---
license: cc-by-nc-sa-4.0
pipeline_tag: image-text-to-text
library_name: transformers
---

# DocLayLLM: An Efficient and Effective Multi-modal Extension of Large Language Models for Text-rich Document Understanding

This model is presented in the paper [DocLayLLM: An Efficient and Effective Multi-modal Extension of Large Language Models for Text-rich Document Understanding](https://huggingface.co/papers/2408.15045). DocLayLLM is designed for text-rich document understanding, integrating visual patch tokens and 2D positional tokens into LLMs to enhance their document comprehension and OCR information perception.

## How to Use

A more complete usage example will be added when available. For now, a basic example:

```python
from transformers import pipeline

pipe = pipeline("text-generation", model="your_model_id")  # replace your_model_id
result = pipe("Your input text here.")
print(result)
```

Replace `"your_model_id"` with the actual Hugging Face model ID.