Files
amd-strix-halo-vllm-toolboxes/benchmarks/benchmark_results_rocm/btbtyler09_Qwen3-Coder-30B-A3B-Instruct-gptq-4bit_tp1_throughput.json
T

7 lines
190 B
JSON

{
"elapsed_time": 621.1952276929987,
"num_requests": 100,
"total_num_tokens": 75285,
"requests_per_second": 0.16097998751758127,
"tokens_per_second": 121.19378360261106
}