[ioseba]

This PR translates Sections 4 through 7 of Chapter 12 into Spanish:

• Section 4: Implementing GRPO in TRL (GRPOTrainer, configuration, minimal training loop)
• Section 5: Practical Exercise to Fine-tune a model with GRPO (GSM8K reasoning, reward functions, TRL setup)
• Section 6: Practical Exercise with Unsloth (efficient GRPO fine-tuning, memory savings, fast inference)
• Section 7: Schedule and wrap-up (Upcoming live cohort details, Discord community)
• Updated chapters/es/_toctree.yml with the corresponding sections.

All Python code blocks formatted and validated with utils/code_formatter.py (blackify).

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.