Change hyperparameters between epochs

I know about training functions with adaptive learning rates, but I need a fine control of the hyperparameters. From my experiments, it appears that all trainable parameters of a model is reset when calling 'train', but I could be doing this too simplistic. Is there a way of training for N epoch, modify the hyperparameters of the model and repeat until some end criterion decided by me?

回答(0 个)

类别

帮助中心File Exchange 中查找有关 Deep Learning Toolbox 的更多信息

产品

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by