load_moe_qwen35_fp8_experts

metrale_model_layers::weight_map::ssm_qwen35

Function load_moe_qwen35_fp8_experts 

Source 
pub fn load_moe_qwen35_fp8_experts(
    store: &WeightStore,
    layer_prefix: &str,
    num_experts: usize,
    gpu: &dyn GpuBackend,
    config: &ModelConfig,
) -> Result<Vec<Fp8ExpertWeight>>
Expand description

2026-09-25: Load the routed experts of {layer_prefix}.mlp as native FP8 block-scaled tables, one entry per expert id. Experts this rank does not hold are NULL entries tagged Fp8BlockScaled.