Upgrade to Gymnasium interface

[alexander-zap]

Proposed change(s)

I adjusted the Unity Gym Wrapper to support the gymnasium interface (gymnasium>=0.25).

This includes:

• returning terminated, truncated on step
• returning info on reset
• accept seed and options on reset
• not accept mode on render

Also I relaxed the dependencies for pettingzoo, numpy and Python for mlagents_env.
They didn't seem required and, at least personally for me, are blocking the usage of mlagents_env independently (e.g., if wishing to connect to other training frameworks).

Useful links (Github issues, JIRA tickets, ML-Agents forum threads etc.)

This also could be helpful for: #5890

Types of change(s)

• Bug fix
• [ X ] New feature
• Code refactor
• [ X ] Breaking change
• [ X ] Documentation update
• Other (please describe)

Checklist

• Added tests that prove my fix is effective or that my feature works
• Updated the changelog (if applicable)
• Updated the documentation (if applicable)
• Updated the migration guide (if applicable)

Other comments

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.

---

a_zap seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[alexander-zap]

#5890

[alexander-zap]

@elliottower Since this feature should partially solve your issue, is there anything I missed?