dodox/fancy_gym
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

fixed seeding and tests

Browse Source
This commit is contained in:
Fabian
2022-07-12 15:43:46 +02:00
parent 0339361656
commit d64cb614fa
10 changed files with 36 additions and 34 deletions
Download Patch File Download Diff File Expand all files Collapse all files
alr_envs/envs/__init__.py
+10 -10
View File
@@ -154,14 +154,14 @@ register(
)
register(
id='ALRAntJump-v0',
id='AntJump-v0',
entry_point='alr_envs.envs.mujoco:AntJumpEnv',
max_episode_steps=MAX_EPISODE_STEPS_ANTJUMP,
)
register(
id='ALRHalfCheetahJump-v0',
entry_point='alr_envs.envs.mujoco:ALRHalfCheetahJumpEnv',
id='HalfCheetahJump-v0',
entry_point='alr_envs.envs.mujoco:HalfCheetahJumpEnv',
max_episode_steps=MAX_EPISODE_STEPS_HALFCHEETAHJUMP,
)
@@ -173,19 +173,19 @@ register(
register(
id='ALRHopperThrow-v0',
entry_point='alr_envs.envs.mujoco:ALRHopperThrowEnv',
entry_point='alr_envs.envs.mujoco:HopperThrowEnv',
max_episode_steps=MAX_EPISODE_STEPS_HOPPERTHROW,
)
register(
id='ALRHopperThrowInBasket-v0',
entry_point='alr_envs.envs.mujoco:ALRHopperThrowInBasketEnv',
entry_point='alr_envs.envs.mujoco:HopperThrowInBasketEnv',
max_episode_steps=MAX_EPISODE_STEPS_HOPPERTHROWINBASKET,
)
register(
id='ALRWalker2DJump-v0',
entry_point='alr_envs.envs.mujoco:ALRWalker2dJumpEnv',
entry_point='alr_envs.envs.mujoco:Walker2dJumpEnv',
max_episode_steps=MAX_EPISODE_STEPS_WALKERJUMP,
)
@@ -518,7 +518,7 @@ register(
# CtxtFree are v0, Contextual are v1
register(
id='ALRHalfCheetahJump-v0',
entry_point='alr_envs.envs.mujoco:ALRHalfCheetahJumpEnv',
entry_point='alr_envs.envs.mujoco:HalfCheetahJumpEnv',
max_episode_steps=MAX_EPISODE_STEPS_HALFCHEETAHJUMP,
kwargs={
"max_episode_steps": MAX_EPISODE_STEPS_HALFCHEETAHJUMP,
@@ -626,7 +626,7 @@ for i in _vs:
)
register(
id='ALRHopperThrow-v0',
entry_point='alr_envs.envs.mujoco:ALRHopperThrowEnv',
entry_point='alr_envs.envs.mujoco:HopperThrowEnv',
max_episode_steps=MAX_EPISODE_STEPS_HOPPERTHROW,
kwargs={
"max_episode_steps": MAX_EPISODE_STEPS_HOPPERTHROW,
@@ -635,7 +635,7 @@ for i in _vs:
)
register(
id='ALRHopperThrowInBasket-v0',
entry_point='alr_envs.envs.mujoco:ALRHopperThrowInBasketEnv',
entry_point='alr_envs.envs.mujoco:HopperThrowInBasketEnv',
max_episode_steps=MAX_EPISODE_STEPS_HOPPERTHROWINBASKET,
kwargs={
"max_episode_steps": MAX_EPISODE_STEPS_HOPPERTHROWINBASKET,
@@ -644,7 +644,7 @@ for i in _vs:
)
register(
id='ALRWalker2DJump-v0',
entry_point='alr_envs.envs.mujoco:ALRWalker2dJumpEnv',
entry_point='alr_envs.envs.mujoco:Walker2dJumpEnv',
max_episode_steps=MAX_EPISODE_STEPS_WALKERJUMP,
kwargs={
"max_episode_steps": MAX_EPISODE_STEPS_WALKERJUMP,
alr_envs/envs/mujoco/__init__.py
+4 -4
View File
@@ -1,9 +1,9 @@
from .beerpong.beerpong import BeerPongEnv, BeerPongEnvFixedReleaseStep, BeerPongEnvStepBasedEpisodicReward
from .ant_jump.ant_jump import AntJumpEnv
from .half_cheetah_jump.half_cheetah_jump import ALRHalfCheetahJumpEnv
from .half_cheetah_jump.half_cheetah_jump import HalfCheetahJumpEnv
