class MiniHackMultiRoomN10Lava(MiniGridHack):
+    def __init__(self, *args, **kwargs):
+        kwargs["max_episode_steps"] = kwargs.pop("max_episode_steps", 360)
+        kwargs["lava_walls"] = True
+        super().__init__(*args, env_name="MiniGrid-MultiRoom-N10-v0", **kwargs)
+
+
+class MiniHackMultiRoomN6LavaOpenDoor(MiniGridHack):
+    def __init__(self, *args, **kwargs):
+        kwargs["max_episode_steps"] = kwargs.pop("max_episode_steps", 240)
+        kwargs["lava_walls"] = True
+        kwargs["door_state"] = kwargs.pop("door_state", "open")
+        super().__init__(*args, env_name="MiniGrid-MultiRoom-N6-v0", **kwargs)
+
+
+class MiniHackMultiRoomN10LavaOpenDoor(MiniGridHack):
+    def __init__(self, *args, **kwargs):
+        kwargs["max_episode_steps"] = kwargs.pop("max_episode_steps", 360)
+        kwargs["lava_walls"] = True
+        kwargs["door_state"] = kwargs.pop("door_state", "open")
+        super().__init__(*args, env_name="MiniGrid-MultiRoom-N10-v0", **kwargs)
+
+
 register(
     id="MiniHack-MultiRoom-N2-Lava-v0",
     entry_point="minihack.envs.minigrid:MiniHackMultiRoomN2Lava",
@@ -247,6 +313,18 @@ def __init__(self, *args, **kwargs):
     id="MiniHack-MultiRoom-N6-Lava-v0",
     entry_point="minihack.envs.minigrid:MiniHackMultiRoomN6Lava",
 )
+register(
+    id="MiniHack-MultiRoom-N10-Lava-v0",
+    entry_point="minihack.envs.minigrid:MiniHackMultiRoomN10Lava",
+)
+register(
+    id="MiniHack-MultiRoom-N6-Lava-OpenDoor-v0",
+    entry_point="minihack.envs.minigrid:MiniHackMultiRoomN6LavaOpenDoor",
+)
+register(
+    id="MiniHack-MultiRoom-N10-Lava-OpenDoor-v0",
+    entry_point="minihack.envs.minigrid:MiniHackMultiRoomN10LavaOpenDoor",
+)
 
 
 # MiniGrid: MonsterpedMultiRoom
@@ -335,7 +413,66 @@ def __init__(self, *args, **kwargs):
     entry_point="minihack.envs.minigrid:MiniHackMultiRoomN6Extreme",
 )
 
+
+# Note: the N6-Extreme-v0 env above is impossible to solve consistently even for a human
+# due to too many monsters. Here are some easier envs with lava and fewer monsters.
+
+
+class MiniHackMultiRoomN2LavaMonsters(MiniGridHack):
+    def __init__(self, *args, **kwargs):
+        kwargs["max_episode_steps"] = kwargs.pop("max_episode_steps", 40)
+        kwargs["num_mon"] = 1
+        kwargs["lava_walls"] = True
+        super().__init__(
+            *args, env_name="MiniGrid-MultiRoom-N2-S4-v0", **kwargs
+        )
+
+
+class MiniHackMultiRoomN4LavaMonsters(MiniGridHack):
+    def __init__(self, *args, **kwargs):
+        kwargs["max_episode_steps"] = kwargs.pop("max_episode_steps", 120)
+        kwargs["num_mon"] = 2
+        kwargs["lava_walls"] = True
+        super().__init__(
+            *args, env_name="MiniGrid-MultiRoom-N4-S5-v0", **kwargs
+        )
+
+
+class MiniHackMultiRoomN6LavaMonsters(MiniGridHack):
+    def __init__(self, *args, **kwargs):
+        kwargs["max_episode_steps"] = kwargs.pop("max_episode_steps", 240)
+        kwargs["num_mon"] = 3
+        kwargs["lava_walls"] = True
+        super().__init__(*args, env_name="MiniGrid-MultiRoom-N6-v0", **kwargs)
+
+
+register(
+    id="MiniHack-MultiRoom-N2-LavaMonsters-v0",
+    entry_point="minihack.envs.minigrid:MiniHackMultiRoomN2LavaMonsters",
+)
+register(
+    id="MiniHack-MultiRoom-N4-LavaMonsters-v0",
+    entry_point="minihack.envs.minigrid:MiniHackMultiRoomN4LavaMonsters",
+)
+register(
+    id="MiniHack-MultiRoom-N6-LavaMonsters-v0",
+    entry_point="minihack.envs.minigrid:MiniHackMultiRoomN6LavaMonsters",
+)
+
+
 # MiniGrid: LavaCrossing
+register(
+    id="MiniGrid-LavaCrossingS19N13-v0",
+    entry_point="gym_minigrid.envs:CrossingEnv",
+    kwargs={"size": 19, "num_crossings": 13},
+)
+register(
+    id="MiniGrid-LavaCrossingS19N17-v0",
+    entry_point="gym_minigrid.envs:CrossingEnv",
+    kwargs={"size": 19, "num_crossings": 17},
+)
+
+
 register(
     id="MiniHack-LavaCrossingS9N1-v0",
     entry_point="minihack.envs.minigrid:MiniGridHack",
@@ -356,6 +493,18 @@ def __init__(self, *args, **kwargs):
     entry_point="minihack.envs.minigrid:MiniGridHack",
     kwargs={"env_name": "MiniGrid-LavaCrossingS11N5-v0"},
 )
+register(
+    id="MiniHack-LavaCrossingS19N13-v0",
+    entry_point="minihack.envs.minigrid:MiniGridHack",
+    kwargs={"env_name": "MiniGrid-LavaCrossingS19N13-v0"},
+)
+
+register(
+    id="MiniHack-LavaCrossingS19N17-v0",
+    entry_point="minihack.envs.minigrid:MiniGridHack",
+    kwargs={"env_name": "MiniGrid-LavaCrossingS19N17-v0"},
+)
+
 
