fixed_nvfp4_proj

metrale_model_layers::layers::ops::w4a4_proj

Function fixed_nvfp4_proj 

Source 
pub fn fixed_nvfp4_proj(
    gpu: &dyn GpuBackend,
    family: ProjFamily,
    weight: &QuantizedWeight,
    input: DevicePtr,
    output: DevicePtr,
    m: u32,
    n: u32,
    k: u32,
    stream: u64,
) -> Result<bool>
Expand description

2026-10-01: [m, k] x [n, k]^T through nvfp4_proj_mx when family takes a fixed nvfp4 activation format at m rows (--activation-quantization) and weight is present; inputs and outputs are contiguous [m, k] and [m, n] rows. Ok(false) launches nothing.