amd-strix-halo-vllm-toolboxes/benchmarks/benchmark_results/meta-llama_Meta-Llama-3.1-8B-Instruct_tp1_throughput.json

{
    "elapsed_time": 395.08209386101225,
    "num_requests": 200,
    "total_num_tokens": 145877,
    "requests_per_second": 0.5062239041143659,
    "tokens_per_second": 369.23212230245684
}