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 creating these models, but in implementing them efficiently in practical scenarios. This is where machine learning inference becomes crucial, emerging as a primary concern for research