search

TianhongDai / reinforcement-learning-algorithms

This repository contains most of pytorch implementation based classic deep reinforcement learning algorithms, including - DQN, DDQN, Dueling Network, DDPG, SAC, A2C, PPO, TRPO. (More algorithms are still in progress)
664 stars 109 forks source link

the same code？ #3

Closed zhaolongkzz closed 5 years ago

zhaolongkzz commented 5 years ago

Thanks for your code for DRL. But I find the dueling_agent.py of dueling dqn is the same as ddqn_agent.py of double dqn. So where do you implement the dueling dqn algorithm into code?

TianhongDai commented 5 years ago

@zhaolongkzz Hi, the difference between the Double DQN and Dueling DQN is in the network architecture - So, there's no difference between the file dueling_agent.py and ddqn_agent.py. The only difference is in the models.py: https://github.com/TianhongDai/reinforcement-learning-algorithms/blob/master/03-dueling-network/models.py#L31

And also thanks for your issues. i think I should make DQN, Double DQN and Dueling DQN into one code, and they can be selected by using different Flags. Thanks!