Does anyone know how to split a dataset created by the dataset API (tf.data.Dataset) in Tensorflow into Test and Train?
Assuming you have all_dataset variable of tf.data.Dataset type:
all_dataset
tf.data.Dataset
test_dataset = all_dataset.take(1000) train_dataset = all_dataset.skip(1000)
Test dataset now has first 1000 elements and the rest goes for training.