Function load_mla_attn 

pub fn load_mla_attn(
    shard: &ShardedSafetensors,
    prefix: &str,
    num_heads: usize,
    q_lora_rank: usize,
    kv_lora_rank: usize,
    qk_nope_head_dim: usize,
    qk_rope_head_dim: usize,
    v_head_dim: usize,
    hidden_dim: usize,
) -> Result<MlaAttnWeights, KimiLoadError>

Loads one Gated-MLA-attention layer’s weights (real tensor names under {prefix}.self_attn.*).