Glossary term
Glossary term
Infrastructure and Serving
A high-level TensorFlow API for reading data and transforming it into a form that a machine learning algorithm requires. A tf.data.Dataset object represents a sequence of elements, in which each element contains one or more Tensors. A tf.data.Iterator object provides access to the elements of a Dataset.
Created for this library
A computer vision team uses the tf.data Dataset API to prefetch and decode JPEGs in parallel with training, keeping the GPU saturated.
A speech recognition team uses tf.data to stream audio files from cloud storage so training is not bottlenecked by disk reads.
An ML platform team standardizes input pipelines on tf.data so feature transformations are consistent across training and serving.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License