Glossary term
Glossary term
Infrastructure and Serving
In TensorFlow, a computation specification. Nodes in the graph represent operations. Edges are directed and represent passing the result of an operation (a Tensor) as an operand to another operation. Use TensorBoard to visualize a graph.
Created for this library
An ML platform team uses a static computation graph for production serving to gain throughput from compiler optimizations.
A search team represents its retrieval scoring as a graph of operations that compiles to optimized device code at deploy time.
An ML team uses graph-mode execution for its production inference path while keeping eager execution for research notebooks.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License