Dateien
amd-strix-halo-vllm-toolboxes/benchmarks/benchmark_results/btbtyler09_Qwen3-Coder-30B-A3B-Instruct-gptq-8bit_tp1_throughput.json
T

7 Zeilen
187 B
JSON

{
"elapsed_time": 865.5675225800005,
"num_requests": 100,
"total_num_tokens": 75285,
"requests_per_second": 0.115531136960788,
"tokens_per_second": 86.97761646092924
}