[Feature Request] Support Card-Game Environment

[mjlaali]

Motivation

Card games are useful to study RL algorithms in control settings, with simple interface and simple to grasp.

Solution

Create a wrapper on top of RLCard games for TorchRL. This is 100% python library.

Alternatives

Create a wrapper on top of open_spiel games for TorchRL While this is also a promising library and has implementation in C++ for games. I am not clear on difficulty of adding this library.

Additional context

I don't have a strong data point on these libraries and which one is the best option for TorchRL

Checklist

• [ x ] I have checked that there is no similar issue in the repo (required)

[vmoens]

Related: #2133

I'd be keen on integrate both of these. The only thing I need to get started is an example of a rollout with these libs.

[mjlaali]

Does this help?

https://github.com/google-deepmind/open_spiel/blob/master/open_spiel/python/examples/tic_tac_toe_qlearner.py

It shows how to roll out and train an agent.