I want to understand if i can replicate the graphrag paper in under 30 mins ? In usual rag, a question is asked and it triggers a vector search or semantic search. The results are then ranked and reranked and sent to the llm for creating a better answer.

In a graph rag, we use llm to create the graph with the help of llm.