load_qkvo_tp

Function load_qkvo_tp 

Source
pub fn load_qkvo_tp<F, T>(
    config: &ModelConfig,
    proj_loader: F,
) -> Result<[T; 4]>
where F: FnMut(&str, usize, usize, TpShardKind) -> Result<T>,
Expand description

Sequence the four Q/K/V/O loads via a loader-supplied closure. The closure receives (name, full_out, full_in, kind) and returns the loader’s representation of that projection (BF16 dense, NVFP4 quantized, FP8 block-scaled — varies by format).

Returns [Q, K, V, O]; callers destructure with let [q, k, v, o] = load_qkvo_tp(config, |name, n, k, kind| { ... })?;.