|
--- |
|
license: apache-2.0 |
|
datasets: |
|
- openbmb/VisRAG-Ret-Train-In-domain-data |
|
- openbmb/VisRAG-Ret-Train-Synthetic-data |
|
language: |
|
- en |
|
base_model: |
|
- openbmb/MiniCPM-V-2 |
|
tags: |
|
- VisRAG |
|
pipeline_tag: feature-extraction |
|
--- |
|
# VisRAG: Vision-based Retrieval-augmented Generation on Multi-modality Documents |
|
<div style="display: flex; align-items: center;"> |
|
<a href="https://huggingface.co/openbmb/VisRAG-Ret" style="margin-right: 10px;"> |
|
<img src="https://img.shields.io/badge/VisRAG_Ret-fcd022?style=for-the-badge&logo=huggingface&logoColor=000" alt="VisRAG Ret"> |
|
</a> |
|
<a href="https://huggingface.co/collections/openbmb/visrag-6717bbfb471bb018a49f1c69" style="margin-right: 10px;"> |
|
<img src="https://img.shields.io/badge/VisRAG_Collection-fcd022?style=for-the-badge&logo=huggingface&logoColor=000" alt="VisRAG Collection"> |
|
</a> |
|
<a href="https://arxiv.org/abs/2410.10594" style="margin-right: 10px;"> |
|
<img src="https://img.shields.io/badge/arXiv-2410.10594-ff0000.svg?style=for-the-badge" alt="arXiv"> |
|
</a> |
|
<a href="https://colab.research.google.com/drive/11KV9adDNXPfHiuFAfXNOvtYJKcyR8JZH?usp=sharing" style="margin-right: 10px;"> |
|
<img src="https://img.shields.io/badge/VisRAG_Pipeline-f9ab00?style=for-the-badge&logo=googlecolab&logoColor=000" alt="Google Colab"> |
|
</a> |
|
<a href="https://github.com/openbmb/VisRAG" style="margin-right: 10px;"> |
|
<img src="https://img.shields.io/badge/VisRAG-000000?style=for-the-badge&logo=github&logoColor=white" alt="GitHub"> |
|
</a> |
|
</div> |
|
|
|
<p align="center">• |
|
<a href="#📖-introduction"> 📖 Introduction </a> • |
|
<a href="#🎉-news">🎉 News</a> • |
|
<a href="#✨-visrag-pipeline">✨ VisRAG Pipeline</a> • |
|
<a href="#⚡️-training">⚡️ Training</a> |
|
</p> |
|
<p align="center">• |
|
<a href="#📦-requirements">📦 Requirements</a> • |
|
<a href="#🔧-usage">🔧 Usage</a> • |
|
<a href="#📄-license">📄 Lisense</a> • |
|
<a href="#📑-citation">📑 Citation</a> • |
|
<a href="#📧-contact">📧 Contact</a> |
|
</p> |
|
|
|
# 📖 Introduction |
|
**VisRAG** is a novel vision-language model (VLM)-based RAG pipeline. In this pipeline, instead of first parsing the document to obtain text, the document is directly embedded using a VLM as an image and then retrieved to enhance the generation of a VLM.Compared to traditional text-based RAG, **VisRAG** maximizes the retention and utilization of the data information in the original documents, eliminating the information loss introduced during the parsing process. |
|
<p align="center"><img width=800 src="https://github.com/openbmb/VisRAG/blob/master/assets/main_figure.png?raw=true"/></p> |
|
|
|
# 🎉 News |
|
|
|
* 20241015: Released our train data and test data on Hugging Face which can be found in the [VisRAG](https://huggingface.co/collections/openbmb/visrag-6717bbfb471bb018a49f1c69) Collection on Hugging Face. It is referenced at the beginning of this page. |
|
* 20241014: Released our [Paper](https://arxiv.org/abs/2410.10594) on arXiv. Released our [Model](https://huggingface.co/openbmb/VisRAG-Ret) on Hugging Face. Released our [Code](https://github.com/OpenBMB/VisRAG) on GitHub. |
|
|
|
# ✨ VisRAG Pipeline |
|
|
|
## VisRAG-Ret |
|
**VisRAG-Ret** is a document embedding model built on [MiniCPM-V 2.0](https://huggingface.co/openbmb/MiniCPM-V-2), a vision-language model that integrates [SigLIP](https://huggingface.co/google/siglip-so400m-patch14-384) as the vision encoder and [MiniCPM-2B](https://huggingface.co/openbmb/MiniCPM-2B-sft-bf16) as the language model. |
|
|
|
## VisRAG-Gen |
|
In the paper, We use MiniCPM-V 2.0, MiniCPM-V 2.6 and GPT-4o as the generators. Actually you can use any VLMs you like! |
|
|
|
# ⚡️ Training |
|
|
|
## VisRAG-Ret |
|
