Classifier Training Options

After you create a new classifier, you can set training options for it.

Set or modify training options using the actions SetClassifierTrainingOption and UnsetClassifierTrainingOption. When you change a training option all training associated with the classifier becomes invalid and you must retrain the classifier using the BuildClassifier action. For more information about these actions, refer to the Media Server Reference.

You can set the following training options:

Training Option Description Default value
batchsize The number of images to train concurrently. The value must be 1,2,4,8,16, or 32. Larger batch sizes result in slightly faster training but require more memory (more GPU memory if you are training with a GPU). If you are training with a GPU that has limited memory you might need to reduce the batch size. 32
iterations The number of iterations to perform when training a neural network classifier. For information about how to set this training option, see Training Requirements. 500

For an example that shows how to add a new classifier and set training options, see Create a Classifier.


_FT_HTML5_bannerTitle.htm