Ruan de Kock

Results 9 comments of Ruan de Kock

Hi @davidedomini, thank you for reaching out :) To answer your first question: * `MeanEpisodeReturn` refers to the mean over the returns obtained by each agent for a given episode....

Thank you @bricksdont, that got things going up until it hit this error, https://github.com/masakhane-io/masakhane-mt/issues/166#issue-982924960

Thank for the help @bricksdont. It is now working perfectly, I summary to fix everything: - in the starter notebook the line that installs 1.8.0 explicitly should be removed. -...

A [link to the branch](https://github.com/instadeepai/Mava/tree/fix/recurrent-ppo) with this fix.

An update on this, the suggested fix uses `timestep.first()` to reset the hidden states, but since no timestep will ever be a `timestep.first()` when using Jumanji's auto reset wrapper the...

Hi @UsaidPro, thank you for raising this issue. We will investigate this further over the coming days. For now, what solved the problem for me is to downgrade Flax and...