Quantization is a technique used to compact LLMs. What methods exist and how to quickly start using them?| TensorOps