MLP-framework (pure numpy) and DDQN-framework for OpenAI's Gym games. +test code for PPO added. +Hindsight Experience Replay(HER) bitflip-DQN example. +prioritized replay.
gamenumpydeep-reinforcement-learningopenai-gymdeep-q-networkddqnprioritized-replayppoadvantage-actor-criticpolicy-networkddqn-frameworkmlp-frameworkhindsight-experience-replay
-
Updated
May 24, 2018 - Jupyter Notebook