Add self-contained GRPO training notebook for HF Jupyter

Add root health check route to fix HF 404

Add alternating self-play training scaffolding.

Add GRPO training pipeline + remove shield emoji

Initial commit of CyberSOC upgraded RLVR environment