bge large en v1.5

提供商BAAI
分类feature-extraction
许可证mit
下载量19.0K
星标7

简介

bge-large-en-v1.5 是由北京智源人工智能研究院（BAAI）推出的高性能英文文本向量模型。它并非用于对话的 LLM，而是将文本转化为高维向量的特征提取工具，在检索增强生成（RAG）架构中扮演关键角色。该模型在语义匹配和文档检索方面表现强劲，能有效将用户查询与知识库内容进行精准对齐。对于开发者而言，它能无缝集成到 Milvus 或 FAISS 等向量数据库中，上手难度低，是构建英文专业知识库、语义搜索系统的理想底层组件。

核心亮点

  • 由智源研究院开发，英文语义检索能力顶尖
  • RAG 架构核心组件，大幅提升知识库检索精度
  • 兼容主流向量数据库，部署与集成门槛极低
  • MIT 协议开源，支持商业化快速落地

使用方法

安装依赖
# 安装 Hugging Face transformers
pip install transformers torch
SDK 使用
# 使用 transformers 加载模型
from transformers import AutoModel, AutoTokenizer

model = AutoModel.from_pretrained("BAAI/bge-large-en-v1.5")
tokenizer = AutoTokenizer.from_pretrained("BAAI/bge-large-en-v1.5")

Hugging Face 下载

我们推荐使用命令行或者 Hugging Face Hub SDK 来进行模型的下载。

操作指引：在下载前，请先通过如下命令安装 huggingface_hub：

操作指引 
pip install -U huggingface_hub

命令行下载

下载完整模型库

下载完整模型库 
huggingface-cli download BAAI/bge-large-en-v1.5

下载单个文件到指定本地文件夹（以下载 config.json 到当前路径下 ./dir 目录为例）

下载单个文件到指定本地文件夹（以下载 config.json 到当前路径下 ./dir 目录为例） 
huggingface-cli download BAAI/bge-large-en-v1.5 config.json --local-dir ./dir

更多命令行下载选项，可参见官方文档

SDK 下载

SDK 下载 
# 模型下载
from huggingface_hub import snapshot_download
model_dir = snapshot_download('BAAI/bge-large-en-v1.5')

Git 下载

请确保 lfs 已经被正确安装

Git 下载 
git lfs install
git clone https://huggingface.co/BAAI/bge-large-en-v1.5

如果您希望跳过 lfs 大文件下载，可以使用如下命令

跳过 LFS 
GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/BAAI/bge-large-en-v1.5

模型文件托管在 Hugging Face Hub，使用 HF CLI / SDK / Git 直接下载，不经过本站。

PyTorch / Transformers 使用

安装 Transformers

安装 Transformers 
pip install -U transformers torch

模型加载和推理

模型加载和推理 
from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained('BAAI/bge-large-en-v1.5')
tokenizer = AutoTokenizer.from_pretrained('BAAI/bge-large-en-v1.5')

模型下载

我们推荐使用命令行或者 ModelScope SDK 来进行模型的下载。

操作指引：在下载前，请先通过如下命令安装 ModelScope：

操作指引 
pip install modelscope

命令行下载

下载完整模型库

下载完整模型库 
modelscope download --model BAAI/bge-large-en-v1.5

下载单个文件到指定本地文件夹（以下载 README.md 到当前路径下 dir 目录为例）

下载单个文件到指定本地文件夹（以下载 README.md 到当前路径下 dir 目录为例） 
modelscope download --model BAAI/bge-large-en-v1.5 README.md --local_dir ./dir

更多更丰富的命令行下载选项，可参见具体文档

SDK 下载

SDK 下载 
# 模型下载
from modelscope import snapshot_download
model_dir = snapshot_download('BAAI/bge-large-en-v1.5')

Git 下载

请确保 lfs 已经被正确安装

Git 下载 
git lfs install
git clone https://www.modelscope.cn/BAAI/bge-large-en-v1.5.git

