dequant_fp8_to_bf16_into

metrale_model_layers::weight_map::fp8_dequant

Function dequant_fp8_to_bf16_into 

Source 
pub fn dequant_fp8_to_bf16_into(
    store: &WeightStore,
    prefix: &str,
    gpu: &dyn GpuBackend,
    dest: DevicePtr,
) -> Result<DenseWeight>
Expand description

2026-09-25: Dequantize FP8 E4M3 + per-tensor scale → BF16 into dest, a caller-provided buffer of at least 2 × elements bytes, and return it as a DenseWeight. The GPU path does not synchronise, so the caller must order any reuse of dest after the launch (e.g. after quantize_to_nvfp4, which synchronises its stream).