Closed Chiniklas closed 1 year ago
Hello @Chiniklas ,
Thank you for your pull request! I have the following remarks and questions regarding your submission:
.idea and .vscode Are these folders needed for something in this repository? If not please delete them from the pull request.
Dockerfile Please add the python dependency shimmy to the src/python/setup.py file. Please change the line
RUN git clone https://github.com/Chiniklas/double_pendulum.git
to the main repository instead of your fork. Please comment the line
RUN git checkout v0.1.0
policy files Consider placing the policy files, i.e.
examples/reinforcement_learning/SAC/best_model/acrobot_model.zip
examples/reinforcement_learning/SAC/best_model/pendubot_model.zip
leaderboard/robustness/acrobot/SAC/sac_model.zip
leaderboard/robustness/pendubot/SAC/sac_model.zip
leaderboard/simulation/acrobot/SAC/sac_model.zip
leaderboard/simulation/pendubot/SAC/sac_model.zip
src/python/double_pendulum/controller/SAC/sac_training/best_model/best_model.zip
in the main data folder. I would suggest the structure
data/policies/<design_X.X>/<model_X.X>/<robot>/SAC/<your_policy>
similar to the stored trajectories in the same data folder. For the naming of design and model see here. <robot>
would be either 'acrobot' or 'pendubot'. If you have multiple policies for the same design, model, robot and method you can create subfolders on the last level.
examples/reinforcement_learning/SAC/check_state.py
The script returns an error because the load_path =data/pendubot/lqr/roa/rho
in line 5 points to the folder data
instead of lqr_data
.
examples/reinforcement_learning/SAC/sac_training I think, this folder contains you training attempts. These do not need to be in this examples folder. Please remove them from the pull request.
experiments/tmotors_donothing.py You changed a few minor things in this experiment file which are not related to your controller. Please undo these changes.
leaderboard controller files
The SACController
class is redefined in every con_sac_lqr.py
file. Is there a reason for this or could the controller also be imported from src/python/double_pendulum/controller/SAC/SAC_controller.py?
leaderboard data The leaderboard data does not need to be commited. The leaderboard results will be computed with a dedicated workflow and stored in a separate repository dedicated to the leaderboard. Please remove the result files
leaderboard/robustness/acrobot/data/*
leaderboard/simulation/acrobot/data/*
leaderboard/robustness/pendubot/data/*
leaderboard/simulation/pendubot/data/*
leaderboard/simulation/pendubot/data_loc/*
from the pull request. I am sorry, that was probably not so clear in the competition/repository rules.
leaderboard/robustness/pendubot/plot.py Do you want this plotting script to be here? Otherwise, I would propose to remove it.
leaderboard/simulation/acrobot and leaderboard/simulation/pendubot Do you want the files
leaderboard/simulation/acrobot/create_urdf.py
leaderboard/simulation/acrobot/dev_con_ilqrfree_lqr.py
leaderboard/simulation/acrobot/timetest.py
leaderboard/simulation/pendubot/create_urdf.py
here? Otherwise, I would propose to remove them.
src/python/double_pendulum
The src folder is intended only for the reusable classes and functions which will be installed with this python package.
In src/python/double_pendulum/controller/SAC there is a plots
and a sac_training
folder containing data of your training, I assume. Please either remove these folders or if they are needed move the data somewhere else in the repository, such as the data folder in the root directory. If there is data you want to submit and you are unsure where to put it feel free to contact me.
Thanks again for submitting a cool controller!
@Chiniklas Thank you for addressing the remarks, I merged the PR into the main repository
Hello,
A new policy-based controller is submitted, which is SAC controller based on Stable Baseline3 library.
Chi Zhang