semantic_router.encoders.openai

OpenAIEncoder Objects

class OpenAIEncoder(DenseEncoder)

OpenAI encoder class for generating embeddings using OpenAI API.

The OpenAIEncoder class is a subclass of DenseEncoder and utilizes the OpenAI API to generate embeddings for given documents. It requires an OpenAI API key.

token_limit

default value, should be replaced by config

__init__

def __init__(name: Optional[str] = None,
             openai_base_url: Optional[str] = None,
             openai_api_key: Optional[str] = None,
             openai_org_id: Optional[str] = None,
             dimensions: Union[int, NotGiven] = NotGiven(),
             max_retries: int = 3)

Initialize the OpenAIEncoder.

Arguments:

• name (str): The name of the embedding model to use.
• openai_base_url (str): The base URL for the OpenAI API.
• openai_api_key (str): The OpenAI API key.
• openai_org_id (str): The OpenAI organization ID.
• dimensions (int): The dimensions of the embeddings.
• str0 (int): The maximum number of retries for the OpenAI API call.

__call__

def __call__(docs: List[str], truncate: bool = True) -> List[List[float]]

Encode a list of text documents into embeddings using OpenAI API.

Arguments:

• docs: List of text documents to encode.
• truncate: Whether to truncate the documents to token limit. If False and a document exceeds the token limit, an error will be raised.

Returns:

List of embeddings for each document.

acall

async def acall(docs: List[str], truncate: bool = True) -> List[List[float]]

Encode a list of text documents into embeddings using OpenAI API asynchronously.

Arguments:

• docs: List of text documents to encode.
• truncate: Whether to truncate the documents to token limit. If False and a document exceeds the token limit, an error will be raised.

Returns:

List of embeddings for each document.