Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Added MS Ant Walk and Run Tasks #533

Merged
merged 5 commits into from
Aug 29, 2024
Merged

Added MS Ant Walk and Run Tasks #533

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
d50fa0b
ant walk and run ppo trainable
Xander-Hinrichsen Aug 29, 2024
211df15
added to docs
Xander-Hinrichsen Aug 29, 2024
fce76b0
formatting
Xander-Hinrichsen Aug 29, 2024
3dcbffb
removed unnecessary lines
Xander-Hinrichsen Aug 29, 2024
3ef489b
added rgb only obs examples
Xander-Hinrichsen Aug 29, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
38 changes: 38 additions & 0 deletions docs/source/tasks/control/index.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -89,6 +89,44 @@ Hopper robot stands upright
- No specific success conditions. We can threshold the episode accumulated reward to determine success.
:::

## MS-AntWalk-v1
![dense-reward][reward-badge]

:::{dropdown} Task Card
:icon: note
:color: primary

**Task Description:**
Ant moves in x direction at 0.5 m/s

**Supported Robots: ant**

**Randomizations:**
- Ant qpos and qvel have added noise from uniform distribution [-1e-2, 1e-2]

**Success Conditions:**
- No specific success conditions. The task is considered successful if the ant walks at 0.5 m/s for the whole episode. We can threshold the episode accumulated reward to determine success.
:::

## MS-AntRun-v1
![dense-reward][reward-badge]

:::{dropdown} Task Card
:icon: note
:color: primary

**Task Description:**
Ant moves in x direction at 4 m/s

**Supported Robots: ant**

**Randomizations:**
- Ant qpos and qvel have added noise from uniform distribution [-1e-2, 1e-2]

**Success Conditions:**
- No specific success conditions. The task is considered successful the ant walks at 4 m/s for the whole episode. We can threshold the episode accumulated reward to determine success.
:::

## MS-HumanoidStand-v1
![dense-reward][reward-badge]

Expand Down
30 changes: 22 additions & 8 deletions examples/baselines/ppo/examples.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -49,17 +49,21 @@ python ppo.py --env_id="MS-CartpoleSwingUp-v1" \
--total_timesteps=10_000_000 --num-steps=250 --num-eval-steps=1000 \
--gamma=0.99 --gae_lambda=0.95 \
--eval_freq=5
python ppo.py --env_id="MS-HumanoidStand-v1" --num_envs=2048 \
python ppo.py --env_id="MS-AntWalk-v1" --num_envs=2048 --eval_freq=10 \
--update_epochs=8 --num_minibatches=32 --total_timesteps=20_000_000 \
--num_eval_steps=1000 --num_steps=200 --gamma=0.97 --ent_coef=1e-3
python ppo.py --env_id="MS-AntRun-v1" --num_envs=2048 --eval_freq=10 \
--update_epochs=8 --num_minibatches=32 --total_timesteps=20_000_000 \
--num_eval_steps=1000 --num_steps=200 --gamma=0.97 --ent_coef=1e-3
python ppo.py --env_id="MS-HumanoidStand-v1" --num_envs=2048 --eval_freq=10 \
--update_epochs=8 --num_minibatches=32 --total_timesteps=40_000_000 \
--eval_freq=10 --num_eval_steps=1000 --num_steps=200 --gamma=0.95
python ppo.py --env_id="MS-HumanoidWalk-v1" --num_envs=2048 \
--num_eval_steps=1000 --num_steps=200 --gamma=0.95
python ppo.py --env_id="MS-HumanoidWalk-v1" --num_envs=2048 --eval_freq=10 \
--update_epochs=8 --num_minibatches=32 --total_timesteps=80_000_000 \
--eval_freq=10 --num_eval_steps=1000 --num_steps=200 --gamma=0.97 \
--ent_coef=1e-3
python ppo.py --env_id="MS-HumanoidRun-v1" --num_envs=2048 \
--num_eval_steps=1000 --num_steps=200 --gamma=0.97 --ent_coef=1e-3
python ppo.py --env_id="MS-HumanoidRun-v1" --num_envs=2048 --eval_freq=10 \
--update_epochs=8 --num_minibatches=32 --total_timesteps=60_000_000 \
--eval_freq=10 --num_eval_steps=1000 --num_steps=200 --gamma=0.97 \
--ent_coef=1e-3
--num_eval_steps=1000 --num_steps=200 --gamma=0.97 --ent_coef=1e-3
python ppo.py --env_id="UnitreeG1PlaceAppleInBowl-v1" \
--num_envs=512 --update_epochs=8 --num_minibatches=32 \
--total_timesteps=50_000_000 --num-steps=100 --num-eval-steps=100
Expand Down Expand Up @@ -109,6 +113,16 @@ python ppo_rgb.py --env_id="PickSingleYCB-v1" \
python ppo_rgb.py --env_id="PushT-v1" \
--num_envs=256 --update_epochs=8 --num_minibatches=8 \
--total_timesteps=25_000_000 --num-steps=100 --num_eval_steps=100 --gamma=0.99
python ppo_rgb.py --env_id="MS-AntWalk-v1" \
--num_envs=256 --update_epochs=8 --num_minibatches=32 \
--total_timesteps=5_000_000 --eval_freq=15 --num_eval_steps=1000 \
--num_steps=200 --gamma=0.97 --no-include-state --render_mode="rgb_array" \
--ent_coef=1e-3
python ppo_rgb.py --env_id="MS-AntRun-v1" \
--num_envs=256 --update_epochs=8 --num_minibatches=32 \
--total_timesteps=15_000_000 --eval_freq=15 --num_eval_steps=1000 \
--num_steps=200 --gamma=0.97 --no-include-state --render_mode="rgb_array" \
--ent_coef=1e-3
python ppo_rgb.py --env_id="MS-HumanoidRun-v1" \
--num_envs=256 --update_epochs=8 --num_minibatches=32 \
--total_timesteps=80_000_000 --eval_freq=15 --num_eval_steps=1000 \
Expand Down
1 change: 1 addition & 0 deletions mani_skill/envs/tasks/control/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,3 +1,4 @@
from .cartpole import CartpoleBalanceEnv, CartpoleSwingUpEnv
from .hopper import HopperHopEnv, HopperStandEnv
from .humanoid import HumanoidRun, HumanoidStand, HumanoidWalk
from .ant import AntWalk, AntRun
Loading