Our training dataset of 362,110 Query-Document (Q-D) Pairs for **VisRAG-Ret** is comprised of train sets of openly available academic datasets (34%) and a synthetic dataset made up of pages from web-crawled PDF documents and augmented with VLM-generated (GPT-4o) pseudo-queries (66%). It can be found in the `VisRAG` Collection on Hugging Face, which is referenced at the beginning of this page. |
|
|
|
|
|
## VisRAG-Gen |
|
The generation part does not use any fine-tuning; we directly use off-the-shelf LLMs/VLMs for generation. |
|
|
|
# 📦 Requirements |
|
``` |
|
torch==2.1.2 |
|
torchvision==0.16.2 |
|
transformers==4.40.2 |
|
sentencepiece==0.1.99 |
|
decord==0.6.0 |
|
Pillow==10.1.0 |
|
``` |
|
|
|
# 🔧 Usage |
|
|
|
## VisRAG-Ret |
|
```python |
|
from transformers import AutoModel, AutoTokenizer |
|
import torch |
|
import torch.nn.functional as F |
|
from PIL import Image |
|
import requests |
|
from io import BytesIO |
|
|
|
def weighted_mean_pooling(hidden, attention_mask): |
|
attention_mask_ = attention_mask * attention_mask.cumsum(dim=1) |
|
s = torch.sum(hidden * attention_mask_.unsqueeze(-1).float(), dim=1) |
|
d = attention_mask_.sum(dim=1, keepdim=True).float() |
|
reps = s / d |
|
return reps |
|
|
|
@torch.no_grad() |
|
def encode(text_or_image_list): |
|
|
|
if (isinstance(text_or_image_list[0], str)): |
|
inputs = { |
|
"text": text_or_image_list, |
|
'image': [None] * len(text_or_image_list), |
|
'tokenizer': tokenizer |
|
} |
|
else: |
|
inputs = { |
|
"text": [''] * len(text_or_image_list), |
|
'image': text_or_image_list, |
|
'tokenizer': tokenizer |
|
} |
|
outputs = model(**inputs) |
|
attention_mask = outputs.attention_mask |
|
hidden = outputs.last_hidden_state |
|
|
|
reps = weighted_mean_pooling(hidden, attention_mask) |
|
embeddings = F.normalize(reps, p=2, dim=1).detach().cpu().numpy() |
|
return embeddings |
|
|
|
model_name_or_path = "openbmb/VisRAG-Ret" |
|
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, trust_remote_code=True) |
|
model = AutoModel.from_pretrained(model_name_or_path, torch_dtype=torch.bfloat16, trust_remote_code=True).cuda() |
|
model.eval() |
|
|
|
queries = ["What does a dog look like?"] |
|
INSTRUCTION = "Represent this query for retrieving relevant documents: " |
|
queries = [INSTRUCTION + query for query in queries] |
|
|
|
print("Downloading images...") |
|
passages = [ |
|
Image.open(BytesIO(requests.get( |
|
'https://github.com/OpenBMB/VisRAG/raw/refs/heads/master/scripts/demo/retriever/test_image/cat.jpeg' |
|
).content)).convert('RGB'), |
|
Image.open(BytesIO(requests.get( |
|
'https://github.com/OpenBMB/VisRAG/raw/refs/heads/master/scripts/demo/retriever/test_image/dog.jpg' |
|
).content)).convert('RGB') |
|
] |
|
print("Images downloaded.") |
|
|
|
embeddings_query = encode(queries) |
|
embeddings_doc = encode(passages) |
|
|
|
scores = (embeddings_query @ embeddings_doc.T) |
|
print(scores.tolist()) |
|
``` |
|
|
|
# 📄 License |
|
|
|
* The code in this repo is released under the [Apache-2.0](https://github.com/OpenBMB/MiniCPM/blob/main/LICENSE) License. |
|
* The usage of **VisRAG-Ret** model weights must strictly follow [MiniCPM Model License.md](https://github.com/OpenBMB/MiniCPM/blob/main/MiniCPM%20Model%20License.md). |
|
* The models and weights of **VisRAG-Ret** are completely free for academic research. After filling out a ["questionnaire"](https://modelbest.feishu.cn/share/base/form/shrcnpV5ZT9EJ6xYjh3Kx0J6v8g) for registration, **VisRAG-Ret** weights are also available for free commercial use. |
|
|
|
# 📑 Citation |
|
|
|
``` |
|
@misc{yu2024visragvisionbasedretrievalaugmentedgeneration, |
|
title={VisRAG: Vision-based Retrieval-augmented Generation on Multi-modality Documents}, |
|
author={Shi Yu and Chaoyue Tang and Bokai Xu and Junbo Cui and Junhao Ran and Yukun Yan and Zhenghao Liu and Shuo Wang and Xu Han and Zhiyuan Liu and Maosong Sun}, |
|
year={2024}, |
|
eprint={2410.10594}, |
|
archivePrefix={arXiv}, |
|
primaryClass={cs.IR}, |
|
url={https://arxiv.org/abs/2410.10594}, |
|
} |
|
``` |
|
|
|
# 📧 Contact |
|
|
|
- Shi Yu: [email protected] |
|
- Chaoyue Tang: [email protected] |
