Class Introduction

Function

Connects to the TEI service and provides the text-to-vector embedding function. The class inherits the langchain_core.embeddings.Embeddings API.

Prototype

from mx_rag.embedding.service import TEIEmbedding
TEIEmbedding(url, client_param, embed_mode)

Parameters

Parameter	Data Type	Required/Optional	Description
url	String	Required	TEI embedding service address. The string contains [1, 128] characters. The /v1/embed, /v1/embeddings, and /embed_sparse APIs are supported. NOTE: Currently, the embedding service created based on the TEI framework does not support the HTTPS protocol. For security purposes, you can set up an Nginx service to ensure that the service and embedding service remain within the same trusted network. When the client accesses the Nginx in HTTPS mode, the Nginx forwards a request to the embedding service.
client_param	ClientParam	Optional	HTTPS client configuration parameter. The default value is ClientParam(). For details, see ClientParam.
embed_mode	String	Optional	(Deprecated) Corresponds to the vectorization type provided by the TEI service. The value can only be sparse or dense (default). sparse indicates sparse vectorization, and dense indicates dense vectorization.

Return Value

TEIEmbedding object.

Example

from paddle.base import libpaddle
from mx_rag.embedding.service import TEIEmbedding
from mx_rag.utils import ClientParam
# Same as tei_embed = TEIEmbedding("https://ip:port/embed", client_param = ClientParam(xxx)).
tei_embed = TEIEmbedding.create(url="https://ip:port/embed",
                                client_param=ClientParam(ca_file="/path/to/ca.crt"))
print(tei_embed.embed_documents(['abc', 'bcd']))
print(tei_embed.embed_query('abc'))

Parent topic: TEIEmbedding