I am trying to test out the Vectara by uploading a sample doc and setting up filtering metadata in order to do queries.
The sample doc looks like this.
{
"documentId": "doc-id-1",
"title": "Book 1",
"metadataJson": "{\"book-name\": \"Jane Eyre\", \"collection\": \"Story\", \"author\": \"Jane\"}",
"section": [
{
"text": "The answer to the ultimate of life, the universe, and everything is 42."
}
]
}
When I create a corpus with TextList, it gives error and I don’t see any Text
type in the FilterMetadata list when creating a corpus.