feat: download GroundingDINO and SAM2 models at startup

fix: track GroundingDINO config and ignore model weights

Fix SAM2 installation for Docker build

Add README files

Remove test images from repository

cleanup: remove redundant snapshot file

auth, LaMa inpainting, PostgreSQL, UI polish, back button, export fix

Add PostgreSQL authentication and UI improvements

Improve OCR background reconstruction and text inpainting

redesigned the home page and accounts page

resolved feature issue, added library page

Add features: lock, group, duplicate, visibilty and fixed code

Week 4: Layer Proposal Engine, logo decomposer, image classifier, component architecture, undo/redo, selection modes, properties panel

Week 3 complete: canvas, drag, resize, delete, export, history

Dockerfile added, requirements.txt updated with pinned versions

layer_extractor and pipeline complete, full AI pipeline working

Day 5: SD inpainting working with OpenCV fallback

Day 4: PaddleOCR reading text with 94-99% confidence

Day 3: GroundingDINO detection and SAM2 prompt generation

Day 2: SAM2 segmentation working

add gitkeep for test_images folder

Day 1: project structure and FastAPI skeleton