litert_bindings library

Classes

AlignedAlloc: Pair of (raw, aligned) pointers returned by allocAligned.
LiteRtBindings: Minimal C-API surface for running an embedding forward pass.
LiteRtLayoutMsvc: LiteRtLayout as packed by MSVC. Used on Windows only.
LiteRtLayoutPosix: LiteRtLayout as packed by GCC/Clang. Used on every platform except Windows. Read LiteRtLayoutView for layout-agnostic accessors.
LiteRtLayoutView: Layout-agnostic view over a LiteRtLayout* returned from the C API. Pick the right struct (POSIX vs MSVC) based on the host compiler ABI at allocation time so callers don't need to think about it.
LiteRtRankedTensorTypeMsvc
LiteRtRankedTensorTypePosix
LiteRtRankedTensorTypeView: Layout-agnostic builder for LiteRtRankedTensorType*. Use this to fill in element type, rank, and dimensions without caring which compiler produced the LiteRT shared library.

LiteRtStatusX on int: Throws if a LiteRT C call returned a non-OK status code.

kLiteRtElementTypeFloat32 → const int: LiteRtElementType values used for embedding pipelines.
kLiteRtElementTypeInt32 → const int
kLiteRtHostMemoryAlignment → const int: Host memory alignment required by LiteRtCreateTensorBufferFromHostMemory (see LITERT_HOST_MEMORY_BUFFER_ALIGNMENT in litert_tensor_buffer_types.h).
kLiteRtHwAcceleratorCpu → const int
kLiteRtHwAcceleratorGpu → const int
kLiteRtHwAcceleratorNone → const int: LiteRtHwAcceleratorSet bit flags (litert_common.h). CPU is enough for embedding; GPU/NPU can be wired later.
kLiteRtHwAcceleratorNpu → const int
kLiteRtMaxRank → const int
kLiteRtStatusOk → const int
kLiteRtTensorBufferLockModeRead → const int: LiteRtTensorBufferLockMode values (litert_common.h).
kLiteRtTensorBufferLockModeReadWrite → const int
kLiteRtTensorBufferLockModeWrite → const int

allocAligned(int bytes, {int align = kLiteRtHostMemoryAlignment}) → AlignedAlloc: Allocate bytes of host memory and return a AlignedAlloc whose .aligned pointer is aligned to align bytes (default 64, the LITERT_HOST_MEMORY_BUFFER_ALIGNMENT value).