Performance/optimization related work across the compiler and runtime
Shared code generation infrastructure and dialects
Relating to the IREE compiler dialects (flow, hal, vm)
Relating to build systems, CI, or testing
LLVM code generation compiler backend
SPIR-V code generation compiler backend
Relating to the IREE runtime library
Run default x86_64 benchmarks
StableHLO (JAX/TensorFlow/etc) import and conversion
Run default Android CPU benchmarks
Nice things are nice; let's have some
Run default compilation statistics benchmarks
Extra attention is needed
Runtime Host/CPU-based HAL backend
NVVM code generation compiler backend
Tensor Operator Set Architecture (TOSA) import, tests, etc.
Tracking issues for related deliverables (as in Agile dev)
IREE's compiler tooling (iree-compile, iree-opt, etc)
Python wrapping IREE's C API
Request support or ask a question
Run default CUDA benchmarks
Relating to benchmarking infrastructure
✏️ Improvements or additions to documentation
Run default Android GPU benchmarks
Runtime Vulkan GPU HAL backend