# Image Data Generator Tensorflow-keras uses generators to handle image loading which is nice Generates batches of image data with real-time data augmentation ```python from tensorflow.keras.preprocessing.image import ImageDataGenerator ``` There are lots of options with multiple augmentation techniques ```python train_datagen = ImageDataGenerator(rescale=1/255) ``` Creates a generator. Lots of nice options ```python train_genertor = train_datagen.flow_from_directory('/tmp/horse-or-human/', target_size=(300,300), batch_size=128, class_mode='binary') ```