Glossary term
Glossary term
Evaluation and Benchmarks
A TensorFlow API for evaluating models. For example, tf.metrics.accuracy determines how often a model's predictions match labels.
Created for this library
An ML platform team uses tf.metrics to compute streaming evaluation metrics during distributed training without manual aggregation.
A research team uses tf.metrics to maintain consistent metric implementations across training and evaluation code paths.
An ML platform team adopts tf.metrics in its baseline templates so metric definitions are reproducible across projects.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License