Bisheng-unstructured is an open-source unstructured data parsing library built to power LLM applications like pretrain, finetune, prompting engineering. Bisheng-unstructured makes the unstructured data porcessing more easily and provides a consistent user experience regardless of any file types.

The project is a sub-project of bisheng.

• High precision pdf layout parser
• High precision table structure recovering
• High precision OCR ability
• More friendly for token prossing for the visual text element, like table, list

Use as a chain node ElemUnstructureLoader

We provide a open cloud service for easily use. See free trial.

• Install from pip: pip install bisheng-unstructured
• Quick Start Guide

For guidance on installation, development, deployment, and administration, check out bisheng-unstructured Docs.

Reporting problems, asking questions We appreciate any feedback, questions or bug reporting regarding this project.

User can posting Issues, follow the process outlined in the Stack Overflow document.

For questions, we recommend posting in our community GitHub Discussions.

bisheng-unstructured adopts dependencies from the following:

• Thanks to unstructured for the main framework.