Difference between revisions of "K-Scale Manipulation Suite"

From Humanoid Robots Wiki
Jump to: navigation, search
 
(4 intermediate revisions by the same user not shown)
Line 1: Line 1:
 
== Setup - Linux ==
 
== Setup - Linux ==
 +
 +
[https://github.com/kscalelabs/gym-kmanip https://github.com/kscalelabs/gym-kmanip]
 +
 
'''Clone and install dependencies'''
 
'''Clone and install dependencies'''
 
<pre>
 
<pre>
Line 19: Line 22:
 
git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip
 
git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip
 
pip install -e .
 
pip install -e .
 +
</pre>
 +
 +
== Usage - Basic ==
 +
'''Visualize the MuJoCo scene'''
 +
<pre>
 +
python gym_kmanip/examples/1_view_env.py
 +
</pre>
 +
'''Record a video of the MuJoCo scene'''
 +
<pre>
 +
python gym_kmanip/examples/2_record_video.py
 +
</pre>
 +
 +
== Usage - Recording Data ==
 +
'''K-Scale HuggingFace Datasets'''
 +
'''Data is recorded via teleop, this requires additional dependencies'''
 +
<pre>
 +
pip install opencv-python==4.9.0.80
 +
pip install vuer==0.0.30
 +
pip install rerun-sdk==0.16.0
 +
</pre>
 +
'''Start the server on the robot computer'''
 +
<pre>
 +
python gym_kmanip/examples/4_record_data_teleop.py
 +
</pre>
 +
'''Start ngrok on the robot computer'''
 +
<pre>
 +
ngrok http 8012
 +
</pre>
 +
'''Open the browser app on the VR headset and go to the ngrok URL'''
 +
 +
 +
== Usage - Visualizing Data ==
 +
'''Data is visualized using rerun'''
 +
<pre>
 +
rerun gym_kmanip/data/test.rrd
 +
</pre>
 +
 +
== Usage - MuJoCo Sim Visualizer ==
 +
'''MuJoCo provides a nice visualizer where you can directly control the robot'''
 +
'''Download standalone MuJoCo'''
 +
<pre>
 +
tar -xzf ~/Downloads/mujoco-3.1.5-linux-x86_64.tar.gz -C /path/to/mujoco-3.1.5
 +
</pre>
 +
'''Run the simulator'''
 +
<pre>
 +
/path/to/mujoco-3.1.5/bin/simulate gym_kmanip/assets/_env_solo_arm.xml
 +
</pre>
 +
 +
== Citation ==
 +
<pre>
 +
@misc{teleop-2024,
 +
  title={gym-kmanip},
 +
  author={Hugo Ponte},
 +
  year={2024},
 +
  url={https://github.com/kscalelabs/gym-kmanip}
 +
}
 
</pre>
 
</pre>

Latest revision as of 20:51, 22 May 2024

Setup - Linux[edit]

https://github.com/kscalelabs/gym-kmanip

Clone and install dependencies

git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip
conda create -y -n gym-kmanip python=3.10 && conda activate gym-kmanip
pip install -e .

Run tests

pip install pytest
pytest tests/test_env.py


Setup - Jetson Orin AGX[edit]

No conda on ARM64, install on bare metal

sudo apt-get install libhdf5-dev
git clone https://github.com/kscalelabs/gym-kmanip.git && cd gym-kmanip
pip install -e .

Usage - Basic[edit]

Visualize the MuJoCo scene

python gym_kmanip/examples/1_view_env.py

Record a video of the MuJoCo scene

python gym_kmanip/examples/2_record_video.py

Usage - Recording Data[edit]

K-Scale HuggingFace Datasets Data is recorded via teleop, this requires additional dependencies

pip install opencv-python==4.9.0.80
pip install vuer==0.0.30
pip install rerun-sdk==0.16.0

Start the server on the robot computer

python gym_kmanip/examples/4_record_data_teleop.py

Start ngrok on the robot computer

ngrok http 8012

Open the browser app on the VR headset and go to the ngrok URL


Usage - Visualizing Data[edit]

Data is visualized using rerun

rerun gym_kmanip/data/test.rrd

Usage - MuJoCo Sim Visualizer[edit]

MuJoCo provides a nice visualizer where you can directly control the robot Download standalone MuJoCo

tar -xzf ~/Downloads/mujoco-3.1.5-linux-x86_64.tar.gz -C /path/to/mujoco-3.1.5

Run the simulator

/path/to/mujoco-3.1.5/bin/simulate gym_kmanip/assets/_env_solo_arm.xml

Citation[edit]

@misc{teleop-2024,
  title={gym-kmanip},
  author={Hugo Ponte},
  year={2024},
  url={https://github.com/kscalelabs/gym-kmanip}
}