nomic embed text v1

提供商nomic-ai
分类sentence-similarity
许可证apache-2.0
下载量1.6K
星标0

简介

nomic-embed-text-v1 是一款高性能的开源文本向量化模型，旨在将文本转化为高维向量以实现语义搜索和相似度计算。相比于闭源的 OpenAI embedding 接口，它最大的优势在于开源且支持长文本输入，能有效处理更复杂的文档片段。对于国内开发者而言，它是构建本地 RAG（检索增强生成）知识库的理想选择，可以替代昂贵的 API 调用，在保证检索精度的同时，显著降低部署成本并提升数据隐私性。上手难度较低，可通过 Hugging Face 或 Ollama 快速集成到现有工作流中。

核心亮点

  • 支持长文本输入，大幅提升 RAG 检索召回率
  • Apache-2.0 开源协议，可自由用于商业部署
  • 语义理解精准，是构建本地知识库的优质底座
  • 兼容主流向量数据库，替代闭源 API 降低成本

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("nomic-ai/nomic-embed-text-v1")
tokenizer = AutoTokenizer.from_pretrained("nomic-ai/nomic-embed-text-v1")

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 nomic-ai/nomic-embed-text-v1

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 nomic-ai/nomic-embed-text-v1 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('nomic-ai/nomic-embed-text-v1')

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/nomic-ai/nomic-embed-text-v1

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/nomic-ai/nomic-embed-text-v1

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('nomic-ai/nomic-embed-text-v1')
tokenizer = AutoTokenizer.from_pretrained('nomic-ai/nomic-embed-text-v1')

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 nomic-ai/nomic-embed-text-v1

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 nomic-ai/nomic-embed-text-v1 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('nomic-ai/nomic-embed-text-v1')

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/nomic-ai/nomic-embed-text-v1.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/nomic-ai/nomic-embed-text-v1.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', 'nomic-ai/nomic-embed-text-v1')

完整文档

来源: HuggingFace

---
library_name: sentence-transformers
pipeline_tag: sentence-similarity
tags:

