You need to install Jupyter Notebooks with:

pip install jupyter

or you can jumpstart with ICRA2023 Colab Tutorial

Note: in case the kernel for the environent is not recognized, you can install it with the following commands:

pip install jupyter ipykernel
python -m ipykernel install --user --name= < name of the environment >

You can then remove it with:

jupyter kernelspec uninstall myosuite

You might need also to install ffmpeg to write videos to play them back in the notebooks. You can do it with

conda install conda-forge::ffmpeg

• Get Started
• Load a trained policy and play it *
• Analyse movements from a trained policy *
• Use the DEPRL baseline. For this tutorial, deprl is needed. You can install it with pip install deprl
• Use the MyoReflex baseline. For this tutorial, we provided a wrapper to use MyoReflex together with the tutorial file
• Train a new policy with stable baselines (SB). For this tutorial, stable-baselines3 is needed. You can install it with pip install stable-baselines
• Move single finger of the Hand
• Train policies with SAR. All required installations are included within the SAR tutorial notebook.
• Replicate hand movements with inverse dynamics. For this tutorial, osqp, matplotlib and pandas are needed. You can install it with pip install osqp matplotlib pandas
• Fatigue Modeling
• Inverse Kinematics. This script shows how to perform inverse kinematics using the mink library.
• Computed Muscle Control (CMC)
• Plaback Opensim Mot Files. This tutorial shows how to load OpenSim Mot files and playback them on the MyoSkeleton (follow the instructions prompted by python -m myosuite_init)

*For those tutorial, mjrl and gym==0.13 are needed. You can install them with pip install tabulate matplotlib torch gym==0.13 git+https://github.com/aravindr93/mjrl.git@pvr_beta_1vk