GRPO refactoring

[mydatascience]

Description

Refactoring how to run GRPO with Tunix MaxText vLLM and Pathways (TMVP). We are adding the new files under src/MaxText/rl and the config file rl.yml in src/MaxText/configs. The entry level main file is src/MaxText/rl/train_rl.py
Work was started by @mydatascience and completed by @A9isha

Tests

I have tested the code on Llama3.1-8b and Llama3.1-70b on v7x-128, v5p-64

Checklist

Before submitting this PR, please make sure (put X in square brackets):

• I have performed a self-review of my code. For an optional AI review, add the gemini-review label.
• I have necessary comments in my code, particularly in hard-to-understand areas.
• I have run end-to-end tests tests and provided workload links above if applicable.
• I have made or will make corresponding changes to the doc if needed, including adding new documentation pages to the relevant Table of Contents (toctree directive) as explained in our documentation.

[github-actions]

🤖 Hi @A9isha, I've received your request, and I'm working on it now! You can track my progress in the logs for more details.

[xuefgu]

Thanks for the PR @A9isha !

In addition to the comments - can you please clarify what tests you have performed, i.e. the hardware, model, and more importantly with what configs (since that's the main change here)?

[A9isha]

Dummy approval since an owner approval is needed :D