vlm-run/vlmrun-hub
A collection of Pydantic schemas for structured data extraction from images, documents, and videos using Vision Language Models.

Velocity · 7d
+1.0
★ / day
Trend
→steady
star history
VLM Run Hub is a repository of pre-defined schemas designed to work with VLMs for extracting structured data from unstructured visual sources such as images, videos, and documents. It provides industry-specific schemas optimized for real-world visual ETL use cases, using Pydantic models to validate and structure the extracted outputs.