Fix VLM models in refinement_loop.py to match app.py (Llama-4-Scout, Kimi-K2.6, Qwen3.5-9B)

Fix VLM: use Llama-4-Scout via nscale (confirmed working with images), Kimi-K2.6 and Qwen3.5-9B via together as fallbacks

Complete README with architecture docs, VLM provider findings, and agentic refinement design

Fix VLM models in refinement_loop.py: use actual vision models with correct providers

Fix VLM models: use actual vision models with correct providers (Qwen2.5-VL, Gemma-4 via novita, Kimi via fireworks-ai)

Add Agentic Refinement tab: iterative image→pattern→3D→projection→compare→refine loop

Add scikit-image for SSIM similarity metrics

Add agentic refinement loop module

Pass actual 2D pattern pieces to 3D builder — 3D view now matches 2D patterns

3D view now built FROM actual 2D pattern pieces (triangulate → cylindrical wrap → Mesh3d)

Fix Gradio 6.0 compatibility: move css/theme from Blocks() to launch()

Fix 3D garment visualization: anatomically correct body, proper garment positioning and scaling

Fix build: remove gradio version pin that conflicts with sdk_version 6.13.0

Upload requirements.txt

Rewrite 3D visualizer with proper surface meshes (not wireframes)

Add chat editing + 3D preview + updated models

Add 3D garment visualizer (Plotly)

Add chat editing + 3D garment preview

Upgrade to Qwen3.5-9B, Gemma 4, and Kimi-K2.5 (replace outdated Qwen2.5-VL)\n\nVerified working via Together AI provider:\n- Qwen/Qwen3.5-9B (primary VLM)\n- google/gemma-4-31B-it (Gemma 4)\n- moonshotai/Kimi-K2.5 (fallback)\n\nHandles reasoning+content response format from newer models."

Update README with proper metadata

Add main Gradio app with VLM integration

Add pattern generator module

Add requirements

initial commit