Evaluating a model - AWS DeepComposer

Evaluating a model

By examining trained models, you can learn what a useful model is and the features that a model should include. Always evaluate a model to understand the predictions that it generates. To evaluate a model, you can examine the changes in the loss function of your model over time. You can also explore the training output per 50th epoch on the model details page. This topic covers what makes an effective model and which hyperparameters are available for different models.


This topic assumes that you chose the hyperparameters documented in the topic on training a custom MuseGAN model. If you chose another model or different hyperparameters, your results will differ from those shown in this topic.