This repository is for CSCI-566 Spring 24 course project 'Multi-Task Reinforcement Learning for Physical Reasoning' - Learning a multi-task Reinforcement Learning agent at USC.

Run new codes (eg. stablebaseline 3 benchmarks) on CREATE environment to train multi-task rl agent.

Run the new-actions-rl repository based code to train multi-task rl agent.

When you merge the pull request, please do the 'Squash and merge' instead of 'Merge pull request'.

You can change the merge type by clicking the green downside-triangular button.

Also, please delete your branch after merging it.