quantized_from_fp8

metrale_model_layers::weight_map::nvfp4_detect

Function quantized_from_fp8 

Source 
pub fn quantized_from_fp8(
    store: &WeightStore,
    prefix: &str,
    n: usize,
    k: usize,
    gpu: &dyn GpuBackend,
    absmax_k: KernelHandle,
    quantize_k: KernelHandle,
    stream: u64,
) -> Result<QuantizedWeight>
Expand description

2026-09-25: Dequantize a block-scaled FP8 weight [n, k] to BF16, quantize it to NVFP4, and free the BF16 intermediate.