如果您希望跳过 lfs 大文件下载，可以使用如下命令

跳过 LFS 
GIT_LFS_SKIP_SMUDGE=1 git clone https://www.modelscope.cn/BAAI/bge-large-en-v1.5.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 快速开发

下载并安装 ModelScope library

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

模型加载和推理

模型加载和推理 
from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks

p = pipeline('text-generation', 'BAAI/bge-large-en-v1.5')

完整文档

来源: HuggingFace

---
tags:

  • sentence-transformers

  • feature-extraction

  • sentence-similarity

  • transformers

  • mteb

model-index:
  • name: bge-large-en-v1.5

results:
- task:
type: Classification
dataset:
type: mteb/amazon_counterfactual
name: MTEB AmazonCounterfactualClassification (en)
config: en
split: test
revision: e8379541af4e31359cca9fbcf4b00f2671dba205
metrics:
- type: accuracy
value: 75.8507462686567
- type: ap
value: 38.566457320228245
- type: f1
value: 69.69386648043475
- task:
type: Classification
dataset:
type: mteb/amazon_polarity
name: MTEB AmazonPolarityClassification
config: default
split: test
revision: e2d317d38cd51312af73b3d32a06d1a08b442046
metrics:
- type: accuracy
value: 92.416675
- type: ap
value: 89.1928861155922
- type: f1
value: 92.39477019574215
- task:
type: Classification
dataset:
type: mteb/amazon_reviews_multi
name: MTEB AmazonReviewsClassification (en)
config: en
split: test
revision: 1399c76144fd37290681b995c656ef9b2e06e26d
metrics:
- type: accuracy
value: 48.175999999999995
- type: f1
value: 47.80712792870253
- task:
type: Retrieval
dataset:
type: arguana
name: MTEB ArguAna
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 40.184999999999995
- type: map_at_10
value: 55.654
- type: map_at_100
value: 56.25
- type: map_at_1000
value: 56.255
- type: map_at_3
value: 51.742999999999995
- type: map_at_5
value: 54.129000000000005
- type: mrr_at_1
value: 40.967
- type: mrr_at_10
value: 55.96
- type: mrr_at_100
value: 56.54900000000001
- type: mrr_at_1000
value: 56.554
- type: mrr_at_3
value: 51.980000000000004
- type: mrr_at_5
value: 54.44
- type: ndcg_at_1
value: 40.184999999999995
- type: ndcg_at_10
value: 63.542
- type: ndcg_at_100
value: 65.96499999999999
- type: ndcg_at_1000
value: 66.08699999999999
- type: ndcg_at_3
value: 55.582
- type: ndcg_at_5
value: 59.855000000000004
- type: precision_at_1
value: 40.184999999999995
- type: precision_at_10
value: 8.841000000000001
- type: precision_at_100
value: 0.987
- type: precision_at_1000
value: 0.1
- type: precision_at_3
value: 22.238
- type: precision_at_5
value: 15.405
- type: recall_at_1
value: 40.184999999999995
- type: recall_at_10
value: 88.407
- type: recall_at_100
value: 98.72
- type: recall_at_1000
value: 99.644
- type: recall_at_3
value: 66.714
- type: recall_at_5
value: 77.027
- task:
type: Clustering
dataset:
type: mteb/arxiv-clustering-p2p
name: MTEB ArxivClusteringP2P
config: default
split: test
revision: a122ad7f3f0291bf49cc6f4d32aa80929df69d5d
metrics:
- type: v_measure
value: 48.567077926750066
- task:
type: Clustering
dataset:
type: mteb/arxiv-clustering-s2s
name: MTEB ArxivClusteringS2S
config: default
split: test
revision: f910caf1a6075f7329cdf8c1a6135696f37dbd53
metrics:
- type: v_measure
value: 43.19453389182364
- task:
type: Reranking
dataset:
type: mteb/askubuntudupquestions-reranking
name: MTEB AskUbuntuDupQuestions
config: default
split: test
