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.
game numpy deep-reinforcement-learning openai-gym deep-q-network ddqn prioritized-replay ppo advantage-actor-critic policy-network ddqn-framework mlp-framework hindsight-experience-replay
-
Updated
May 24, 2018 - Jupyter Notebook