amd-strix-halo-vllm-toolboxes/benchmarks/benchmark_results_rocm/btbtyler09_Qwen3-Coder-30B-A3B-Instruct-gptq-4bit_cluster_tp2_throughput.json

{
    "elapsed_time": 438.29837328500435,
    "num_requests": 100,
    "total_num_tokens": 75285,
    "requests_per_second": 0.2281550790401287,
    "tokens_per_second": 171.7665512553609
}