bprabhakar/upside-down-reinforcement-learning

Pytorch based implementation of Upside Down Reinforcement Learning (UDRL) by J. Schmidhuber et al.

28
/ 100
Experimental

This project helps machine learning researchers and practitioners explore an alternative approach to training AI agents for tasks where actions are rewarded after a sequence of steps. It takes in a simulated environment and produces a trained agent capable of completing specific goals, like landing a spacecraft. This is for anyone researching or implementing advanced AI decision-making systems.

No commits in the last 6 months.

Use this if you are a reinforcement learning researcher or practitioner interested in experimenting with a novel, supervised learning-based method for agent training on episodic tasks.

Not ideal if you are looking for a plug-and-play solution for common reinforcement learning problems without delving into experimental algorithm implementations.

reinforcement-learning ai-agent-training machine-learning-research episodic-tasks supervised-learning
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 5 / 25
Maturity 16 / 25
Community 7 / 25

How are scores calculated?

Stars

11

Forks

1

Language

Jupyter Notebook

License

MIT

Last pushed

May 01, 2020

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/bprabhakar/upside-down-reinforcement-learning"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.