Reduce ZeroGPU quota use for anonymous visitors

Lower ZeroGPU reservation for anonymous users

Add GitHub repo link to Space README

Tag Space for Build Small judging

Pin Gradio client asset

Pin Gradio client asset

Add project social links

Add video walkthrough links

Clarify sample data button wording

Add Field Notes article link

Update field note quote

Update README field note

Add user field note quote

Add hackathon sponsor acknowledgements

Fix print export and warm hosted model

Bust custom asset cache for demo button

Add demo cases and full-report export copy

Tighten hero copy spacing

Add About page with repository link

Remove environment labels from UI

Add prep export actions

Clarify Space-local processing copy

Update custom Server UI

Sync About page GitHub link

Avoid invented same-day timing

Cap generated section output

Tighten separate section prompts

Generate appointment sections separately

Remove ineffective ZeroGPU warmup

Serialize Gradio llama.cpp GPU events

Serialize llama.cpp warmup and generation

Warm llama.cpp model on page load

Tune llama.cpp latency

Keep concise GPU diagnostics

Add llama.cpp GPU verification logs

Tighten 4B report prompt

Use MedGemma 1.5 4B ZeroGPU llama.cpp UI

Use runnable MedGemma 27B llama.cpp GGUF

Force hosted llama.cpp defaults

Use CUDA wheel for llama-cpp Space build

Switch Space to llama.cpp MedGemma 27B GGUF

Update medication search and patient-friendly copy

Update app UI and model settings

Fix ZeroGPU inference argument pickling

Fix Space host binding

Deploy initial Gradio Space

initial commit