amd-strix-halo-vllm-toolboxes/benchmarks/benchmark_results/dazipe_Qwen3-Next-80B-A3B-Instruct-GPTQ-Int4A16_tp1_throughput.json

{
    "elapsed_time": 1303.4944151099999,
    "num_requests": 200,
    "total_num_tokens": 146805,
    "requests_per_second": 0.15343372221746138,
    "tokens_per_second": 112.62418795067208
}