Vertex

VertexEmbeddingMode

Bases: str, Enum

VertexAI embedding mode.

Attributes:

Name	Type	Description
DEFAULT_MODE	str	The default embedding mode, for older models before August 2023, that does not support task_type
CLASSIFICATION_MODE	str	Optimizes embeddings for classification tasks.
CLUSTERING_MODE	str	Optimizes embeddings for clustering tasks.
SEMANTIC_SIMILARITY_MODE	str	Optimizes embeddings for tasks that require assessments of semantic similarity.
RETRIEVAL_MODE	str	Optimizes embeddings for retrieval tasks, including search and document retrieval.

Source code in llama_index/embeddings/vertex/base.py

@deprecated.deprecated(
    reason=(
        "Should use `llama-index-embeddings-google-genai` instead, using Google's latest unified SDK. "
        "See: https://docs.llamaindex.ai/en/stable/examples/embeddings/google_genai/"
    )
)
class VertexEmbeddingMode(str, Enum):
    """
    VertexAI embedding mode.

    Attributes:
        DEFAULT_MODE (str): The default embedding mode, for older models before August 2023,
                            that does not support task_type
        CLASSIFICATION_MODE (str): Optimizes embeddings for classification tasks.
        CLUSTERING_MODE (str): Optimizes embeddings for clustering tasks.
        SEMANTIC_SIMILARITY_MODE (str): Optimizes embeddings for tasks that require assessments of semantic similarity.
        RETRIEVAL_MODE (str): Optimizes embeddings for retrieval tasks, including search and document retrieval.

    """

    DEFAULT_MODE = "default"
    CLASSIFICATION_MODE = "classification"
    CLUSTERING_MODE = "clustering"
    SEMANTIC_SIMILARITY_MODE = "similarity"
    RETRIEVAL_MODE = "retrieval"

options: members: - VertexMultiModalEmbedding - VertexTextEmbedding