Langchain

Modules concerning custom classes that extend LangChain's Embeddings class.

EMInvokerEmbeddings

Bases: BaseModel, Embeddings

An adapter class that enables an EMInvoker to be used as a LangChain Embeddings.

Attributes:

Name	Type	Description
em_invoker	BaseEMInvoker	The EMInvoker instance to be interacted with.

Usage example:

from gllm_inference.em_invoker.langchain import EMInvokerEmbeddings
from gllm_inference.em_invoker import OpenAIEMInvoker

em_invoker = OpenAIEMInvoker(...)
embeddings = EMInvokerEmbeddings(em_invoker=em_invoker)

aembed_documents(texts, **kwargs) async

Asynchronously embed documents using the EMInvoker.

Parameters:

Name	Type	Description	Default
texts	list[str]	The list of texts to embed.	required
**kwargs	Any	Additional keyword arguments to pass to the EMInvoker's invoke method.	{}

Returns:

Type	Description
list[Vector]	list[Vector]: List of embeddings, one for each text.

aembed_query(text, **kwargs) async

Asynchronously embed query using the EMInvoker.

Parameters:

Name	Type	Description	Default
text	str	The text to embed.	required
**kwargs	Any	Additional keyword arguments to pass to the EMInvoker's invoke method.	{}

Returns:

Name	Type	Description
Vector	Vector	Embeddings for the text.

embed_documents(texts, **kwargs)

Embed documents using the EMInvoker.

Parameters:

Name	Type	Description	Default
texts	list[str]	The list of texts to embed.	required
**kwargs	Any	Additional keyword arguments to pass to the EMInvoker's invoke method.	{}

Returns:

Type	Description
list[Vector]	list[Vector]: List of embeddings, one for each text.

embed_query(text, **kwargs)

Embed query using the EMInvoker.

Parameters:

Name	Type	Description	Default
text	str	The text to embed.	required
**kwargs	Any	Additional keyword arguments to pass to the EMInvoker's invoke method.	{}

Returns:

Name	Type	Description
Vector	Vector	Embeddings for the text.

TEIEmbeddings

Bases: BaseModel, Embeddings

A custom LangChain Embeddings class to interact with Text Embeddings Inference (TEI).

Attributes:

Name	Type	Description
url	str	The URL of the TEI service that hosts the embedding model.
api_key	str | None	The API key to the TEI service. Defaults to None.
client	InferenceClient	The client instance to interact with the TEI service.
query_prefix	str	The additional prefix to be added when embedding a query.
document_prefix	str	The additional prefix to be added when embedding documents.

Initialize with URL and API key example:

from gllm_inference.em_invoker.langchain import TEIEmbeddings

embeddings = TEIEmbeddings(url="<url-to-tei-service>", api_key="<my-api-key>")

Initialize with only URL example:

from gllm_inference.em_invoker.langchain import TEIEmbeddings

embeddings = TEIEmbeddings(url="<url-to-tei-service>")

Initialize with client example:

from gllm_inference.em_invoker.langchain import TEIEmbeddings
from huggingface_hub import InferenceClient

client = InferenceClient(model="<url-to-tei-service>", api_key="<my-api-key>")
embeddings = TEIEmbeddings(client=client)

embed_documents(texts)

Embed documents using TEI's hosted embedding model.

Parameters:

Name	Type	Description	Default
texts	list[str]	The list of texts to embed.	required

Returns:

Type	Description
list[Vector]	list[Vector]: List of embeddings, one for each text.

embed_query(text)

Embed query using TEI's hosted embedding model.

Parameters:

Name	Type	Description	Default
text	str	The text to embed.	required

Returns:

Name	Type	Description
Vector	Vector	Embeddings for the text.

validate_environment()

Validates that the TEI service URL and python package exists in environment.

The validation is done in the following order: 1. If neither url nor client is provided, an error will be raised. 2. If an invalid client is provided, an error will be raised. 3. If url is provided, it will be used to initialize the TEI service, along with an optional api_key.