Skip to content

alvaroprat97/MetaRobot

Repository files navigation

Probabilistic Embeddings for hybrid meta-Reinforcement and Imitation Learning

Requirements

This project requires Python v3.6.x and pip3 to run.

Setup

Requirements:

  • Python 3
  • pip3
  1. Install dependencies

Dependency management is done through pip.

pip3 install -r requirements.txt

Conditions

Before running PERIL, you must select the task families you want to meta-train on. Access this in experiment_demos.

Configurations for meta-training and meta-testing can be modified in Configs/default_configs

Available Scripts

From the root directory, you can run:

python3 -m experiment_demos to run the PERIL training pipeline.

python3 -m demos_tester to test the algorithm on unseeen tasks.

Documents

Please read our works on PERIL in the provided pdf document (XXXX). This document will be available after assessment of the project.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages