Hyperparameter Tuning - CI/CD Glossary Definition
Hyperparameter tuning searches over pre-training settings like learning rate to maximize validation performance.
Hyperparameter tuning searches over model configuration values (learning rate, tree depth, regularization) that are set before training to find the combination that maximizes validation performance.
Hyperparameters are the knobs you set before training; tuning automates the search for the best combination.
Search strategies
Grid search tries every combination, random search samples, and Bayesian optimization uses past results to guide the next trial. Each trial is logged via experiment tracking.
Related guides
Experiment Tracking - CI/CD Glossary DefinitionExperiment Tracking: Experiment tracking records the parameters, code version, data version, metrics, and art…
Model Validation - CI/CD Glossary DefinitionModel Validation: Model validation evaluates a trained model on held-out data against quality thresholds (acc…
Validation Set - CI/CD Glossary DefinitionValidation Set: A validation set is data held out from training and used to tune hyperparameters and select m…