Strip non-functional buttons; prefill prompt; use NVIDIA logo

Rebuild frontend as image-first overlay UI (Firefly-style)

Move pipeline load to module top (build-stage cache); collapse advanced on mobile

Reskin frontend in NVIDIA style; mobile-optimized layout

Defer pipeline load to first request; dynamic GPU duration

ZeroGPU: module-level model load, xlarge GPU + 120s duration

Fix license metadata: use HF-recognized 'other' value

Add gradio.Server app with custom frontend for Qwen-Image-Flash

initial commit