submission 742282

# MOE-MXFP4: No online tuning (avoid timeout), use default heuristics
# AITER_ONLINE_TUNE adds ~5min overhead which causes leaderboard timeout
import torch
from task import input_t, output_t
from aiter import ActivationType, QuantType
from aiter.fused_moe import fused_moe

def custom_kernel(data: input_t) -> output_t:
    (hidden_states, gate_up_weight, down_weight,
     gate_up_weight_scale, down_weight_scale,
     gate_up_weight_shuffled, down_weight_shuffled,
     gate_up_weight_scale_shuffled, down_weight_weight_scale_shuffled,
     topk_weights, topk_ids, config) = data
    
    hidden_pad = config["d_hidden_pad"] - config["d_hidden"]
    intermediate_pad = config["d_expert_pad"] - config["d_expert"]
    
    return fused_moe(
        hidden_states, gate_up_weight_shuffled, down_weight_shuffled,
        topk_weights, topk_ids, expert_mask=None,
        activation=ActivationType.Silu, quant_type=QuantType.per_1x32,
        doweight_stage1=False,
        w1_scale=gate_up_weight_scale_shuffled,
        w2_scale=down_weight_weight_scale_shuffled,
        a1_scale=None, a2_scale=None,
        hidden_pad=hidden_pad, intermediate_pad=intermediate_pad
    )