Struct Glm5NextMhcSiteWeights 

pub struct Glm5NextMhcSiteWeights {
    pub hc_fn: DevicePtr,
    pub hc_fn_bf16: bool,
    pub hc_scale: DevicePtr,
    pub hc_base: DevicePtr,
    pub mix: DevicePtr,
}

Per-site mHC weights. One set for the attention site, one for the FFN site.

🪤 hc_fn is FP32 to the kernel and BF16 on disk — the loader upcasts. Reading it at the on-disk width is the #341/#347 dtype-mismatch class, and the shapes do not say so.

Fields

hc_fn: DevicePtr

[mix_hc, hc_mult * hidden], where mix_hc = (2 + hc_mult) * hc_mult. FP32 unless hc_fn_bf16, in which case BF16 — the width the checkpoint actually stores.

hc_fn_bf16: bool

Is hc_fn BF16? Production sets it; the microtests and the split gate build their own F32 weights and leave it false, so the oracle they compare against is unchanged.

hc_scale: DevicePtr

[3] FP32 — the three logit scales (pre, post, comb), in that order.

hc_base: DevicePtr

[mix_hc] FP32.

mix: DevicePtr

[MHC_MIX_MAX_TOKENS, mix_hc] FP32 scratch: hc_mix writes it, hc_finish reads it. Per-site, so the two sites of a layer cannot alias; both run on one stream in order.

Trait Implementations

impl Clone for Glm5NextMhcSiteWeights

fn clone(&self) -> Glm5NextMhcSiteWeights

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Glm5NextMhcSiteWeights

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

Formats the value using the given formatter.

impl Copy for Glm5NextMhcSiteWeights