from .hopper_jump.hopper_jump_on_box import HopperJumpOnBoxEnv
from .hopper_throw.hopper_throw import ALRHopperThrowEnv
from .hopper_throw.hopper_throw_in_basket import ALRHopperThrowInBasketEnv
from .hopper_throw.hopper_throw import HopperThrowEnv
from .hopper_throw.hopper_throw_in_basket import HopperThrowInBasketEnv
from .reacher.reacher import ReacherEnv
from .walker_2d_jump.walker_2d_jump import ALRWalker2dJumpEnv
from .walker_2d_jump.walker_2d_jump import Walker2dJumpEnv
from .hopper_jump.hopper_jump import HopperJumpEnv
alr_envs/envs/mujoco/beerpong/deprecated/beerpong.py
+1 -1
View File
@@ -155,7 +155,7 @@ class BeerPongEnvStepBasedEpisodicReward(BeerPongEnv):
return ob, reward, done, infos
# class ALRBeerBongEnvStepBased(ALRBeerBongEnv):
# class BeerBongEnvStepBased(ALRBeerBongEnv):
# def __init__(self, frame_skip=1, apply_gravity_comp=True, noisy=False, rndm_goal=False, cup_goal_pos=None):
# super().__init__(frame_skip, apply_gravity_comp, noisy, rndm_goal, cup_goal_pos)
# self.release_step = 62 # empirically evaluated for frame_skip=2!
alr_envs/envs/mujoco/half_cheetah_jump/half_cheetah_jump.py
+1 -1
View File
@@ -8,7 +8,7 @@ import numpy as np
MAX_EPISODE_STEPS_HALFCHEETAHJUMP = 100
class ALRHalfCheetahJumpEnv(HalfCheetahEnv):
class HalfCheetahJumpEnv(HalfCheetahEnv):
"""
ctrl_cost_weight 0.1 -> 0.0
"""
alr_envs/envs/mujoco/hopper_throw/hopper_throw.py
+2 -2
View File
@@ -7,7 +7,7 @@ import numpy as np
MAX_EPISODE_STEPS_HOPPERTHROW = 250
class ALRHopperThrowEnv(HopperEnv):
class HopperThrowEnv(HopperEnv):
"""
Initialization changes to normal Hopper:
- healthy_reward: 1.0 -> 0.0 -> 0.1
@@ -98,7 +98,7 @@ class ALRHopperThrowEnv(HopperEnv):
if __name__ == '__main__':
render_mode = "human" # "human" or "partial" or "final"
env = ALRHopperThrowEnv()
env = HopperThrowEnv()
obs = env.reset()
for i in range(2000):
alr_envs/envs/mujoco/hopper_throw/hopper_throw_in_basket.py
+2 -2
View File
@@ -8,7 +8,7 @@ import numpy as np
MAX_EPISODE_STEPS_HOPPERTHROWINBASKET = 250
class ALRHopperThrowInBasketEnv(HopperEnv):
class HopperThrowInBasketEnv(HopperEnv):
"""
Initialization changes to normal Hopper:
- healthy_reward: 1.0 -> 0.0
@@ -130,7 +130,7 @@ class ALRHopperThrowInBasketEnv(HopperEnv):
if __name__ == '__main__':
render_mode = "human" # "human" or "partial" or "final"
env = ALRHopperThrowInBasketEnv()
env = HopperThrowInBasketEnv()
obs = env.reset()
for i in range(2000):
alr_envs/envs/mujoco/walker_2d_jump/walker_2d_jump.py
+2 -2
View File
@@ -12,7 +12,7 @@ MAX_EPISODE_STEPS_WALKERJUMP = 300
# as possible, while landing at a specific target position
class ALRWalker2dJumpEnv(Walker2dEnv):
class Walker2dJumpEnv(Walker2dEnv):
"""
healthy reward 1.0 -> 0.005 -> 0.0025 not from alex
penalty 10 -> 0 not from alex
@@ -95,7 +95,7 @@ class ALRWalker2dJumpEnv(Walker2dEnv):
if __name__ == '__main__':
render_mode = "human" # "human" or "partial" or "final"
env = ALRWalker2dJumpEnv()
env = Walker2dJumpEnv()
obs = env.reset()
for i in range(6000):