Bo Liu

Hi there @samdow , thanks for your quick and detailed feedback. > is this a current bottleneck for your examples? I think I can call it bottleneck in some way,...

> > @vmoens Hey Vincent, I've adapted the maml_torchrl script using lastest torchrl and corrected my own MAML-RL implementation, but the results are still not aligned. Mind taking a look...

> > > > @vmoens Hey Vincent, I've adapted the maml_torchrl script using lastest torchrl and corrected my own MAML-RL implementation, but the results are still not aligned. Mind taking...

> > @vmoens Hey Vincent, I've adapted the maml_torchrl script using lastest torchrl and corrected my own MAML-RL implementation, but the results are still not aligned. Mind taking a look...

@ViktorM Yes, envpool plans to support all tasks in dm_control.locomotion, the multi-agent soccer will be supported too. It can be one of the multi-agent env that envpool supported.

> CHANGELOG added

> Wondering whether there is a road blocker for this? Bazel dependency on OpenGL, also unittest

> Does this mean translating pysc2 to c++? No, actually [SMAC](https://github.com/oxwhirl/smac) is a just python-based multi-agent env based on Blizzard's [StarCraft II Machine Learning API](https://github.com/Blizzard/s2client-proto), which has a [C++](https://github.com/Blizzard/s2client-api) version,...

It does interest me if applicable for incorporating envpool into distributed reinforcement learning

I am afraid that if you want to accelerate the environments in robogym using envpool, you may have to translate the python code in robogym into c++ 🙃 But I...