Chanwoong joo



@kslazarev Hi, I used that config. but only NumEnv is 128 and MaxStepPerEpisode is 4500. In paper, author did not announce Advantage Norm and Noisynet. so I disabled that config.

@kslazarev per_rollout and per_epi is not same scale. per_rollout means just one global update(enter agent.train_model()). but per_epi means Env’s one episode info that is one of parallel env. If one...

@kslazarev I want you to create an issue for each question. :)

@shuang-liu Could you briefly explain why this doesn't work?

My model got 6100 when it run 16k rollout. :)

how many env did you use? I used 128 env

It was also mentioned in the paper that performance was not very good. Why don't you learn more? Reinforcement learning is always a battle of patience. :)