Optimizing Rag With Llms: Exploring Chunking Techniques And Reranking For Enhanced Results