bge m3

ProviderBAAI
Categorysentence-similarity
Licensemit
Downloads3.1M
Stars307

Overview

BGE-M3 is a versatile embedding model designed for high-performance retrieval across diverse linguistic landscapes. Unlike traditional embedding models limited to a single modality or language, M3 focuses on 'multi-linguality, multi-functionality, and multi-granularity.' For developers, this means a single model can handle dense retrieval, sparse retrieval (BM25-style), and multi-vector reranking simultaneously. It significantly expands the context window compared to earlier BGE iterations, allowing for the processing of longer documents without aggressive truncation. This makes it an ideal backbone for RAG pipelines where hybrid search is required to balance semantic meaning with keyword precision across global datasets.

Highlights

  • Supports dense, sparse, and multi-vector retrieval modes.
  • Native multi-lingual support for global application deployment.
  • Extended context window for processing longer documents.
  • MIT licensed for flexible commercial and private integration.

Usage

Install
# Install Hugging Face transformers
pip install transformers torch
SDK Usage
# Load model with transformers
from transformers import AutoModel, AutoTokenizer

model = AutoModel.from_pretrained("BAAI/bge-m3")
tokenizer = AutoTokenizer.from_pretrained("BAAI/bge-m3")

Hugging Face Download

We recommend downloading the model via the Hugging Face CLI or Hub SDK.

Guidance：Before downloading, install huggingface_hub with:

Guidance 
pip install -U huggingface_hub

CLI Download

Download the full repository

Download the full repository 
huggingface-cli download BAAI/bge-m3

Download a single file to a local folder (e.g. config.json into ./dir)

Download a single file to a local folder (e.g. config.json into ./dir) 
huggingface-cli download BAAI/bge-m3 config.json --local-dir ./dir

See the official docs for more CLI options

SDK Download

SDK Download 
# 模型下载
from huggingface_hub import snapshot_download
model_dir = snapshot_download('BAAI/bge-m3')

Git Download

Make sure git-lfs is installed first

Git Download 
git lfs install
git clone https://huggingface.co/BAAI/bge-m3

To skip LFS large-file downloads, use:

Skip LFS 
GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/BAAI/bge-m3

Model files are hosted on the Hugging Face Hub — download directly via HF CLI / SDK / Git, not through this site.

PyTorch / Transformers Usage

Install Transformers

Install Transformers 
pip install -U transformers torch

Load the model and run inference

Load the model and run inference 
from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained('BAAI/bge-m3')
tokenizer = AutoTokenizer.from_pretrained('BAAI/bge-m3')

Model Download

We recommend downloading the model via the ModelScope CLI or SDK.

Guidance：Before downloading, install ModelScope with:

Guidance 
pip install modelscope

CLI Download

Download the full repository

Download the full repository 
modelscope download --model BAAI/bge-m3

Download a single file to a local folder (e.g. README.md into ./dir)

Download a single file to a local folder (e.g. README.md into ./dir) 
modelscope download --model BAAI/bge-m3 README.md --local_dir ./dir

See the docs for more CLI options

SDK Download

SDK Download 
# 模型下载
from modelscope import snapshot_download
model_dir = snapshot_download('BAAI/bge-m3')

Git Download

Make sure git-lfs is installed first

Git Download 
git lfs install
git clone https://www.modelscope.cn/BAAI/bge-m3.git

To skip LFS large-file downloads, use:

Skip LFS 
GIT_LFS_SKIP_SMUDGE=1 git clone https://www.modelscope.cn/BAAI/bge-m3.git

ModelScope 模型页直接下载模型文件；无需将模型文件放在本站服务器。

Notebook Quickstart

Install the ModelScope library

Install the ModelScope library 
pip install "modelscope[audio,cv,nlp,multi-modal,science]" -f https://modelscope.oss-cn-beijing.aliyuncs.com/releases/repo.html

Load the model and run inference

Load the model and run inference 
from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks

p = pipeline('text-generation', 'BAAI/bge-m3')

Full Documentation

来源: HuggingFace

---
pipeline_tag: sentence-similarity
tags:

  • sentence-transformers

  • feature-extraction

  • sentence-similarity

license: mit
---

For more details please refer to our github repo: https://github.com/FlagOpen/FlagEmbedding

BGE-M3 (paper, code)

In this project, we introduce BGE-M3, which is distinguished for its versatility in Multi-Functionality, Multi-Linguality, and Multi-Granularity.

  • Multi-Functionality: It can simultaneously perform the three common retrieval functionalities of embedding model: dense retrieval, multi-vector retrieval, and sparse retrieval.

  • Multi-Linguality: It can support more than 100 working languages.

  • Multi-Granularity: It is able to process inputs of different granularities, spanning from short sentences to long documents of up to 8192 tokens.

