semantic_router.encoders.ollama

​

OllamaEncoder Objects

class OllamaEncoder(DenseEncoder)

from semantic_router.encoders.ollama import OllamaEncoder

encoder = OllamaEncoder(base_url="http://localhost:11434")
embeddings = encoder(["document1", "document2"])

• client - An instance of the TextEmbeddingModel client.
• type - The type of the encoder, which is “ollama”.

​

__init__

def __init__(name: Optional[str] = None,
             score_threshold: float = 0.5,
             base_url: str | None = None)

• name (str): The name of the pre-trained model to use for embedding. If not provided, the default model specified in EncoderDefault will be used.
• score_threshold (float): The threshold for similarity scores.
• base_url (str): The API endpoint for OLLAMA. If not provided, it will be retrieved from the OLLAMA_BASE_URL environment variable.

• ValueError: If the hosted base url is not provided properly or if the ollama client fails to initialize.

​

__call__

def __call__(docs: List[str]) -> List[List[float]]

• docs (List[str]): A list of strings representing the documents to embed.

• ValueError: If the Google AI Platform client is not initialized or if the API call fails.