DDPG replication

[sahiliitm]

Hi,

I believe that in DDPG the Value function output is a single scalar and not same as action size. Hence this line in CriticModel.py should be

V = Dense(1,activation='linear')(h3)

Corresponding in ddpg.py the definition of y_t can be changed to

y_t = np.zeros((states.shape[0],1))

Although I'm not sure how this would affect learning, I believe this is the right way to replicate ddpg.

[quhezheng]

I think you are right. Becuase reward is a scaler, which can not server a vector Q value

[LevineYang]

Hi @sahiliitm , I wanna ask did you get the same or better result after changing the two lines here?