Enum MoeVariant 

#[non_exhaustive]
pub enum MoeVariant {
    ScalarGguf,
    Wmma,
    WmmaGguf,
}

Selector for the MoE variant.

#[non_exhaustive] — additional MoE backend variants (FP8 expert weights, BF16+WMMA on Hopper, multi-block routing) may land in future phases. Match arms must include a _ => catch-all.

Variants (Non-exhaustive)

Non-exhaustive enums could have additional variants added in future. Therefore, when matching against variants of non-exhaustive enums, an extra wildcard arm must be added to account for any future variants.

ScalarGguf

Scalar dispatch over GGUF-packed expert weights, f32 activations.

Wmma

WMMA tensor cores over FP expert weights, f16/bf16 activations.

WmmaGguf

WMMA tensor cores + GGUF-packed expert weights, f16/bf16 activations. The combined hot path for quantized LLM inference.

Trait Implementations

impl Clone for MoeVariant

fn clone(&self) -> MoeVariant

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Copy for MoeVariant

impl Debug for MoeVariant

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Eq for MoeVariant

impl Hash for MoeVariant

fn hash<__H: Hasher>(&self, state: &mut __H)

Feeds this value into the given Hasher.

fn hash_slice<H>(data: &[Self], state: &mut H)

where H: Hasher, Self: Sized,

Feeds a slice of this type into the given Hasher.

impl PartialEq for MoeVariant

fn eq(&self, other: &MoeVariant) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl StructuralPartialEq for MoeVariant