GraphRAG Chat
Last updated
Last updated
The GraphRAG Chat tab is a place where you can test your active RAG models. This feature allows users to query their model and receive responses based on them. The chat system searches through the graph knowledge to provide relevant answers.
Once you have an active RAG model, you can test it via GraphRAG Chat, using the following steps:
Select a model - choose a graph model from the list of active models.
Enter a query - in the question field, type a query or request information related to your dataset. A default prompt is prefilled in the input field. Click Show Prompt to preview or modify it before submitting your query.
This test version is optimized for English language queries.
The system processes the query and provides an answer based on the retrieved knowledge.
On the right side, you will find retrieved text panel. It displays details of the source text used to generate the response, including matched entities, confidence scores and related data nodes. You can click on the result to preview the graph for it.
Additionally, if the model remains inactive for 15 minutes, it enters a sleep mode. When this happens, the first query after inactivity may take longer to process, as it wakes the model up. The wake-up process is triggered by sending a new query.
In this example, a RAG model was built using a file containing basic information about Marie Curie.
User query: When was Marie Curie born?
System response: Marie Curie was born in 1867.
The system processes the query and returns the following response.
Underneath: a preview of the graph visualization of the retrieved knowledge for this node.