Spaces:
Sleeping
Sleeping
Commit History
-
feat: add Qdrant configuration and handle disabled state in QdrantService
SuZeAI27e079c -
feat: update subscription response to track AI credits instead of tokens
SuZeAI437ffc0 -
feat: add token usage tracking for AI recommendations and subscription management
SuZeAIff5991d -
feat: add subscription management with user plan tracking and response schema
SuZeAId249b9c
-
Remove compiled Python cache files from backend and tests
SuZeAIc890de3 -
Add example images for virtual try-on feature
SuZeAIbab5292 -
Add chat conversations and try-on history endpoints
SuZeAIPersist stylist chats grouped into conversations (new chat_conversations table and conversation_id on chat_messages) with list/messages/delete endpoints. Add a try-on task listing endpoint for history, and a startup migration for the new column.8ee0940 -
Use OpenAI GPT-4.1-mini for AI Stylist recommendations
SuZeAIAdd an LLM provider selector (openai/nvidia) defaulting to OpenAI's gpt-4.1-mini with JSON-mode responses, falling back to NVIDIA then mock. Read the OpenAI key from OPENAI_API_KEY or the legacy GPT_API_KEY.af27d60 -
Add virtual try-on for Marketplace and Dashboard items
SuZeAIImport a catalog image into the wardrobe by URL (new /wardrobe/import-url endpoint) and preselect it on the try-on page. Adds a hover Try-on action to marketplace product cards and dashboard outfit cards.75b9ab0 -
Add OpenAI vision model configuration and garment masking functionality
SuZeAI- Updated .env.example to include OPENAI_USE_MASK and OPENAI_VISION_MODEL. - Enhanced Settings class in config.py with new configuration options for garment masking. - Modified TryOnService to support garment region segmentation and mask generation. - Added instructions for running the backend locally in run_backend_local.md.23f80df
-
Update .gitignore to include specific __pycache__ directories and adjust patterns for Python files
SuZeAI628e09a -
Fail fast on hard Gemini quota instead of spamming every key
SuZeAIA 429 with 'limit: 0' means the model is unavailable on the plan for every key (e.g. free tier gives 0 image requests), so rotating through all keys just made N redundant calls and still failed. Detect that case and stop after one call. Surface a clear quota/billing message on the try-on task instead of a raw error.d47f0ef -
Add Gemini API key rotation
SuZeAIPool multiple Gemini keys (GEMINI_API_KEYS, comma-separated) and rotate to the next key on an auth/quota/server failure or network error, cycling through all keys before giving up. Shared by the try-on and vision services so a key that hits its quota is skipped everywhere. Falls back to the single GEMINI_API_KEY when no list is set.6cab4e8 -
Update .gitignore for Python and add pitch deck slide
SuZeAI3b2bbcd -
Fix Qdrant healthcheck in docker-compose
SuZeAIThe qdrant image ships no curl/wget, so the curl-based healthcheck always failed and reported the container unhealthy. Probe the REST port with bash /dev/tcp instead.470e1c9 -
Update try-on tests for the API-based service
SuZeAIMock tryon_service.generate instead of CatVTON/AutoMasker; add tests for the Gemini provider image parsing and the missing-key error path.3135f39 -
Trim backend dependencies to an API-only stack
SuZeAIDrop torch, transformers, opencv, diffusers, accelerate, peft, ultralytics and the detectron2/densepose build deps now that no model runs locally. This also removes ~37 nvidia-cuda packages and shrinks the lockfile from ~200 to 69 packages, so the image installs fast on a CPU-only host.3073d0f -
Switch try-on and garment analysis to external image APIs
SuZeAIReplace the local CatVTON+mask pipeline with a multi-provider image-generation service (Gemini / OpenAI gpt-image-1 / Replicate), selectable via TRYON_PROVIDER. Wardrobe item analysis now classifies garments via the Gemini Vision API instead of a local SegFormer model, so the backend needs no torch/transformers. - tryon_service: person + garment images -> generated try-on (sequential for multi-garment), provider chosen by env, clear errors when keys are missing - vision_service: Gemini garment classification (category/color/style_tags) with a simulated-metadata fallback - tasks: rewrite run_idm_vton_tryon -> run_tryon; drop CatVTON/SegFormer loaders - config/.env.example: add TRYON_PROVIDER, GEMINI_IMAGE_MODEL, GEMINI_VISION_MODEL, REPLICATE_*; drop CLOTHES_SEGMENTATION_MODEL, TRYON_MODEL and CUDA settings25d14ad -
Remove vendored CatVTON model and segmentation pipeline
SuZeAIDelete the bundled CatVTON repo (SD-inpainting try-on, DensePose and SCHP human-parsing, detectron2/densepose vendored trees). Virtual try-on no longer runs a local diffusion+segmentation pipeline.75f84ed
-
feat: implement conversational chatbot with routing and chat history support
265cc9d -
feat: default STORAGE_MODE to cloudinary and force real CatVTON execution on CPU
85a11ff -
feat: add CPU simulation fallback mode for tryon to prevent hangs/OOM on CPU spaces
6aa0990 -
perf: optimize CatVTON on CPU with float32 and 20 steps, increase frontend polling to 10 min
f222800 -
feat: proactively create chest_cm, waist_cm, hips_cm columns on startup if missing
c1affc0 -
feat: add Cloudinary support as an alternative image storage mode
afee3ee -
feat: add chest, waist, and hips measurements support in profile and LLM service
54cc2f8 -
fix: set cfg.MODEL.DEVICE in DensePose initialization for CPU nodes
fca1b3f -
chore: add missing ML dependencies for CatVTON
a4ac20d
-
chore: move CatVTON inside backend and resolve paths dynamically without binary resources
f32f6d0 -
fix: auth headers and postgresql connection for cloud deployment
3745765
-
Auto-append ssl_cert_reqs=CERT_NONE to rediss URL in config
dd8eb63 -
Add REDIS_URL and QDRANT_URL environment variable support
cd337a4 -
Fix OpenGL package in Dockerfile for newer Debian images
8e90cb9 -
Move Docker deployment files inside backend/
8390e42
-
docs: add env template examples for backend and frontend
79cf4bb -
feat: integrate local CatVTON virtual try-on and secure env files
9a36956
-
Update SERVER_HOST_URL to port 8001 in .env templates
e4a3e5d -
Allow Vite preview port 4173 in CORS origins
db25f46 -
Update README.md execution instructions with cd backend and port 8001
7b93e81 -
Update README.md with local storage mode guides and correct Redis port
cb9acf9 -
Support and default to STORAGE_MODE=local for zero-key local development
640f288 -
Update Redis port to 6380 in docker-compose.yml to avoid conflicts
88efd0d -
Create comprehensive README.md for backend codebase
b6f3a4c -
Initialize production-grade FastAPI backend with models, routes, celery workers, and test suite
b18b789