In machine learning, normalizing flow is a technique that helps create complex distributions from simpler ones. Flows for physics modeling will be discussed in this seminar, with an emphasis on the final stages of the physics analysis. The talk will begin with an introduction to the fundamental theoretical ideas of flow design and then go on to describe two different architectures. There will be a demonstration of flow modeling using a physics dataset. At the end, the precision and reliability of the generated data will be discussed.