A cider-press workshop trains a grade predictor and watches a single number fall as predicted juice acidity gets closer to lab truth. What is that number?
Select an answer to reveal the explanation.
Short Explanation
The number that falls as predicted acidity gets closer to lab truth is the training loss. Held-out accuracy is a later report, GPU utilization is a hardware stat, and batch size is not the error signal.
Full Explanation
The training objective that falls as predictions match labels is the loss. Held-out accuracy is a later evaluation metric, and GPU utilization is a hardware stat. Neither is the signal that updates weights.