athrael-soju/Snappy
π Snappy's unique approach unifies vision-language late interaction with structured OCR for region-level knowledge retrieval. Like the project? Drop a star! β
Implements region-level document retrieval by mapping ColPali's patch-level similarity scores to OCR-extracted bounding boxes via spatial coordinate propagation, enabling fine-grained retrieval without additional training. Operates as a two-stage pipeline: efficient candidate prefetching using mean-pooled embeddings followed by full-resolution region reranking with interpretability heatmaps. Built on FastAPI and Next.js with Qdrant vector storage, DeepSeek-OCR-2 integration, and Docker deployment for production RAG systems.
Stars
83
Forks
15
Language
Python
License
MIT
Category
Last pushed
Feb 09, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/vector-db/athrael-soju/Snappy"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related tools
fredsiika/huxley-pdf
Upload personal docs and Chat with your PDF files with this GPT4-powered app. Built with...
aakashsharan/research-vault
AI research assistant that extracts structured patterns from papers using RAG, LangGraph, and...
roberto729a/OllamaRAG
π€ Build a smart AI assistant that learns from any website using a Retrieval-Augmented Generation...
waterpare833/Novel-Assistant
λ‘컬 λλ ν΄λΌμ°λ LLMμ μ¬μ©ν΄ λ¬Έμ ν΄λ κΈ°λ°μΌλ‘ κ²μΒ·λ΅λ³νλ RAG μ΄μμ€ν΄νΈ
Srijan-D/pdf.ai
Edge compatible PDF.ai Chat with any PDF document You can ask questions, get summaries, find...