 # MiniGrid: Simple Crossing
 register(

minihack/envs/skills_freeze.py

@@ -1,6 +1,7 @@
 # Copyright (c) Facebook, Inc. and its affiliates.
 from minihack import MiniHackSkill, LevelGenerator, RewardManager
 from minihack.envs import register
+from nle import nethack
 
 freeze_msgs = [
     "The bolt of cold bounces!",  # checks if cold bounces from the wall
@@ -26,6 +27,22 @@ def __init__(self, *args, **kwargs):
         super().__init__(*args, des_file=des_file, **kwargs)
 
 
+class MiniHackFreezeWandRestrictedActions(MiniHackFreeze):
+    def __init__(self, *args, **kwargs):
+        lvl_gen = LevelGenerator(w=8, h=8, lit=True)
+        lvl_gen.add_object("cold", "/", cursestate="blessed")
+        des_file = lvl_gen.get_des()
+
+        ACTIONS = tuple(nethack.CompassDirection) + (
+            nethack.Command.PICKUP,
+            nethack.Command.ZAP,
+            nethack.Command.FIRE,
+        )
+        kwargs["actions"] = ACTIONS
+
+        super().__init__(*args, des_file=des_file, **kwargs)
+
+
 class MiniHackFreezeHorn(MiniHackFreeze):
     def __init__(self, *args, **kwargs):
         lvl_gen = LevelGenerator(w=8, h=8, lit=True)
@@ -35,6 +52,22 @@ def __init__(self, *args, **kwargs):
         super().__init__(*args, des_file=des_file, **kwargs)
 
 
+class MiniHackFreezeHornRestrictedActions(MiniHackFreeze):
+    def __init__(self, *args, **kwargs):
+        lvl_gen = LevelGenerator(w=8, h=8, lit=True)
+        lvl_gen.add_object("frost horn", "(", cursestate="blessed")
+        des_file = lvl_gen.get_des()
+
+        ACTIONS = tuple(nethack.CompassDirection) + (
+            nethack.Command.PICKUP,
+            nethack.Command.APPLY,
+            nethack.Command.FIRE,
+        )
+        kwargs["actions"] = ACTIONS
+
+        super().__init__(*args, des_file=des_file, **kwargs)
+
+
 class MiniHackFreezeRandom(MiniHackFreeze):
     def __init__(self, *args, **kwargs):
         des_file = """
@@ -63,6 +96,43 @@ def __init__(self, *args, **kwargs):
         super().__init__(*args, des_file=des_file, **kwargs)
 
