Glossary term
Glossary term
Infrastructure and Serving
An open-source machine learning compiler for GPUs, CPUs, and ML accelerators.
The XLA compiler takes models from popular ML frameworks such as PyTorch, TensorFlow, and JAX, and optimizes them for high-performance execution across different hardware platforms including GPUs, CPUs, and ML accelerators.
Created for this library
An ML platform team uses XLA to compile and fuse operations in its model graph for faster training and serving.
A research engineer relies on XLA-compiled graphs to maximize accelerator utilization in JAX-based training.
An ML platform team uses XLA optimizations in its production serving stack to lower inference cost across many models.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License