Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.neum.ai/llms.txt

Use this file to discover all available pages before exploring further.

The JSONLoader class handles the loading of JSON files into the Neum AI ecosystem by creating NeumDocument objects. It offers flexible data extraction, accommodating various JSON structures.

​
Properties

Required properties:
  • None
Optional properties:
  • id_key: The key used to identify the ID field within the JSON structure. Defaults to “id”. Provide a value if you have a different key for your objects.
Available metadata:
  • custom: Metadata fields can be customized based on the contents of the JSON object. Simply pass a list of columns. (i.e. [“field1” , “field2”])
Available content:
  • custom: Content is dynamically derived from the JSON object based on specified keys or entire row data. Simply pass a list of columns. (i.e. [“field1” , “field2”])
To access nested fields within the JSON objects, you can use a dot notation. For example: “field1.field2”
from neumai.Shared import Selector
from neumai.Loaders import JSONLoader

json_loader = JSONLoader(
    loader_information={
        "id_key": "id"
    },
    selector=Selector(
        to_embed=["data"],
        to_metadata=["id", "category"]
    )
)