【python qdrant 向量数据库 完整示例代码】

测试一下python版本的dqrant向量数据库的效果,完整代码如下:

安装库

复制代码
!pip install qdrant-client>=1.1.1
!pip install -U sentence-transformers

导入

复制代码
from qdrant_client import models, QdrantClient
from sentence_transformers import SentenceTransformer

encoder = SentenceTransformer("all-MiniLM-L6-v2", device="cuda")

准备测试数据集

复制代码
documents = [
    {
        "name": "The Time Machine",
        "description": "A man travels through time and witnesses the evolution of humanity."
        * 8,
        "author": "H.G. Wells",
        "year": 1895,
    },
    {
        "name": "Ender's Game",
        "description": "A young boy is trained to become a military leader in a war against an alien race."
        * 4,
        "author": "Orson Scott Card",
        "year": 1985,
    },
    {
        "name": "Brave New World",
        "description": "A dystopian society where people are genetically engineered and conditioned to conform to a strict social hierarchy."
        * 6,
        "author": "Aldous Huxley",
        "year": 1932,
    },
] * 50000

print(len(documents))

创建存储库

复制代码
qdrant = QdrantClient(":memory:")  # 内存中
# qdrant = QdrantClient(path='./qdrant')  # 存储到本地

在数据库中创建一个collection(类似一个存储桶)

复制代码
qdrant.recreate_collection(
    collection_name="my_books",
    vectors_config=models.VectorParams(
        size=encoder.get_sentence_embedding_dimension(),  # Vector size is defined by used model
        distance=models.Distance.COSINE,
    ),
)

对文档进行向量化

复制代码
import hashlib
from tqdm import tqdm

def sha256(text):

    hash_object = hashlib.sha256()
    hash_object.update(text.encode("utf-8"))
    hash_value = hash_object.hexdigest()
    return hash_value

records = []
bs = 256
for i in tqdm(range(0, len(documents), bs)):
    docs = documents[i : i + bs]
    vectors = encoder.encode(
        [doc["description"] for doc in docs], normalize_embeddings=True
    ).tolist()

    record = [
        models.Record(id=idx, vector=vec, payload=doc)  # sha256(doc['description'])
        for idx, vec, doc in zip(range(i, i + bs), vectors, docs)
    ]

    records.extend(record)

上传到向量数据库中指定的collection

复制代码
qdrant.upload_points(
    collection_name="my_books", points=records, batch_size=128, parallel=12
)

语义搜索

复制代码
query = "Aliens attack our planet"
hits = qdrant.search(
    collection_name="my_books",
    query_vector=encoder.encode(query).tolist(),
    limit=6,
)
for hit in hits:
    print(hit.payload, "score:", hit.score)

条件搜索

search only for books from 21st century

复制代码
hits = qdrant.search(
    collection_name="my_books",
    query_vector=encoder.encode("Tyranic society").tolist(),
    query_filter=models.Filter(
        must=[models.FieldCondition(key="year", range=models.Range(gte=1980))]
    ),
    limit=3,
)
for hit in hits:
    print(hit.payload, "score:", hit.score)

参考官方GitHub

github

colab

相关推荐
screenCui1 小时前
macOS运行python程序遇libiomp5.dylib库冲突错误解决方案
开发语言·python·macos
小眼睛羊羊1 小时前
pyinstaller打包paddleocr
python
java1234_小锋1 小时前
基于Python的旅游推荐协同过滤算法系统(去哪儿网数据分析及可视化(Django+echarts))
python·数据分析·旅游
蓝婷儿1 小时前
Python 机器学习核心入门与实战进阶 Day 4 - 支持向量机(SVM)原理与分类实战
python·机器学习·支持向量机
%d%d22 小时前
python 在运行时没有加载修改后的版本
java·服务器·python
amazinging3 小时前
北京-4年功能测试2年空窗-报培训班学测开-第四十七天
python·学习·selenium
Freak嵌入式3 小时前
一文速通 Python 并行计算:13 Python 异步编程-基本概念与事件循环和回调机制
开发语言·python·嵌入式·协程·硬件·异步编程
一个天蝎座 白勺 程序猿3 小时前
Python练习(1)Python基础类型操作语法实战:20道实战题解与案例分析(上)
开发语言·python·学习
巨人张3 小时前
信息素养Python编程题
开发语言·python
站大爷IP3 小时前
Python爬虫动态IP代理报错全解析:从问题定位到实战优化
python