iree-org/iree
An MLIR-based compiler and runtime toolkit for deploying machine learning models to diverse hardware backends.
Velocity · 7d
+1.5
★ / day
Trend
→steady
star history
IREE compiles ML models expressed in frameworks like JAX, PyTorch, ONNX, and TensorFlow to a unified intermediate representation, then lowers and executes them on various hardware targets including CUDA, ROCm, Vulkan/SPIR-V, and CPU backends. It is designed for scalable deployment from datacenter to mobile and edge devices.