  • feature-extraction

  • sentence-similarity

  • mteb

  • transformers

  • transformers.js

model-index:
  • name: epoch_0_model

results:
- task:
type: Classification
dataset:
type: mteb/amazon_counterfactual
name: MTEB AmazonCounterfactualClassification (en)
config: en
split: test
revision: e8379541af4e31359cca9fbcf4b00f2671dba205
metrics:
- type: accuracy
value: 76.8507462686567
- type: ap
value: 40.592189159090495
- type: f1
value: 71.01634655512476
- task:
type: Classification
dataset:
type: mteb/amazon_polarity
name: MTEB AmazonPolarityClassification
config: default
split: test
revision: e2d317d38cd51312af73b3d32a06d1a08b442046
metrics:
- type: accuracy
value: 91.51892500000001
- type: ap
value: 88.50346762975335
- type: f1
value: 91.50342077459624
- task:
type: Classification
dataset:
type: mteb/amazon_reviews_multi
name: MTEB AmazonReviewsClassification (en)
config: en
split: test
revision: 1399c76144fd37290681b995c656ef9b2e06e26d
metrics:
- type: accuracy
value: 47.364
- type: f1
value: 46.72708080922794
- task:
type: Retrieval
dataset:
type: arguana
name: MTEB ArguAna
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 25.178
- type: map_at_10
value: 40.244
- type: map_at_100
value: 41.321999999999996
- type: map_at_1000
value: 41.331
- type: map_at_3
value: 35.016999999999996
- type: map_at_5
value: 37.99
- type: mrr_at_1
value: 25.605
- type: mrr_at_10
value: 40.422000000000004
- type: mrr_at_100
value: 41.507
- type: mrr_at_1000
value: 41.516
- type: mrr_at_3
value: 35.23
- type: mrr_at_5
value: 38.15
- type: ndcg_at_1
value: 25.178
- type: ndcg_at_10
value: 49.258
- type: ndcg_at_100
value: 53.776
- type: ndcg_at_1000
value: 53.995000000000005
- type: ndcg_at_3
value: 38.429
- type: ndcg_at_5
value: 43.803
- type: precision_at_1
value: 25.178
- type: precision_at_10
value: 7.831
- type: precision_at_100
value: 0.979
- type: precision_at_1000
value: 0.1
- type: precision_at_3
value: 16.121
- type: precision_at_5
value: 12.29
- type: recall_at_1
value: 25.178
- type: recall_at_10
value: 78.307
- type: recall_at_100
value: 97.866
- type: recall_at_1000
value: 99.57300000000001
- type: recall_at_3
value: 48.364000000000004
- type: recall_at_5
value: 61.451
- task:
type: Clustering
dataset:
type: mteb/arxiv-clustering-p2p
name: MTEB ArxivClusteringP2P
config: default
split: test
revision: a122ad7f3f0291bf49cc6f4d32aa80929df69d5d
metrics:
- type: v_measure
value: 45.93034494751465
- task:
type: Clustering
dataset:
type: mteb/arxiv-clustering-s2s
name: MTEB ArxivClusteringS2S
config: default
split: test
revision: f910caf1a6075f7329cdf8c1a6135696f37dbd53
metrics:
- type: v_measure
value: 36.64579480054327
- task:
type: Reranking
dataset:
type: mteb/askubuntudupquestions-reranking
name: MTEB AskUbuntuDupQuestions
config: default
split: test
revision: 2000358ca161889fa9c082cb41daa8dcfb161a54
metrics:
- type: map
value: 60.601310529222054
- type: mrr
value: 75.04484896451656
- task:
type: STS
dataset:
type: mteb/biosses-sts
name: MTEB BIOSSES
config: default
split: test
revision: d3fb88f8f02e40887cd149695127462bbcf29b4a
metrics:
- type: cos_sim_pearson
value: 88.57797718095814
- type: cos_sim_spearman
value: 86.47064499110101
- type: euclidean_pearson
value: 87.4559602783142
- type: euclidean_spearman
value: 86.47064499110101
- type: manhattan_pearson
value: 87.7232764230245
- type: manhattan_spearman
value: 86.91222131777742
- task:
type: Classification
dataset:
type: mteb/banking77
name: MTEB Banking77Classification
config: default
split: test
revision: 0fd18e25b25c072e09e0d92ab615fda904d66300
metrics:
- type: accuracy
value: 84.5422077922078
- type: f1
value: 84.47657456950589
- task:
type: Clustering
dataset:
type: mteb/biorxiv-clustering-p2p
name: MTEB BiorxivClusteringP2P
config: default
split: test
revision: 65b79d1d13f80053f67aca9498d9402c2d9f1f40
metrics:
- type: v_measure
value: 38.48953561974464
- task:
type: Clustering
dataset:
type: mteb/biorxiv-clustering-s2s
name: MTEB BiorxivClusteringS2S
config: default
split: test
revision: 258694dd0231531bc1fd9de6ceb52a0853c6d908
metrics:
- type: v_measure
value: 32.75995857510105
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackAndroidRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 30.008000000000003
- type: map_at_10
value: 39.51
- type: map_at_100
value: 40.841
- type: map_at_1000
value: 40.973
- type: map_at_3
value: 36.248999999999995
- type: map_at_5
value: 38.096999999999994
- type: mrr_at_1
value: 36.481
- type: mrr_at_10
value: 44.818000000000005
- type: mrr_at_100
value: 45.64
- type: mrr_at_1000
value: 45.687
- type: mrr_at_3
value: 42.036
- type: mrr_at_5
value: 43.782
- type: ndcg_at_1
value: 36.481
- type: ndcg_at_10
value: 45.152
- type: ndcg_at_100
value: 50.449
- type: ndcg_at_1000
value: 52.76499999999999
- type: ndcg_at_3
value: 40.161
- type: ndcg_at_5
value: 42.577999999999996
- type: precision_at_1
value: 36.481
- type: precision_at_10
value: 8.369
- type: precision_at_100
value: 1.373
- type: precision_at_1000
value: 0.186
- type: precision_at_3
value: 18.693
- type: precision_at_5
value: 13.533999999999999
- type: recall_at_1
value: 30.008000000000003
- type: recall_at_10
value: 56.108999999999995
- type: recall_at_100
value: 78.55499999999999
- type: recall_at_1000
value: 93.659
- type: recall_at_3
value: 41.754999999999995
- type: recall_at_5
value: 48.296
- task:
type: Retrieval
dataset:
type: BeIR/cqadupstack
name: MTEB CQADupstackEnglishRetrieval
config: default
split: test
revision: None
metrics:
- type: map_at_1
value: 30.262
- type: map_at_10
value: 40.139
- type: map_at_100
value: 41.394
- type: map_at_1000
value: 41.526
- type: map_at_3
value: 37.155
- type: map_at_5
value: 38.785
- type: mrr_at_1
value: 38.153
- type: mrr_at_10
value: 46.369
- type: mrr_at_100
value: 47.072
- type: mrr_at_1000
value: 47.111999999999995
- type: mrr_at_3
value: 44.268
- type: mrr_at_5
value: 45.389
- type: ndcg_at_1
value: 38.153
- type: ndcg_at_10
value: 45.925
- type: ndcg_at_100
value: 50.394000000000005
- type: ndcg_at_1000
value: 52.37500000000001
- type: ndcg_at_3
value: 41.754000000000005
- type: ndcg_at_5
value: 43.574
- type: precision_at_1
value: 38.153
- type: precision_at_10
value: 8.796
- type: precision_at_100

查看官方来源