Glossary term
Glossary term
Architecture
A mechanism by which data is loaded into a neural network.
An input generator can be thought of as a component responsible for processing raw data into tensors which are iterated over to generate batches for training, evaluation, and inference.
Created for this library
An ML platform team writes input generators for production training so disk reads, decoding, and augmentation happen in parallel with computation.
A speech recognition team builds an input generator that streams audio from cloud storage so training is not bottlenecked by IO.
A computer vision team uses an input generator with on-the-fly augmentation so each epoch sees slightly different versions of the same images.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License