Hyperparameter

Hyperparameter                                        Machine Learning

noun

Definition: A configuration value set before training (or controlled externally) that governs the learning process or model structure, such as learning rate, depth, or batch size [Google ML Glossary]. 

Example in context:Adding the number of prompts as a hyperparameter could introduce additional complexity to the model and the training process.” [Liang et al. 2024]

Related terms: training hyperparameter, hyperparameter tuning

 

Добавить комментарий 0

Ваш электронный адрес не будет опубликован. Обязательные поля помечены *