User Guide¶
Learn how to use Duckling effectively.
Overview¶
Duckling provides a comprehensive interface for document conversion with advanced features like OCR, table extraction, and RAG chunking.
Sections¶
-
Features
Explore all the capabilities of Duckling
-
Supported Formats
Input and output format reference
-
Configuration
Customize OCR, tables, images, and performance settings
Quick Tips¶
Batch Processing
Toggle batch mode to convert multiple files simultaneously. The system processes up to 2 files in parallel to balance speed and memory usage.
OCR Selection
- EasyOCR: Best for multi-language documents with GPU support
- Tesseract: Reliable for simple documents
- macOS Vision: Fastest on Mac with Apple Silicon
- RapidOCR: Lightweight and fast
RAG Chunking
Enable chunking in settings to generate document chunks optimized for retrieval-augmented generation. Chunks include metadata like headings and page numbers.