PFRL is a PyTorch deep reinforcement learning library. Part of my work at Preferred Networks was devoted to developing algorithms and infrastructure for PFRL.

Model Agnostic Meta Learning

A PyTorch implementation of Model Agnostic Meta Learning (MAML) on the Sinusoid task.

Atari Behavioral Cloning

A Behavioral cloning implementation on Atari. This was used in our TREX paper and modeled after DQfD's Behavioral cloning.

Reproducible Deep Q-networks

A high-quality implementation of Human-level control through deep reinforcement learning. This is a deterministic implementation, aimed at reproducibility. This was built with PyTorch.

Bayesian Inverse RL

A basic implementation of Bayesian Inverse Reinforcement Learning.