From e818dba016c390b7a85afb2cb941af8f2928fe3f Mon Sep 17 00:00:00 2001
From: zhumingjue138 <zhumingjue@huawei.com>
Date: Thu, 14 May 2026 10:47:14 +0800
Subject: [PATCH] [Test] Add stability tests for HunyuanImage-3-Instruct
 (#3504)

Signed-off-by: zhumingjue <zhumingjue@huawei.com>
---
 .../scripts/test_stability_hunyuan_image.py   | 69 +++++++++++++++++++
 .../stability/tests/test_hunyuan_image.json   | 36 ++++++++++
 2 files changed, 105 insertions(+)
 create mode 100644 tests/dfx/stability/scripts/test_stability_hunyuan_image.py
 create mode 100644 tests/dfx/stability/tests/test_hunyuan_image.json

diff --git a/tests/dfx/stability/scripts/test_stability_hunyuan_image.py b/tests/dfx/stability/scripts/test_stability_hunyuan_image.py
new file mode 100644
index 000000000..38858047f
--- /dev/null
+++ b/tests/dfx/stability/scripts/test_stability_hunyuan_image.py
@@ -0,0 +1,69 @@
+"""
+Hunyuan-Image stability: OmniServer (diffusion) + ``diffusion_benchmark_serving.py``.
+
+Configuration: ``tests/dfx/stability/tests/test_hunyuan_image.json``.
+"""
+
+from __future__ import annotations
+
+from pathlib import Path
+
+import pytest
+
+from tests.dfx.conftest import (
+    create_benchmark_indices,
+    create_test_parameter_mapping,
+    create_unique_server_params,
+    load_configs,
+)
+from tests.dfx.stability.helpers import _run_one_diffusion_batch, run_stability_benchmark_loop
+
+STABILITY_DIR = Path(__file__).resolve().parent.parent
+DEPLOY_CONFIGS_DIR = STABILITY_DIR / "deploy"
+CONFIG_FILE_PATH = str(STABILITY_DIR / "tests" / "test_hunyuan_image.json")
+DEFAULT_NUM_PROMPTS_PER_BATCH = 20
+STABILITY_SERVER_TIMEOUT_ARGS = ["--stage-init-timeout", "600", "--init-timeout", "900"]
+
+try:
+    BENCHMARK_CONFIGS = load_configs(CONFIG_FILE_PATH)
+except FileNotFoundError:
+    BENCHMARK_CONFIGS = []
+
+test_params = create_unique_server_params(BENCHMARK_CONFIGS, DEPLOY_CONFIGS_DIR) if BENCHMARK_CONFIGS else []
+server_to_benchmark_mapping = create_test_parameter_mapping(BENCHMARK_CONFIGS) if BENCHMARK_CONFIGS else {}
+benchmark_indices = create_benchmark_indices(BENCHMARK_CONFIGS, server_to_benchmark_mapping)
+
+
+@pytest.mark.slow
+@pytest.mark.diffusion
+@pytest.mark.parametrize("omni_server", test_params, indirect=True)
+@pytest.mark.parametrize("stability_benchmark_params", benchmark_indices, indirect=True)
+def test_stability_hunyuan_image(omni_server, stability_benchmark_params):
+    test_name = stability_benchmark_params["test_name"]
+    params = stability_benchmark_params["params"]
+    duration_sec = params.get("duration_sec", 300)
+    num_prompts_per_batch = params.get("num_prompts_per_batch", DEFAULT_NUM_PROMPTS_PER_BATCH)
+    request_rate = params.get("request_rate")
+    max_concurrency = params.get("max_concurrency")
+
+    bench_params = {
+        k: v
+        for k, v in params.items()
+        if k not in ("duration_sec", "request_rate", "max_concurrency", "num_prompts_per_batch")
+    }
+
+    result = run_stability_benchmark_loop(
+        host=omni_server.host,
+        port=omni_server.port,
+        model=omni_server.model,
+        duration_sec=duration_sec,
+        params=bench_params,
+        request_rate=request_rate,
+        max_concurrency=max_concurrency,
+        result_dir=str(STABILITY_DIR),
+        num_prompts_per_batch=num_prompts_per_batch,
+        run_one_batch=_run_one_diffusion_batch,
+    )
+
+    assert result.get("failed", 0) == 0, f"[{test_name}] Failed requests detected: {result.get('errors', [])}"
+    assert result.get("completed", 0) > 0, f"[{test_name}] No requests completed"
diff --git a/tests/dfx/stability/tests/test_hunyuan_image.json b/tests/dfx/stability/tests/test_hunyuan_image.json
new file mode 100644
index 000000000..b78fa812c
--- /dev/null
+++ b/tests/dfx/stability/tests/test_hunyuan_image.json
@@ -0,0 +1,36 @@
+[
+    {
+        "test_name": "test_hunyuan_image_stability_tp2_fp8_sp2",
+        "server_params": {
+            "model": "tencent/HunyuanImage-3.0-Instruct",
+            "serve_args": {
+                "tensor-parallel-size": 2,
+                "usp": 2,
+                "quantization": "fp8",
+                "distributed-executor-backend": "mp",
+                "enforce-eager": true,
+                "enable-diffusion-pipeline-profiler": true
+            }
+        },
+        "benchmark_params": [
+            {
+                "dataset": "random",
+                "task": "t2i",
+                "backend": "vllm-omni",
+                "duration_sec": 86400,
+                "max_concurrency": 1,
+                "num_prompts_per_batch": 50,
+                "width": {
+                    "min": 512,
+                    "max": 1024
+                },
+                "height": {
+                    "min": 512,
+                    "max": 1024
+                },
+                "num_inference_steps": 50,
+                "enable_negative_prompt": true
+            }
+        ]
+    }
+]