Selecting the retrieval service model

You can use the retrieval service to initiate a search query and retrieve results for enhancing AI-generated responses.

Model selection is a key step in configuring Retrieval-Augmented Generation (RAG) workflows. These models are used to generate responses based on the documents retrieved from your knowledge base.

To configure watsonx.data for selecting the retrieval service model:

1. Log in to the watsonx.data console.
2. From the navigation menu, select Configurations.
3. Click the Retrieval service model tile.
4. Select the model and click Select in the confirmation window that appears. The available model options are:
   • granite-3-8b-instruct
   • llama-3-3-70b-instruct

The nofication message Success appears when the model selection is successful.