VertexAIDocumentEmbedder

Embed documents using Vertex AI Embeddings API.

Key Features

• Embeds documents using the Vertex AI Embeddings API
• Configurable task type (RETRIEVAL_DOCUMENT, RETRIEVAL_QUERY, and others)
• Batch processing support for efficient embedding of large document sets
• Built-in retry logic for reliability
• Stores embeddings in each document's embedding field

Configuration

1. Drag the VertexAIDocumentEmbedder component onto the canvas from the Component Library.
2. Click on the component to open the configuration panel.
3. On the General tab:
   1. Enter your GCP project ID and region. Create secrets with the keys GCP_PROJECT_ID and GCP_DEFAULT_REGION. For detailed instructions, see Create Secrets.
   2. Select a model. For supported models, see the official Google documentation.
   3. Select a task type. Use RETRIEVAL_DOCUMENT when embedding documents for storage.
4. Go to the Advanced tab to configure batch size, retries, truncation, and metadata field settings.

Connections

VertexAIDocumentEmbedder accepts a list of Document objects through its documents input. It outputs a list of Document objects with embeddings stored in the embedding field.

Use this component in an indexing pipeline. Connect preprocessors like DocumentSplitter to its documents input, and connect its documents output to DocumentWriter.

Embedding Models in Query Pipelines and Indexes

The embedding model you use to embed documents in your indexing pipeline must be the same as the embedding model you use to embed the query in your query pipeline.

This means the embedders for your indexing and query pipelines must match. For example, if you use CohereDocumentEmbedder to embed your documents, you should use CohereTextEmbedder with the same model to embed your queries.

Source Code

To check this component's source code, open document_embedder.py in the Haystack Core Integrations repository.

Usage Examples

Basic Configuration

  VertexAIDocumentEmbedder:
    type: haystack_integrations.components.embedders.google_vertex.document_embedder.VertexAIDocumentEmbedder
    init_parameters:
      model: text-embedding-005
      task_type: RETRIEVAL_DOCUMENT
      gcp_region_name:
        type: env_var
        env_vars:
        - GCP_DEFAULT_REGION
        strict: false
      gcp_project_id:
        type: env_var
        env_vars:
        - GCP_PROJECT_ID
        strict: false
      batch_size: 32
      max_tokens_total: 20000
      time_sleep: 30
      retries: 3
      progress_bar: true
      embedding_separator: "\n"

This index uses VertexAIDocumentEmbedder to embed documents before storing them:

components:
  TextFileToDocument:
    type: haystack.components.converters.txt.TextFileToDocument
    init_parameters:
      encoding: utf-8
      store_full_path: false

  DocumentSplitter:
    type: haystack.components.preprocessors.document_splitter.DocumentSplitter
    init_parameters:
      split_by: sentence
      split_length: 5
      split_overlap: 1

  VertexAIDocumentEmbedder:
    type: haystack_integrations.components.embedders.google_vertex.document_embedder.VertexAIDocumentEmbedder
    init_parameters:
      model: text-embedding-005
      task_type: RETRIEVAL_DOCUMENT
      gcp_region_name:
        type: env_var
        env_vars:
          - GCP_DEFAULT_REGION
        strict: false
      gcp_project_id:
        type: env_var
        env_vars:
          - GCP_PROJECT_ID
        strict: false
      batch_size: 32
      max_tokens_total: 20000
      time_sleep: 30
      retries: 3
      progress_bar: true
      truncate_dim:
      meta_fields_to_embed:
      embedding_separator: "\n"

  DocumentWriter:
    type: haystack.components.writers.document_writer.DocumentWriter
    init_parameters:
      document_store:
        type: haystack_integrations.document_stores.opensearch.document_store.OpenSearchDocumentStore
        init_parameters:
          hosts:
          index: 'vertex-embeddings'
          max_chunk_bytes: 104857600
          embedding_dim: 768
          return_embedding: false
          method:
          mappings:
          settings:
          create_index: true
          http_auth:
          use_ssl:
          verify_certs:
          timeout:
      policy: OVERWRITE

connections:
  - sender: TextFileToDocument.documents
    receiver: DocumentSplitter.documents
  - sender: DocumentSplitter.documents
    receiver: VertexAIDocumentEmbedder.documents
  - sender: VertexAIDocumentEmbedder.documents
    receiver: DocumentWriter.documents

inputs:
  files:
    - TextFileToDocument.sources

max_runs_per_component: 100

metadata: {}

Parameters

Inputs

Parameter	Type	Description
documents	List[Document]	A list of documents to embed.

Outputs

Parameter	Type	Description
documents	List[Document]	A list of documents with embeddings.

Init Parameters

These are the parameters you can configure in Pipeline Builder:

Parameter	Type	Default	Description
model	Literal['text-embedding-004', 'text-embedding-005', ...]		Name of the model to use.
task_type	Literal['RETRIEVAL_DOCUMENT', 'RETRIEVAL_QUERY', ...]	RETRIEVAL_DOCUMENT	The type of task for which the embeddings are being generated. See Google documentation.
gcp_region_name	Optional[Secret]	Secret.from_env_var('GCP_DEFAULT_REGION', strict=False)	The default location to use when making API calls. If not set, uses us-central-1.
gcp_project_id	Optional[Secret]	Secret.from_env_var('GCP_PROJECT_ID', strict=False)	ID of the GCP project to use.
batch_size	int	32	The number of documents to process in a single batch.
max_tokens_total	int	20000	The maximum number of tokens to process in total.
time_sleep	int	30	The time to sleep between retries in seconds.
retries	int	3	The number of retries in case of failure.
progress_bar	bool	True	Whether to display a progress bar during processing.
truncate_dim	Optional[int]	None	The dimension to truncate the embeddings to, if specified.
meta_fields_to_embed	Optional[List[str]]	None	A list of metadata fields to include in the embeddings.
embedding_separator	str	\n	The separator to use between different embeddings.

Run Method Parameters

These are the parameters you can configure for the component's run() method. This means you can pass these parameters at query time through the API, in Playground, or when running a job. For details, see Modify Pipeline Parameters at Query Time.

Parameter	Type	Description
documents	List[Document]	A list of documents to embed.

Related Information

• Create Secrets
• Google Cloud Application Default Credentials