add: IQL Example

[JemmaLDaniel]

What?

• An item buffer example: independent Q-learning in Matrax.

Why?

• To provide another user-friendly implementation showcasing the potential of the flashbax item buffer.

[CLAassistant]

All committers have signed the CLA.

[callumtilbury]

Still working through the noteobook, but here are some comments in the interim.

In addition to the attached comments, also:

• Re: naming convention, we do sometimes use ALL_CAPS for constants, but not for other things. e.g. Q_NETWORK can just be q_network, etc.
• Could we rename the notebook itself to match the pattern seen elsewhere?—e.g. quickstart_iql_matrax.ipynb or similar.

It's looking fab though! Great comments, and clear structure. Will continue looking soon :)

[callumtilbury]

Great job, @JemmaLDaniel!