47 lines
1.1 KiB
JSON
47 lines
1.1 KiB
JSON
{
|
|
"backend": "vllm-omni",
|
|
"task": "fl2va",
|
|
"gpus_per_instance": 8,
|
|
"dit_tensor_parallel_size": 2,
|
|
"ulysses_degree": 4,
|
|
"replicas": 1,
|
|
"expected_requests": 32,
|
|
"requests_recorded": 32,
|
|
"completed": 32,
|
|
"failed": 0,
|
|
"machine_wall_s": 3330.181683778763,
|
|
"machine_qps": 0.009609085340860305,
|
|
"latency_mean_s": 104.06762664610869,
|
|
"latency_p50_s": 75.86134355794638,
|
|
"latency_p95_s": 240.5564539189101,
|
|
"by_short_edge": {
|
|
"480": {
|
|
"requests": 8,
|
|
"completed": 8,
|
|
"failed": 0,
|
|
"latency_mean_s": 23.865465252136346,
|
|
"latency_p95_s": 23.923671398696023
|
|
},
|
|
"720": {
|
|
"requests": 8,
|
|
"completed": 8,
|
|
"failed": 0,
|
|
"latency_mean_s": 70.80472628047573,
|
|
"latency_p95_s": 71.00687271173811
|
|
},
|
|
"768": {
|
|
"requests": 8,
|
|
"completed": 8,
|
|
"failed": 0,
|
|
"latency_mean_s": 80.94421810278436,
|
|
"latency_p95_s": 81.08086508190027
|
|
},
|
|
"1080": {
|
|
"requests": 8,
|
|
"completed": 8,
|
|
"failed": 0,
|
|
"latency_mean_s": 240.65609694903833,
|
|
"latency_p95_s": 241.86744526363907
|
|
}
|
|
}
|
|
} |