DeepsetNvidiaTextEmbedder

Embed strings of text using embedding models by NVIDIA Triton on optimized hardware.

Basic Information

• Type: deepset_cloud_custom_nodes.embedders.nvidia.text_embedder.DeepsetNvidiaTextEmbedder
• Components it most often connects with:
  • Input: DeepsetNvidiaTextEmbedder receives the query to embed from Input.
  • Embedding Retrievers: DeepsetNvidiaTextEmbedder can send the embedded query to an Embedding Retriever that uses it to find matching documents.

Inputs

Parameter	Type	Default	Description
text	str		The text to embed.

Outputs

Parameter	Type	Default	Description
embedding	List[float]		Embeddng of the text.
meta	Dict[str, Any]		Metadata on usage statistics.

Overview

NvidiaTextEmbedder uses NVIDIA Triton models to embed a string, such as a query. This is useful if your pipeline performs vector-based retrieval. The Embedding Retriever can then used the embedded query to find matching documents in the document store.

This component runs on optimized hardware in deepset AI Platform, which means it doesn't work if you export it to a local Python file. If you're planning to export, use SentenceTransformersTextEmbedder instead.

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.

Usage Example

Initializing the Component

components:
  DeepsetNvidiaTextEmbedder:
    type: embedders.nvidia.text_embedder.DeepsetNvidiaTextEmbedder
    init_parameters:

Using the Component in a Pipeline

This is an example of a DeepsetNvidiaTextEmbedder used in a query pipeline. It receives the text to embed from Input and then sends the embedded query to OpenSearchEmbeddingRetriever:

Here's the YAML configuration:

components:
  DeepsetNvidiaTextEmbedder:
    type: deepset_cloud_custom_nodes.embedders.nvidia.text_embedder.DeepsetNvidiaTextEmbedder
    init_parameters:
      model: intfloat/multilingual-e5-base
      prefix: ''
      suffix: ''
      truncate: null
      normalize_embeddings: true
      timeout: null
      backend_kwargs: null
  OpenSearchEmbeddingRetriever:
    type: haystack_integrations.components.retrievers.opensearch.embedding_retriever.OpenSearchEmbeddingRetriever
    init_parameters:
      document_store:
        type: haystack_integrations.document_stores.opensearch.document_store.OpenSearchDocumentStore
        init_parameters:
          use_ssl: true
          verify_certs: false
          hosts:
            - ${OPENSEARCH_HOST}
          http_auth:
            - ${OPENSEARCH_USER}
            - ${OPENSEARCH_PASSWORD}
          embedding_dim: 1024
          similarity: cosine
      filters: null
      top_k: 10
      filter_policy: replace
      custom_query: null
      raise_on_failure: true
      efficient_filtering: false
connections:
  - sender: DeepsetNvidiaTextEmbedder.embedding
    receiver: OpenSearchEmbeddingRetriever.query_embedding
max_runs_per_component: 100
metadata: {}
inputs:
  query:
    - DeepsetNvidiaTextEmbedder.text

Parameters

Init Parameters

These are the parameters you can configure in Pipeline Builder:

Parameter	Type	Default	Description
model	DeepsetNVIDIAEmbeddingModels	DeepsetNVIDIAEmbeddingModels.INTFLOAT_MULTILINGUAL_E5_BASE	The model to use for calculating embeddings. Choose the model from the list on the component card.
prefix	str		A string to add at the beginning of the string being embedded. Can be used to prepend the text with an instruction, as required by some embedding models, such as E5 and bge.
suffix	str		A string to add at the end of the string being embedded.
truncate	Optional[EmbeddingTruncateMode]	None	Specifies how to truncate inputs longer than the maximum token length. Possible options are: START, END, NONE. If set to START, the input is truncated from the start. If set to END, the input is truncated from the end. If set to NONE, returns an error if the input is too long.
normalize_embeddings	bool	True	Whether to normalize the embeddings. Normalization is done by dividing the embedding by its L2 norm.
timeout	Optional[float]	None	Timeout for request calls in seconds.
backend_kwargs	Optional[Dict[str, Any]]	None	Keyword arguments to further customize model behavior.

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	Default	Description
text	str		The text to embed.