Acerca de
The Live Training Monitor skill provides a comprehensive framework for tracking machine learning model training within the Claude Code environment. It streamlines the process of visualizing metrics through support for TensorBoard, Weights & Biases, and terminal-based ASCII dashboards. By offering standardized paths for identifying logging backends and setting up alerts for critical issues like NaN values or gradient explosions, this skill ensures that data scientists can maintain high visibility over their experiments and respond quickly to training instabilities.
