Struct InferencePathPlan 

pub struct InferencePathPlan {
    pub wgpu_backend: Option<String>,
    pub compute_lane: ComputeLane,
    pub quant: QuantProfile,
    pub prefill_prefer_tc: bool,
    pub decode_is_gemv: bool,
    pub rationale: String,
    pub winning_circuit: Option<String>,
    pub decode_proxy_tok_s: Option<f64>,
    pub gemv_ms: Option<f64>,
}

Full selected plan for this process / host.

Fields

wgpu_backend: Option<String>

dx12 | vulkan | metal | gl | unknown

compute_lane: ComputeLane

quant: QuantProfile

prefill_prefer_tc: bool

Use tensor-core dense GEMM for prefill when dims allow (CUDA lane).

decode_is_gemv: bool

Decode uses GEMV (m=1); never pretend TC replaces decode.

rationale: String

Human-readable reason chain.

winning_circuit: Option<String>

Passport circuit label that won, if any.

decode_proxy_tok_s: Option<f64>

gemv_ms: Option<f64>

Trait Implementations

impl Clone for InferencePathPlan

fn clone(&self) -> InferencePathPlan

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for InferencePathPlan

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

Formats the value using the given formatter.

impl PartialEq for InferencePathPlan

fn eq(&self, other: &InferencePathPlan) -> 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 InferencePathPlan