Alireza Kazemipour

icon indicating a website link alirezakazemipour.github.io icon indicating an email [email protected]

icon company University of Alberta icon location Edmonton, AB icon location MSc in CS

Results 4 comments of Alireza Kazemipour

Why is the reward value so designed？

@DannyLee1991 It is my question too!

RewardForwardFilter to compute intrinsic returns for normalize intrinsic reward

Exactly. :+1: I think they have made a mistake!!! It must have been `self.I.buf_rews_int.T[::-1]` as [**4kasha**](https://github.com/openai/random-network-distillation/issues/16#issuecomment-488387659) has mentioned.

`tasks_per_node=1` does not keep the number of tasks to 1 for the `LocalExecutor`

Hi! Is there any updates on this? Is it solved? The same thing happens using `Slurm launcher` in hydra on clusters for me!

Wrong PPO Model architecture.

@xiaioding The difference is in kernel sizes (`rf`.)