Elasticsearch嵌入式向量¶

如果您在colab上打开这个笔记本，您可能需要安装LlamaIndex 🦙。

In [ ]:

%pip install llama-index-vector-stores-elasticsearch
%pip install llama-index-embeddings-elasticsearch

%pip install llama-index-vector-stores-elasticsearch %pip install llama-index-embeddings-elasticsearch

In [ ]:

!pip install llama-index

!pip install llama-index

In [ ]:

# 导入

from llama_index.embeddings.elasticsearch import ElasticsearchEmbedding
from llama_index.vector_stores.elasticsearch import ElasticsearchStore
from llama_index.core import StorageContext，VectorStoreIndex
from llama_index.core import Settings

# 导入 from llama_index.embeddings.elasticsearch import ElasticsearchEmbedding from llama_index.vector_stores.elasticsearch import ElasticsearchStore from llama_index.core import StorageContext，VectorStoreIndex from llama_index.core import Settings

In [ ]:

# 获取凭证并创建嵌入

import os

# 从环境变量中获取主机地址，默认为localhost:9200
host = os.environ.get("ES_HOST", "localhost:9200")
# 从环境变量中获取用户名，默认为elastic
username = os.environ.get("ES_USERNAME", "elastic")
# 从环境变量中获取密码，默认为changeme
password = os.environ.get("ES_PASSWORD", "changeme")
# 从环境变量中获取索引名称，默认为your-index-name
index_name = os.environ.get("INDEX_NAME", "your-index-name")
# 从环境变量中获取模型ID，默认为your-model-id
model_id = os.environ.get("MODEL_ID", "your-model-id")

# 通过凭证创建嵌入
embeddings = ElasticsearchEmbedding.from_credentials(
    model_id=model_id, es_url=host, es_username=username, es_password=password
)

# 获取凭证并创建嵌入 import os # 从环境变量中获取主机地址，默认为localhost:9200 host = os.environ.get("ES_HOST", "localhost:9200") # 从环境变量中获取用户名，默认为elastic username = os.environ.get("ES_USERNAME", "elastic") # 从环境变量中获取密码，默认为changeme password = os.environ.get("ES_PASSWORD", "changeme") # 从环境变量中获取索引名称，默认为your-index-name index_name = os.environ.get("INDEX_NAME", "your-index-name") # 从环境变量中获取模型ID，默认为your-model-id model_id = os.environ.get("MODEL_ID", "your-model-id") # 通过凭证创建嵌入 embeddings = ElasticsearchEmbedding.from_credentials( model_id=model_id, es_url=host, es_username=username, es_password=password )

In [ ]:

# 设置全局设置
Settings.embed_model = embeddings
Settings.chunk_size = 512

# 设置全局设置 Settings.embed_model = embeddings Settings.chunk_size = 512

In [ ]:

# 用于与elasticsearch向量存储一起使用

vector_store = ElasticsearchStore(
    index_name=index_name, es_url=host, es_user=username, es_password=password
)

storage_context = StorageContext.from_defaults(vector_store=vector_store)

index = VectorStoreIndex.from_vector_store(
    vector_store=vector_store,
    storage_context=storage_context,
)

query_engine = index.as_query_engine()


response = query_engine.query("hello world")

# 用于与elasticsearch向量存储一起使用 vector_store = ElasticsearchStore( index_name=index_name, es_url=host, es_user=username, es_password=password ) storage_context = StorageContext.from_defaults(vector_store=vector_store) index = VectorStoreIndex.from_vector_store( vector_store=vector_store, storage_context=storage_context, ) query_engine = index.as_query_engine() response = query_engine.query("hello world")