amd-strix-halo-vllm-toolboxes/benchmarks/benchmark_results/dazipe_Qwen3-Next-80B-A3B-Instruct-GPTQ-Int4A16_cluster_tp2_eth_throughput.json

{
    "elapsed_time": 805.9022228560061,
    "num_requests": 200,
    "total_num_tokens": 146805,
    "requests_per_second": 0.24816906360082697,
    "tokens_per_second": 182.16229690959702
}