Changing the Content Assistant embedding model for an object store

You can review the Content Assistant embedding model for an object store.

Before you begin

If the index state is set to active, you cannot change the embedding model for an object store. Before you change the Content Assistant embedding model for an object store, you need to disable vector indexing for the object store. For more information, see the topic Disabling vector indexing for an object store.

About this task

You can change the embedding model to improve how the Content Assistant processes text and delivers more accurate results for features like document comparison and summarization.
Important: To ensure that all documents are correctly processed and searchable with the new embedding model, you must reindex the object store after changing the model.
If you plan to use the default IBM® Content Assistant entitlement, you are authorized to use the IBM Slate embedding model only. For more information, see the topic Slate foundation models External link opens a new window or tab.

If you plan to use the IBM Content Assistant along with a IBM watsonx.ai account that you own, you can configure the object store to use any of the embedding models supported by IBM watsonx.ai. For more information, see the topic Supported encoder foundation models in IBM watsonx.ai External link opens a new window or tab.

Procedure

  1. In the Administration Console for Content Platform Engine (ACCE), go to the object store where you want to use the Content Assistant.
  2. In the left-pane, go to Search.
  3. Click New Object Store Search.
  4. In Simple View, specify the Class as Gen AI Configuration and click Run.
    The search results display a single Gen AI Configuration object.
  5. From the search results, click the object ID to access the properties for the Gen AI Configuration object.
  6. In the Properties tab, specify the embedding model in Gen AI Embedding Model Name and click Save.

What to do next

After you change the embedding model for the object store, you need to reindex the document content. For more information, see the topic Reindexing document content.