Consolidate CLI RL logic

[yatlas20098]

Consolidate RL CLI logic into a new base pipeline.

This CL introduces tunix/cli/base_rl_main.py with a BasePipeline abstract class to house common Reinforcement Learning (RL) CLI logic, refactored out of tunix/cli/grpo_rl_main.py. This change is being made to allow for code reuse when implementing a new PPO CLI.

Updates the GRPO CLI pipeline to inherit from BasePipeline and modifies GRPO CLI tests accordingly.

Change-Id: I88188941bef08525d96b5cdfd03146e198d3a0d1

Resolves #506889007

It's a good idea to open an issue first for discussion.

Reference

Colab Notebook

Checklist

• I have added all the necessary unit tests for my change.
• I have verified that my change does not break existing code and all unit tests pass.
• I have added all appropriate doc-strings/documentation.
• My PR is based on the latest changes of the main branch (if unsure, rebase the code).
• I have signed the Contributor License Agreement.
• I have followed Contribution Guidelines.