Part 1 of 2 in blog series in which we explore the limits of LLMs with respect to memory overhead and context windows. https://revelry.co/insights/artificial-intelligence/llms-large-context-windows/