Update README.md

Update README.md

Update docs/HF_MINI_BLOG.md

Update docs/HF_MINI_BLOG.md

Update docs/HF_MINI_BLOG.md

Update docs/HF_MINI_BLOG.md

Update docs/HF_MINI_BLOG.md

Update docs/HF_MINI_BLOG.md

README: restructure to hackathon template (demo/results/quickstart)

README: link published HF Post mini-blog

Update docs/HF_MINI_BLOG.md

Update docs/HF_MINI_BLOG.md

mini-blog: reorder (problem→story→market), remove file metrics

mini-blog: add market sizing + ARR wedge paragraph

mini-blog: add quantified snapshot section

sync local env repo (server/training/docs/readme)

README: link mini-blog markdown for submission form

mini-blog: add storytelling + simple end-to-end architecture

README: add mini-blog link placeholder

add HF mini-blog draft (copy/paste into HF Post)

notebook: embed executed GRPO run outputs + plot

fix: auto-reset on first step + dashboard auto-reset on load (no more 500 on click query_policy)

deploy: full Space sync — app + server + training + GRPO run artifacts

README: embed GRPO training curves + reference adapter run

GRPO LoRA adapter: grpo-20260425-223508

GRPO run: grpo-20260425-223508

GRPO run: grpo-20260425-223508

GRPO run: grpo-20260425-223508

make GRPOConfig kwargs version-tolerant

add HF Jobs GRPO training script

wire Colab notebook to TRL GRPOTrainer (real LoRA weight updates)

add runs/20260425-215059/qwen_vs_random.png (re-upload)

add runs/20260425-215059/reward_curves.png (re-upload)

allow runs/**/*.png in tracking

add runs/20260425-215059/qwen_vs_random.png

add runs/20260425-215059/reward_curves.png

hackathon submission: theme-aligned README, refreshed training, Qwen vs random plots

add chart for runs/20260425-215059

training run: runs/20260425-215059

training run: runs/20260425-215059

Add CUDA backend (transformers) for HF Jobs A10G runs

training run: runs/20260425-214433

training run: runs/20260425-214433

training run: runs/20260425-214217

training run: runs/20260425-214217

training run: runs/20260425-213601

training run: runs/20260425-213601

Deploy ClaimSense adjudication gym

initial commit