MODEL OPTIMIZATION OF GENERATIVE AI MODELS: MAKING AI FASTER

Abstract

The field of deep learning has been advancing at a breakneck speed recently, and one can now find examples of its use in every facet of present-day human existence. It is now absolutely necessary to optimize these models in such a way that they can be put to use in real life since every month hundreds of new models are introduced and their capabilities continue to advance with each new model. Most of the time, modelers find it challenging to productionalize their models due to constraints imposed by the hardware or a low throughput rate brought on by models that have not been optimized. In this work, we discuss some of the architecturally enhanced algorithms and innovative approaches such as dynamic networks along with ML frameworks enabling deep learning model optimization across various platforms and devices.