When I upload multiple documents to filestore, they are assigned unique document IDs. However, when I ask questions and get answers, I can’t find the unique ID for each cited document; only the filestore id is returned. Do I need to add the unique document ID to the metadata? I feel like this should be returned automatically. I need the unique document id for pulling out specific document I stored in my S3 bucket.
Here is the JSON object being returned for grounding metadata
{
“retrievedContext”: {
“text”: “…document excerpt…”,
“fileSearchStore”: “fileSearchStores//”
}
},
{
“retrievedContext”: {
“text”: “…another excerpt…”,
“fileSearchStore”: “fileSearchStores//”
}
}
]
}