Update ER_MAP/dashboard.py

Update ER_MAP/dashboard.py

Update ER_MAP/dashboard.py

feat: add support for lowercase Hugging Face Space secrets

Clean up: remove junk files (opus_prompt, mermaid_render, EVALUATION, unused UI, train_ppo)

Submission-ready: README, blog, training pipeline, baseline evidence, OpenEnv compliance

fix(grpo): Unsloth inference-mode swap + smaller LoRA + KV cleanup (T4 OOM #2)

fix(grpo): per-step backward to bound VRAM during update (T4 OOM fix)

fix(grpo): skip reference model when kl_beta=0 to save 5GB VRAM on T4

chore(kaggle): rebuild notebook v3 + clean dev-scratch files

api_router: read ERMAP_<ROLE>_MODEL env vars

train: per-phase reward thresholds for early-stop

kaggle: route Patient + Nurse to 8B-instant pool

kaggle: 8B Doctor + train-until-optimal early-stop

LLM Emotion Adapter for TTS + LLM Empathy Judge + reward rebalance

Implement LLM-as-a-Judge and CLI Tester

Initial commit