DCC Docling Serve

A patched version of docling-serve that bundles DCC docling plugins for enhanced layout detection and OCR.

GitHub Repository: DCC-BS/dcc-docling-serve

Features

• Bundled Plugins: Includes docling-pp-doc-layout and docling-glm-ocr.
• Multi-Environment Support: Provides Docker images for CPU, CUDA 12.8, and CUDA 13.0.
• Flexible Selection: Engines are selectable per-request via the API or the included Gradio UI.
• Full Stack: Includes a Docker Compose setup to run both the Docling API and a vLLM server for OCR.

Installation

1. Prerequisites: Docker with NVIDIA GPU support and a HuggingFace token.
2. Configuration: Copy .env.example to .env and set your HF_TOKEN.
3. Start: Run make docker-up to launch the services.

Usage

• API: Send a POST request to http://localhost:5001/v1/convert/source specifying "ocr_engine": "glm-ocr-remote" and "layout_custom_config": { "kind": "ppdoclayout-v3" }.
• UI: Access the Gradio interface at http://localhost:5001.