Skip to content

Updates: remove mixed resolution and save server params into output json file #36

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
fhfuih merged 2 commits into from
Mar 27, 2026
Merged

Updates: remove mixed resolution and save server params into output json file #36

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
ccdbaab
remove mixed resolution
wtomin Mar 27, 2026
6eb6826
save server params in json
wtomin Mar 27, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 5 additions & 0 deletions tests/dfx/perf/scripts/run_diffusion_benchmark.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -317,6 +317,7 @@ def _unique_server_params(configs: list[dict[str, Any]]) -> list[dict[str, Any]]
"model": cfg["server_params"]["model"],
"serve_args": _build_serve_args(cfg["server_params"].get("serve_args", {})),
"benchmark_backend": "vllm-omni",
"server_params": cfg["server_params"],
}
)
return result
Expand Down Expand Up @@ -363,6 +364,7 @@ def diffusion_server(request):
print(f"\nStarting {server_type} server for test: {test_name}")
with _make_server(server_cfg) as server:
server.test_name = test_name
server.server_params = server_cfg["server_params"]
print(f"{server_type} server started successfully")
yield server
print(f"{server_type} server stopping…")
Expand Down Expand Up @@ -400,6 +402,7 @@ def run_benchmark(
params: dict[str, Any],
test_name: str,
backend: str = "vllm-omni",
server_params: dict[str, Any] | None = None,
) -> dict[str, Any]:
"""Run diffusion_benchmark_serving.py as a subprocess and return parsed metrics.

Expand Down Expand Up @@ -496,6 +499,7 @@ def run_benchmark(
"test_name": test_name,
"backend": backend,
"timestamp": timestamp,
"server_params": server_params,
"benchmark_params": params,
"result": metrics,
"log_file": str(log_file),
Expand Down Expand Up @@ -561,6 +565,7 @@ def test_diffusion_performance_benchmark(diffusion_server, benchmark_params):
params=params,
test_name=test_name,
backend=backend,
server_params=diffusion_server.server_params,
)

print(f"\n{'=' * 60}")
Expand Down
63 changes: 0 additions & 63 deletions tests/dfx/perf/tests/test_qwen_image_vllm_omni.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -43,27 +43,6 @@
"peak_memory_mb_max": 74000,
"peak_memory_mb_mean": 74000
}
},
{
"name": "mixed_resolution",
"dataset": "random",
"task": "t2i",
"num-inference-steps": 20,
"num-prompts": 10,
"max-concurrency": 1,
"enable-negative-prompt": true,
"random-request-config": [
{"width": 512, "height": 512, "num_inference_steps": 20, "weight": 0.15},
{"width": 768, "height": 768, "num_inference_steps": 20, "weight": 0.25},
{"width": 1024, "height": 1024, "num_inference_steps": 25, "weight": 0.45},
{"width": 1536, "height": 1536, "num_inference_steps": 35, "weight": 0.15}
],
"baseline": {
"throughput_qps": 0.14,
"latency_p99": 25.0,
"peak_memory_mb_max": 74000,
"peak_memory_mb_mean": 74000
}
}
]
},
Expand Down Expand Up @@ -116,27 +95,6 @@
"peak_memory_mb_max": 61000,
"peak_memory_mb_mean": 61000
}
},
{
"name": "mixed_resolution",
"dataset": "random",
"task": "t2i",
"num-inference-steps": 20,
"num-prompts": 10,
"max-concurrency": 1,
"enable-negative-prompt": true,
"random-request-config": [
{"width": 512, "height": 512, "num_inference_steps": 20, "weight": 0.15},
{"width": 768, "height": 768, "num_inference_steps": 20, "weight": 0.25},
{"width": 1024, "height": 1024, "num_inference_steps": 25, "weight": 0.45},
{"width": 1536, "height": 1536, "num_inference_steps": 35, "weight": 0.15}
],
"baseline": {
"throughput_qps": 0.20,
"latency_p99": 8.5,
"peak_memory_mb_max": 61000,
"peak_memory_mb_mean": 61000
}
}
]
},
Expand Down Expand Up @@ -199,27 +157,6 @@
"peak_memory_mb_max": 74000,
"peak_memory_mb_mean": 74000
}
},
{
"name": "mixed_resolution",
"dataset": "random",
"task": "t2i",
"num-inference-steps": 20,
"num-prompts": 10,
"max-concurrency": 1,
"enable-negative-prompt": true,
"random-request-config": [
{"width": 512, "height": 512, "num_inference_steps": 20, "weight": 0.15},
{"width": 768, "height": 768, "num_inference_steps": 20, "weight": 0.25},
{"width": 1024, "height": 1024, "num_inference_steps": 25, "weight": 0.45},
{"width": 1536, "height": 1536, "num_inference_steps": 35, "weight": 0.15}
],
"baseline": {
"throughput_qps": 0.35,
"latency_p99": 5.33,
"peak_memory_mb_max": 74000,
"peak_memory_mb_mean": 74000
}
}
]
}
Expand Down