Validate image input before GPU inference

Remove unnecessary Space token dependency

Prevent low-resolution grid artifacts

Quantize model stack for ZeroGPU memory limits

Allow cold four-step edits to finish on ZeroGPU

Harden privacy and pin four-step model stack

Remove undisclosed prompt logging that secretly saves user prompts and images (#2)

Make model loading resilient for duplicated Space

Remove temporary dimension-flow debug prints

Add temporary debug prints for dimension flow (will revert)

Revert "Preserve input image aspect ratio when sliders are at default"

Preserve input image aspect ratio when sliders are at default

Expose a second image input slot in the UI

Add debug diagnostic dashboard

Lower @spaces.GPU duration from 300s to 60s

Disable AOTI compile to dodge spaces==0.49.3 NVML regression

Clean up for public release

Fix device mismatch during AOT compilation warmup

Upgrade to Qwen-Image-Edit-2511 with Phr00t v18 transformer

Reduce memory during AOT compilation by offloading text encoder

Revert "Fix OOM by deferring model loading and AOT compilation to runtime"

Fix model link in UI description

Fix OOM by deferring model loading and AOT compilation to runtime

routine commit

Remove demo files and fix image serving for proper download support

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py (#6)

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py (#5)

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py