Machine learning has advanced considerably in recent years, with models achieving human-level performance in diverse tasks. However, the main hurdle lies not just in developing these models, but in deploying them optimally in everyday use cases. This is where machine learning inference becomes cruci