Struct PredictedDerived 

pub struct PredictedDerived {
    pub attn_fp8_twins: u64,
    pub ssm_fp8_concat: u64,
    pub ffn_gateup_fused: u64,
    pub ffn_gateup_pruned: u64,
    pub twins: TwinsBuilt,
    pub attn_twin_set: Fp8TwinSet,
}

2026-09-25: The derived bytes the native-FP8 dense loader will keep, by term.

Fields

attn_fp8_twins: u64

2026-09-25: Fp8Weight::transpose_for_gemm twins for the projections fp8_residency::DenseFp8Plan selects, summed over full-attention layers.

ssm_fp8_concat: u64

2026-09-25: The fused [QKV|Z] FP8 weight, its block-scale grids, the out_proj block-scale grid and the interleaved in_proj_ba, summed over linear-attention layers (ssm_concat_bytes).

ffn_gateup_fused: u64

2026-09-25: The E4M3 bytes of the fused [2*inter, hidden] gate+up weight, summed over layers; its scale grid is not counted. 0 when the fusion is not selected.

ffn_gateup_pruned: u64

2026-09-25: The mlp.gate_proj.weight and mlp.up_proj.weight bytes that prune_after_load releases on the same layers. Equal to Self::ffn_gateup_fused: the fused weight is those two tensors side by side.

twins: TwinsBuilt

2026-09-25: Which derived families the prediction expects.

attn_twin_set: Fp8TwinSet

2026-09-25: The twin set the attention term was priced at.

Implementations

impl PredictedDerived

pub fn total(&self) -> u64

2026-09-25: Derived bytes above the on-disk checkpoint size, which headroom.rs adds to that size. The fusion enters as a difference, which is zero: the on-disk size still counts the gate and up tensors that prune_after_load releases once the fused copy exists (the_gateup_fusion_is_residency_neutral).

Trait Implementations

impl Clone for PredictedDerived

fn clone(&self) -> PredictedDerived

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for PredictedDerived

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

Formats the value using the given formatter.

impl Default for PredictedDerived

fn default() -> PredictedDerived

Returns the “default value” for a type.

impl PartialEq for PredictedDerived

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

impl Eq for PredictedDerived

impl StructuralPartialEq for PredictedDerived