Enum KVCacheFormat 

pub enum KVCacheFormat {
    F32,
    Int8,
    Fp8E4M3,
    Fp8E5M2,
}

KV cache storage format

Variants

F32

Full precision (default)

Int8

INT8 with per-head symmetric quantization: value = scale * int8_value

Fp8E4M3

FP8 E4M3 format (4 exponent, 3 mantissa bits) - good for inference

Fp8E5M2

FP8 E5M2 format (5 exponent, 2 mantissa bits) - wider range, less precision

Trait Implementations

impl Clone for KVCacheFormat

fn clone(&self) -> KVCacheFormat

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for KVCacheFormat

impl Debug for KVCacheFormat

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

Formats the value using the given formatter.

impl PartialEq for KVCacheFormat

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