C Dac

I had a technical interview focused on Generative AI, Machine Learning, and project-based concepts. The interviewer asked me about my project and tested my understanding of fundamental concepts such as RAG, Transformers, GPUs, and LLMs. The main questions included: Explain RAG (Retrieval-Augmented Generation) and how it works. Explain the project, including its architecture and implementation. What is a Transformer, and how does it work? What is a GPU, and why is it important for deep learning? What are the different types of LLMs? If a very large LLM does not fit into the available GPU VRAM, how would you handle inference? What techniques can be used to reduce the model’s memory requirements during inference, such as quantization, CPU offloading, and model sharding?