Answer: MLflow Tracking

MLflow Tracking is the component designed for logging parameters, metrics, and models during an experiment run. It serves as a centralized platform for tracking and logging experiments, facilitating the comparison and reproduction of results. Parameters (model inputs), metrics (evaluation outcomes), and models (trained models) can all be logged using MLflow Tracking. This component is essential for monitoring experiment progress and comparing results, offering visualization tools to aid in understanding model performance. While MLflow Projects focus on code packaging and execution, MLflow Models on model management and deployment, and MLflow Registry on model versioning and collaboration, none of these are specifically tasked with the logging function during experiment runs, a role exclusively filled by MLflow Tracking.

Author: LeetQuiz Editorial Team