Graph execution

Graph execution                              Deep Learning; MLOps

noun phrase

Definition: A computation mode in which operations are represented as a computational graph and executed as a compiled or optimized graph rather than immediately step by step in eager mode. In TensorFlow, graph execution is typically associated with tf.function and graph-based optimization [TensorFlow Guide].

Example in context:

Synonyms: graph mode; computational graph execution

Related terms: eager execution; tf.function; computational graph

Добавить комментарий 0

Ваш электронный адрес не будет опубликован. Обязательные поля помечены *