Get hands-on experience with 20+ free Google Cloud products and $300 in free credit for new customers.

For Blended search, datastore spec with boost spec says Invalid

Hello There, 

We are been actively using Vertex ai Agent builder search and recently we started looking into blended search. 

Our requirement is too boost specific content type present in one of the datastore, hence we planned to use datastore spec along with boost spec as per documentation mentioned here: https://cloud.google.com/generative-ai-app-builder/docs/reference/rest/v1/DataStoreSpec 

Here is the my payload for your reference:

{
    "query": {
          "text": "xyz",
      },
    "pageSize": 10,
    "dataStoreSpecs": [
        {
            "dataStore": "projects/xxx/locations/us/collections/default_collection/dataStores/xxx",
            "boostSpec": {
                "conditionBoostSpecs": [
                    {
                        "condition": "project_status_nme: ANY('Active')",
                        "boost": 1
                    }
                ]
            }
        }
    ]
}
but we got  400 bad request. 
{
"error": {
"code": 400,
"message": "Setting boost_spec proto fields is not allowed for multi-datastore search.",
"status": "INVALID_ARGUMENT"
}
}
 
If we remove boostSpec in the payload, it works. Can you let me know what is wrong in my request? 
 
Also, I don't see the same feature been supported in python client libraries, so now I have high doubts on this feature. FYR, https://cloud.google.com/python/docs/reference/discoveryengine/latest/google.cloud.discoveryengine_v...
 

 

0 0 105
0 REPLIES 0