What do you do after a model has been trained to perfection? Use it? If the answer is yes, then how do you use it? The answer you're looking for is inference. Simply put, the process of inference is what is needed to ensure that machine learning models can be used for serving the needs of actual users. Formally put, inference is the process of computing trained machine learning models efficiently to serve the user's needs. Inference can be performed on a variety of hardware types including servers, and end user devices such as phones and web browsers. As per user requirements, it can also be performed on different operating systems.
Previous chapters have focused on the process of how to build a model. This chapter will cover a detailed overview of the inference stage. First, you will cover a detailed overview of...
