Problem about updating the actor network.

[Jian-Yin-Shine]

I read your code to update the actor network, it different from the paper. In the paper it is : $-\frac{1}{batch_size}\sum_{i=1}^{batch_size}Q(s_i, a_i)$ Thank you very much for your answer.

[madvn]

Not sure what you are referring to. I have implemented the algorithm in page 5 here - https://arxiv.org/pdf/1509.02971.pdf