Why are Q network and target network the same?

AndreaVidali / Deep-QLearning-Agent-for-Traffic-Signal-Control

A framework where a deep Q-Learning Reinforcement Learning agent tries to choose the correct traffic light phase at an intersection to maximize traffic efficiency.

MIT License

405 stars 146 forks source link

Why are Q network and target network the same? #31

Open QionghuaLiao opened 2 years ago

QionghuaLiao commented 2 years ago

Usually, the Q Network is trained while the parameters of target network are fixed. And every certain steps, the parameters of Q Network will be copied to Target Network. But when I check your code, I find that the Q Network and the Target Network are the same neural network, which confuses me. Could you please help me out?

wxwmd commented 2 years ago

你说的机制是为了减少训练的震荡，这个demo项目就没采用这个机制，直接每步都让q网络更新呗，这有啥confuse的。。。。。。。