A mathematical function that quantifies the difference between the model's prediction and the ground truth. The goal of training is to minimize this value.
The primary metric for optimization algorithms.
Cross-Entropy and MSE are the most common variants.