Enum InferenceMode 

#[repr(u8)]
pub enum InferenceMode {
    Portable = 0,
    CudaTc = 1,
    QuantGraph = 2,
    FastVerify = 3,
}

Active inference approach for this process.

Variants

Portable = 0

wgpu single-fence resident decode/prefill — default, portable.

CudaTc = 1

Prefer CUDA tensor-core dense GEMM via forge when eligible; else Portable.

QuantGraph = 2

Quantized front-end + graph/Webizen grounding for quality recovery.

FastVerify = 3

Full-speed decode (skip mid-token Sentinel rings) then post-turn verify/heal.

Implementations

impl InferenceMode

pub const ALL: [InferenceMode; 4]

pub fn as_str(self) -> &'static str

pub fn parse(s: &str) -> Option<Self>

pub fn description(self) -> &'static str

Trait Implementations

impl Clone for InferenceMode

fn clone(&self) -> InferenceMode

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for InferenceMode

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

Formats the value using the given formatter.

impl Hash for InferenceMode

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 InferenceMode

fn eq(&self, other: &InferenceMode) -> 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 Copy for InferenceMode

impl Eq for InferenceMode

impl StructuralPartialEq for InferenceMode