Deep diving into RAG architectures for custom LLMs. Sharing my experiments with fine-tuning open-source models. x.com/Scourge35596893