Some suggestions for retrieval pipeline in RAG

We recommend to use the following pipeline: hybrid retrieval + re-ranking.

  • Hybrid retrieval leverages the strengths of various methods, offering higher accuracy and stronger generalization capabilities.

A classic example: using both embedding retrieval and the BM25 algorithm.
Now, you can try to use BGE-M3, which supports both embedding and sparse retrieval.
This allows you to obtain token weights (similar to the BM25) without any additional cost when generate dense embeddings.
To use hybrid retrieval, you can refer to Vespa and Milvus.

  • As cross-encoder models, re-ranker demonstrates higher accuracy than bi-encoder embedding model.
Utilizing the re-ranking model (e.g., bge-reranker, bge-reranker-v2) after retrieval can further filter the selected text.

News:

  • 2024/7/1: We update the MIRACL evaluation results of BGE-M3. To reproduce the new results, you can refer to: bge-m3_miracl_2cr. We have also updated our paper on arXiv.
<details> <summary> Details </summary>

The previous test results were lower because we mistakenly removed the passages that have the same id as the query from the search results. After correcting this mistake, the overall performance of BGE-M3 on MIRACL is higher than the previous results, but the experimental conclusion remains unchanged. The other results are not affected by this mistake. To reproduce the previous lower results, you need to add the --remove-query parameter when using pyserini.search.faiss or pyserini.search.lucene to search the passages.

</details>

  • 2024/3/20: Thanks Milvus team! Now you can use hybrid retrieval of bge-m3 in Milvus: pymilvus/examples

/hello_hybrid_sparse_dense.py.
  • 2024/3/8: Thanks for the experimental results from @Yannael. In this benchmark, BGE-M3 achieves top performance in both English and other languages, surpassing models such as OpenAI.

  • 2024/3/2: Release unified fine-tuning example and data


  • 2024/2/1: Thanks for the excellent tool from Vespa. You can easily use multiple modes of BGE-M3 following this notebook

Specs

  • Model

| Model Name | Dimension | Sequence Length | Introduction |
|:----:|:---:|:---:|:---:|
| BAAI/bge-m3 | 1024 | 8192 | multilingual; unified fine-tuning (dense, sparse, and colbert) from bge-m3-unsupervised|
| BAAI/bge-m3-unsupervised | 1024 | 8192 | multilingual; contrastive learning from bge-m3-retromae |
| BAAI/bge-m3-retromae | -- | 8192 | multilingual; extend the max_length of xlm-roberta to 8192 and further pretrained via retromae|
| BAAI/bge-large-en-v1.5 | 1024 | 512 | English model |
| BAAI/bge-base-en-v1.5 | 768 | 512 | English model |
| BAAI/bge-small-en-v1.5 | 384 | 512 | English model |

  • Data

| Dataset | Introduction |
|:----------------------------------------------------------:|:-------------------------------------------------:|
| MLDR | Docuemtn Retrieval Dataset, covering 13 languages |
| bge-m3-data | Fine-tuning data used by bge-m3 |

FAQ

1. Introduction for different retrieval methods

  • Dense retrieval: map the text into a single embedding, e.g., DPR, BGE-v1.5
  • Sparse retrieval (lexical matching): a vector of size equal to the vocabulary, with the majority of positions set to zero, calculating a weight only for tokens present in the text. e.g., BM25, unicoil, and splade
  • Multi-vector retrieval: use multiple vectors to represent a text, e.g., ColBERT.

2. How to use BGE-M3 in other projects?

For embedding retrieval, you can employ the BGE-M3 model using the same approach as BGE.
The only difference is that the BGE-M3 model no longer requires adding instructions to the queries.

For hybrid retrieval, you can use Vespa and Milvus.

3. How to fine-tune bge-M3 model?

You can follow the common in this example
to fine-tune the dense embedding.

If you want to fine-tune all embedding function of m3 (dense, sparse and colbert), you can refer to the unified_fine-tuning example

Usage

Install:

code
git clone https://github.com/FlagOpen/FlagEmbedding.git
cd FlagEmbedding
pip install -e .

or:
code
pip install -U FlagEmbedding

Generate Embedding for text

  • Dense Embedding
```python from FlagEmbedding import BGEM3FlagModel

model = BGEM3FlagModel('BAAI/bge-m3',
use_fp16=True) # Setting use_fp16 to True speeds up computation with a slight performance degradation

sentences_1 = ["What is BGE M3?", "Defination of BM25"]
sentences_2 = ["BGE M3 is an embedding model supporting dense retrieval, lexical matching and multi-vector interaction.",
"BM25 is a bag-of-words retrieval function that ranks a set of documents based on the query terms app

View official source
Join our Telegram