Module generic 

Generic backend that pairs a Driver with a ModelArch.

GenericBackend implements EmbedBackend by delegating to the architecture’s forward() method, which composes driver primitives into the full inference pipeline. This decouples weight loading from the backend interface — any (Driver, ModelArch) pair can serve as an embedding backend.

The _mmap field keeps the memory-mapped safetensors file alive as long as the backend exists, since Metal zero-copy buffers and CPU MmapTensor slices reference its pages.

Structs

GenericBackend

Generic backend that pairs a Driver with a ModelArch.

Enums

MmapHolder

Holds a memory-mapped file, accepting either an owned Mmap or an Arc<Mmap>. CPU backends share the Arc with MmapTensor::Mapped variants; GPU backends pass an owned Mmap.