Flux Baselines

Implementations of reinforcement learning algorithms using Flux NOTE: The code works on Julia 0.6.x. Upgrade to 1.0 will be released soon.

Updated to 1.x

Actor-critic

• A2C

DDPG

• DDPG

DQN

• Double DQN
• DQN
• Duel DQN
• Prioritized Replay DQN

Dependencies

• Flux.jl
• OpenAIGym.jl

Implemented Algorithms

• Deep Q Networks [paper]
• Double DQN [paper]
• Dueling Double DQN [paper]
• Prioritized Experience DQN [paper]
• A2C [paper]
• Deep Deterministic Policy Gradients [paper]