Enum WeightQuantFormat 

pub enum WeightQuantFormat {
    Bf16,
    Fp8PerRow,
    Fp8BlockScaled,
    Fp8SingleScale,
    Nvfp4,
    Mxfp4E8m0,
    PackedQ2_0,
}

2026-09-25: The format of a weight buffer in device memory, as opposed to the on-disk format (Nvfp4Variant). A kernel call site checks it with WeightQuantFormat::expect, because a kernel cannot tell one layout from another by its pointer.

Variants

Bf16

2026-09-25: BF16, not quantized.

Fp8PerRow

2026-09-25: FP8 E4M3 with one F32 scale per row ([N]), as load_fp8_weight returns.

Fp8BlockScaled

2026-09-25: FP8 E4M3 with one F32 scale per 128x128 block, as load_fp8_block_scaled_as_fp8weight and quantize_to_fp8_blockscaled return.

Fp8SingleScale

2026-09-25: FP8 E4M3 with no scale buffer.

Nvfp4

2026-09-25: NVFP4: packed E2M1 nibbles, one FP8 scale per 16 weights, and an F32 per-tensor scale.

Mxfp4E8m0

2026-09-25: Native MXFP4: packed E2M1 nibbles and one E8M0 scale per 32 weights, with no per-tensor scale. The DeepSeek-V4 loader tags its routed experts with it (MoeLayer::experts_scale_kind).

PackedQ2_0

2026-09-25: Keep-packed ternary Q2_0 (ggml id 42): raw block_q2_0 blocks, each an fp16 scale and 2-bit codes, as PackedQ2Weight holds them.

Implementations

impl WeightQuantFormat

pub fn expect(self, expected: WeightQuantFormat, context: &str)

2026-09-25: Panic, naming context, when self is not expected.

Trait Implementations

impl Clone for WeightQuantFormat

fn clone(&self) -> WeightQuantFormat

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for WeightQuantFormat

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

Formats the value using the given formatter.

impl PartialEq for WeightQuantFormat

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

impl Eq for WeightQuantFormat

impl StructuralPartialEq for WeightQuantFormat