47 lines
1.1 KiB
JSON
47 lines
1.1 KiB
JSON
{
|
|
"backend": "vllm-omni",
|
|
"task": "fl2va",
|
|
"gpus_per_instance": 2,
|
|
"dit_tensor_parallel_size": 2,
|
|
"ulysses_degree": 1,
|
|
"replicas": 4,
|
|
"expected_requests": 32,
|
|
"requests_recorded": 32,
|
|
"completed": 32,
|
|
"failed": 0,
|
|
"machine_wall_s": 2980.0244238376617,
|
|
"machine_qps": 0.01073816702441336,
|
|
"latency_mean_s": 372.29218261338247,
|
|
"latency_p50_s": 265.3287420308916,
|
|
"latency_p95_s": 879.4243095606682,
|
|
"by_short_edge": {
|
|
"480": {
|
|
"requests": 8,
|
|
"completed": 8,
|
|
"failed": 0,
|
|
"latency_mean_s": 80.09600442298688,
|
|
"latency_p95_s": 80.27359099447494
|
|
},
|
|
"720": {
|
|
"requests": 8,
|
|
"completed": 8,
|
|
"failed": 0,
|
|
"latency_mean_s": 247.4871746932331,
|
|
"latency_p95_s": 247.78976834390778
|
|
},
|
|
"768": {
|
|
"requests": 8,
|
|
"completed": 8,
|
|
"failed": 0,
|
|
"latency_mean_s": 283.1468455953582,
|
|
"latency_p95_s": 283.65168853149515
|
|
},
|
|
"1080": {
|
|
"requests": 8,
|
|
"completed": 8,
|
|
"failed": 0,
|
|
"latency_mean_s": 878.4387057419517,
|
|
"latency_p95_s": 879.5743667519534
|
|
}
|
|
}
|
|
} |