revision: 2000358ca161889fa9c082cb41daa8dcfb161a54
metrics:
- type: map
value: 64.46555939623092
- type: mrr
value: 77.82361605768807
- task:
type: STS
dataset:
type: mteb/biosses-sts
name: MTEB BIOSSES
config: default
split: test
revision: d3fb88f8f02e40887cd149695127462bbcf29b4a
metrics:
- type: cos_sim_pearson
value: 84.9554128814735
- type: cos_sim_spearman
value: 84.65373612172036
- type: euclidean_pearson
value: 83.2905059954138
- type: euclidean_spearman
value: 84.52240782811128
- type: manhattan_pearson
value: 82.99533802997436
- type: manhattan_spearman
value: 84.20673798475734
- task:
type: Classification
dataset:
type: mteb/banking77
name: MTEB Banking77Classification
config: default
split: test
revision: 0fd18e25b25c072e09e0d92ab615fda904d66300
metrics:
- type: accuracy
value: 87.78896103896103
- type: f1
value: 87.77189310964883
- task:
type: Clustering
dataset:
type: mteb/biorxiv-clustering-p2p
name: MTEB BiorxivClusteringP2P
config: default
split: test
revision: 65b79d1d13f80053f67aca9498d9402c2d9f1f40
metrics:
- type: v_measure
value: 39.714538337650495
- task:
type: Clustering
dataset:
type: mteb/biorxiv-clustering-s2s
name: MTEB BiorxivClusteringS2S
config: default
split: test
revision: 258694dd0231531bc1fd9de6ceb52a0853c6d908
metrics:
- type: v_measure
value: 36.90108349284447
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackAndroidRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 32.795
- type: map_at_10
value: 43.669000000000004
- type: map_at_100
value: 45.151
- type: map_at_1000
value: 45.278
- type: map_at_3
value: 40.006
- type: map_at_5
value: 42.059999999999995
- type: mrr_at_1
value: 39.771
- type: mrr_at_10
value: 49.826
- type: mrr_at_100
value: 50.504000000000005
- type: mrr_at_1000
value: 50.549
- type: mrr_at_3
value: 47.115
- type: mrr_at_5
value: 48.832
- type: ndcg_at_1
value: 39.771
- type: ndcg_at_10
value: 50.217999999999996
- type: ndcg_at_100
value: 55.454
- type: ndcg_at_1000
value: 57.37
- type: ndcg_at_3
value: 44.885000000000005
- type: ndcg_at_5
value: 47.419
- type: precision_at_1
value: 39.771
- type: precision_at_10
value: 9.642000000000001
- type: precision_at_100
value: 1.538
- type: precision_at_1000
value: 0.198
- type: precision_at_3
value: 21.268
- type: precision_at_5
value: 15.536
- type: recall_at_1
value: 32.795
- type: recall_at_10
value: 62.580999999999996
- type: recall_at_100
value: 84.438
- type: recall_at_1000
value: 96.492
- type: recall_at_3
value: 47.071000000000005
- type: recall_at_5
value: 54.079
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackEnglishRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 32.671
- type: map_at_10
value: 43.334
- type: map_at_100
value: 44.566
- type: map_at_1000
value: 44.702999999999996
- type: map_at_3
value: 40.343
- type: map_at_5
value: 41.983
- type: mrr_at_1
value: 40.764
- type: mrr_at_10
value: 49.382
- type: mrr_at_100
value: 49.988
- type: mrr_at_1000
value: 50.03300000000001
- type: mrr_at_3
value: 47.293
- type: mrr_at_5
value: 48.51
- type: ndcg_at_1
value: 40.764
- type: ndcg_at_10
value: 49.039
- type: ndcg_at_100
value: 53.259
- type: ndcg_at_1000
value: 55.253
- type: ndcg_at_3
value: 45.091
- type: ndcg_at_5
value: 46.839999999999996
- type: precision_at_1
value: 40.764
- type: precision_at_10
value: 9.191
- type: precision_at_100
value: 1.476
- type: preci

查看官方来源