Struct Q8Activations 

pub struct Q8Activations {
    pub q: Vec<i8>,
    pub d: Vec<f32>,
}

An activation vector quantized to signed 8-bit in 32-element blocks, each with its own f32 scale (d), so it can feed the integer vec_dot paths against Q8_0 weights. This mirrors llama.cpp’s quantize_row_q8_1 (minus the block sum, which is only needed for asymmetric weight formats): quantizing the shared activation once per matvec turns every weight-row dot into an int8×int8 → int32 reduction (vdotq_s32 / _mm256_maddubs-class ops) plus a single scale, which is what lets llama.cpp’s CPU matmul stay in integer SIMD.

Fields

q: Vec<i8>

Signed 8-bit quantized values, n_blocks * 32 long.

d: Vec<f32>

Per-block scale, n_blocks long. x ≈ q * d.

Implementations

impl Q8Activations

pub fn n_blocks(&self) -> usize

Trait Implementations

impl Clone for Q8Activations

fn clone(&self) -> Q8Activations

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Q8Activations

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

Formats the value using the given formatter.