Zhilong Zheng

icon location PhD student, iDLab, School of Vehicle and Mobility, Tsinghua University. Reinforcement Learning.

Results 2 issues of Zhilong Zheng

a question regarding RecurrentPPO

2 comment

Hi, I'm using RecurrentPPO to train an RecurrentActorCriticPolicy. I noticed that when collecting rollouts data, the hidden states in LSTM at each time steps are also storaged in the rollout...

question

REFACTOR: recieve algo params throught init instead of set_parameters