This example demonstrates how to train the MobileNet v1 network against the Imagenette image classification dataset.
A MobileNet (version 1) network is instantiated from the ImageClassificationModels library of standard models, and applied to an instance of the Imagenette dataset. A custom training loop is defined, and the training and test losses and accuracies for each epoch are shown during training.
To begin, you'll need the latest version of Swift for
TensorFlow
installed. Make sure you've added the correct version of swift to your path.
To train the model, run:
cd swift-models
swift run -c release MobileNetV1-Imagenette