AWS Certified AI Practitioner AIF-C01 Practice Question

Inferencing (sometimes called prediction or scoring) occurs after a model has already been trained. At this stage, the model's learned parameters remain fixed and incoming, typically unseen, data is passed through the model to obtain predictions or classifications. No parameter updates or gradient calculations take place. By contrast, adjusting model weights with labeled data, hyperparameter tuning, or any data-preparation steps belong to the training or pre-processing phases. Therefore, sending new data to a deployed model endpoint so that it can return a prediction exemplifies inferencing.