This documentation contains work-in-progress information for future Elastic Stack and Cloud releases. Use the version selector to view supported release docs. It also contains some Elastic Cloud serverless information. Check out our serverless docs for more details.

« Google Gemini connector and action IBM Resilient connector and action »

› › ›

Inference connector and action

edit

Inference connector and action

edit

The inference connector uses the Elasticsearch client to send requests to an inference service. The connector uses the run connector API to send the request.

Create connectors in Kibana

edit

You can create connectors in Stack Management > Connectors. For example:

Connector configuration

edit

Inference connectors have the following configuration properties:

Name: The name of the connector.
Service: The supported inference service provider.
Task type: The inference task type, it depends on the selected service.
Inference ID: The unique identifier of the inference endpoint.
Provider configuration: Settings for service configuration.
Provider secrets: Configuration for authentication.
Task type configuration: Settings for task type configuration.

Test connectors

edit

You can test connectors using the run connector API or while creating or editing the connector in Kibana. For example:

Inference connector actions

edit

The inference actions have the following configuration properties. Properties depend on the selected task type.

Completion

edit

The following example performs a completion task on the example question.

Input

The text on which you want to perform the inference task. For example:

{
  input: 'What is Elastic?'
}

Text embedding

edit

The following example performs a text embedding task.

Input

The text on which you want to perform the inference task. For example:

{
  input: 'The sky above the port was the color of television tuned to a dead channel.',
  task_settings: {
    input_type: 'ingest'
  }
}

Input type

An optional string that overwrites the connector’s default model.

Reranking

edit

The following example performs a reranking task on the example input.

Input

The text on which you want to perform the inference task. Should be a string array. For example:

{
  input: ['luke', 'like', 'leia', 'chewy', 'r2d2', 'star', 'wars'],
  query: 'star wars main character'
}

Query

The search query text.

Sparse embedding

edit

The following example performs a sparse embedding task on the example sentence.

Input

The text on which you want to perform the inference task. For example:

{
  input: 'The sky above the port was the color of television tuned to a dead channel.'
}

Connector networking configuration

edit

Use the Action configuration settings to customize connector networking configurations, such as proxies, certificates, or TLS settings. You can apply these settings to all your connectors or use xpack.actions.customHostSettings to set per-host configurations.

« Google Gemini connector and action IBM Resilient connector and action »