Home › Glossary › Systems › Eager execution

Beginner · Systems

Eager execution

Visual diagram · (in preparation) · Math · (in preparation) · Worked example · 3 difficulty levels.

TL;DR. A TensorFlow execution mode where operations are performed immediately as they are called.

Technical Definition

A TensorFlow execution mode where operations are performed immediately as they are called.

How it works

Eager execution is a programming style in TensorFlow where operations are evaluated immediately. This is similar to how standard Python code runs, making it more intuitive for debugging and development. In contrast, graph execution builds a computation graph first and then executes it, which can offer performance benefits but is harder to inspect.

Related Concepts

  • Graph execution — A mode of operation in TensorFlow where a computation graph is first defined and then executed, enabling optimizations and portability.

Further Reading

  • Google ML Glossary