Vectors for Semantic Search
You will need to create a vector for the sentence transformers that allow semantic searches. There are a few specific requirements but otherwise it's like setting up any other vector.
For support creating a vector see our guide on creating a new vector.
New vector
There are couple of specific requirements depending on the transformer you are using.
All-mpnet-base-v2 - 768 Dimensions
Type: Dense Vector
Element Type: Select Float from the dropdown list.
Dimensions: 768
Similarity Metric: Select dot_product from the dropdown list.
Count: 10
Index: This must be enabled.

Multi-qa-mpnet-base-dot-v1 - 768 Dimensions
Type: Dense Vector
Element Type: Select Float from the dropdown list.
Dimensions: 768
Similarity Metric: Select dot_product from the dropdown list.
Count: 10
Index: This must be enabled.