 
+class MiniHackFreezeRandomRestrictedActions(MiniHackFreeze):
+    def __init__(self, *args, **kwargs):
+        des_file = """
+MAZE: "mylevel", ' '
+FLAGS:hardfloor
+MESSAGE: "Welcome to MiniHack!"
+INIT_MAP: solidfill,' '
+GEOMETRY:center,center
+MAP
+........
+........
+........
+........
+........
+........
+........
+........
+ENDMAP
+REGION:(0,0,7,7),lit,"ordinary"
+IF [50%] {
+    OBJECT:('/',"cold"),random,blessed
+} ELSE {
+    OBJECT:('(',"frost horn"),random,blessed
+}
+"""
+
+        ACTIONS = tuple(nethack.CompassDirection) + (
+            nethack.Command.PICKUP,
+            nethack.Command.ZAP,
+            nethack.Command.APPLY,
+            nethack.Command.FIRE,
+        )
+        kwargs["actions"] = ACTIONS
+
+        super().__init__(*args, des_file=des_file, **kwargs)
+
+
 class MiniHackFreezeLava(MiniHackSkill):
     def __init__(self, *args, **kwargs):
         des_file = """
@@ -95,19 +165,75 @@ def __init__(self, *args, **kwargs):
         super().__init__(*args, des_file=des_file, **kwargs)
 
 
+class MiniHackFreezeLavaRestrictedActions(MiniHackSkill):
+    def __init__(self, *args, **kwargs):
+        des_file = """
+MAZE: "mylevel", ' '
+FLAGS:hardfloor
+INIT_MAP: solidfill,' '
+GEOMETRY:center,center
+MAP
+-------------
+|.....L.....|
+|.....L.....|
+|.....L.....|
+|.....L.....|
+|.....L.....|
+-------------
+ENDMAP
+REGION:(0,0,12,6),lit,"ordinary"
+$left_bank = selection:fillrect (1,1,5,5)
+$right_bank = selection:fillrect (7,1,11,5)
+IF [50%] {
+    # wand of cold
+    OBJECT:('/',"cold"),rndcoord($left_bank),blessed
+} ELSE {
+    # frost horn
+    OBJECT:('(',"frost horn"),rndcoord($left_bank),blessed
+}
+BRANCH:(1,1,5,5),(0,0,0,0)
+STAIR:rndcoord($right_bank),down
+"""
+        ACTIONS = tuple(nethack.CompassDirection) + (
+            nethack.Command.PICKUP,
+            nethack.Command.ZAP,
+            nethack.Command.APPLY,
+            nethack.Command.FIRE,
+        )
+        kwargs["actions"] = ACTIONS
+
+        super().__init__(*args, des_file=des_file, **kwargs)
+
+
 register(
-    id="MiniHack-Freeze-Wand-v0",
+    id="MiniHack-Freeze-Wand-Full-v0",
     entry_point="minihack.envs.skills_freeze:MiniHackFreezeWand",
 )
 register(
-    id="MiniHack-Freeze-Horn-v0",
+    id="MiniHack-Freeze-Wand-Restricted-v0",
+    entry_point="minihack.envs.skills_freeze:MiniHackFreezeWandRestrictedActions",
+)
+register(
+    id="MiniHack-Freeze-Horn-Full-v0",
     entry_point="minihack.envs.skills_freeze:MiniHackFreezeHorn",
 )
 register(
-    id="MiniHack-Freeze-Random-v0",
+    id="MiniHack-Freeze-Horn-Restricted-v0",
+    entry_point="minihack.envs.skills_freeze:MiniHackFreezeHornRestrictedActions",
+)
+register(
+    id="MiniHack-Freeze-Random-Full-v0",
     entry_point="minihack.envs.skills_freeze:MiniHackFreezeRandom",
 )
 register(
-    id="MiniHack-Freeze-Lava-v0",
+    id="MiniHack-Freeze-Random-Restricted-v0",
+    entry_point="minihack.envs.skills_freeze:MiniHackFreezeRandomRestrictedActions",
+)
+register(
+    id="MiniHack-Freeze-Lava-Full-v0",
     entry_point="minihack.envs.skills_freeze:MiniHackFreezeLava",
 )
+register(
+    id="MiniHack-Freeze-Lava-Restricted-v0",
+    entry_point="minihack.envs.skills_freeze:MiniHackFreezeLavaRestrictedActions",
+)