deepRL

deep-q-learning for atari games with caffe ##Resource Pool Want to brew RL ? Click me !

Referred impls

deep_q_rl | dqn-in-the-caffe

Prerequisites

caffe required

bash setup.sh （only install the ALE environment)

Hacking features

Double-Q learning | Prioritized Experience Replay | Eligibility trace | Async-rl | A3C

Run

[cpu] bash run_cpu.sh

[gpu] bash run_gpu.sh