Spaces:
Sleeping
Sleeping
Commit History
-
perf: fully async DB - results in 5-7s, background check+save
8f86a3c -
perf: async DB save with duplicate check for faster extraction
- Cache init_db() with @st.cache_resource (runs once per session) - Background thread for database saves (fire-and-forget) - Quick duplicate check before queueing save - Accurate status: 'duplicate' vs 'queued' vs 'disabled'4bdd01c -
feat: PDF preview, database integration, and improved error handling
- Add PDF preview support using pdf2image - Enable bounding box overlay visualization for PDFs - Implement database persistence with SQLModel (Invoice, LineItem) - Add InvoiceRepository with save and duplicate detection - Improve DB status messages (show 'unavailable' once at startup) - Show 'Demo Mode' toast only once per session - Fix torch.load and transformers deprecation warnings - Add conda environment.yml for reproducible setup - Update README with conda installation instructions2a944a5
-
Refactor: Replace Tesseract with DocTR and integrate LayoutLMv3-DocTR model
Major overhaul of OCR/Inference pipeline. Swapped Tesseract for DocTR, retrained LayoutLMv3 (~83% F1), and fixed address extraction using Fuzzy Matching.ec0b507
-
feat: add AI detection overlay visualization with bounding boxes on extracted entities
343b0c3
-
feat: Update Dockerfile and requirements for PDF processing, add new dependencies, and refactor API structure
faa3050
-
feat: Enhance pipeline with smart PDF handling, Pydantic validation, and semantic hashing, and refactor API to src.
f74e17e
-
feat: Add Phase 3 generalization scripts and clean up legacy files
d79b7f7
-
feat: LayoutLMv3 integration, Streamlit UI toggle, README refresh, .gitignore
42e1c04
-
Complete Version 0.5 with Streamlit UI and full pipeline
566dc81