Hi! I'm trying to run llamaindex fully local using Ollama as an embedding model. Has anyone managed to do this?

Hi everyone! I was wondering is there a way to remove repeated results in the retrieval pipeline?