[Test][HunyuanImage3] Add e2e offline I2T smoke test (#3332)

Signed-off-by: TaffyOfficial <2324465096@qq.com>
Co-authored-by: TaffyOfficial <2324465096@qq.com>

tests/e2e/offline_inference/test_hunyuanimage3_i2t_expansion.py

@@ -0,0 +1,95 @@
+# SPDX-License-Identifier: Apache-2.0
+# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
+"""Smoke test for HunyuanImage-3.0 Image-to-Text (I2T) pipeline."""
+
+from collections.abc import Generator
+from pathlib import Path
+
+import pytest
+import torch
+from PIL import Image
+
+from tests.helpers.runtime import OmniRunner
+from vllm_omni import Omni
+from vllm_omni.diffusion.models.hunyuan_image3.prompt_utils import build_prompt
+
+MODEL_NAME = "tencent/HunyuanImage-3.0-Instruct"
+REPO_ROOT = Path(__file__).resolve().parents[3]
+STAGE_CONFIG_PATH = REPO_ROOT / "vllm_omni" / "model_executor" / "stage_configs" / "hunyuan_image3_i2t.yaml"
+
+# First 20 generated token IDs from the HF greedy reference on this input.
+# vllm-omni AR output matches this prefix bitwise; the two implementations
+# diverge past this point.
+EXPECTED_PREFIX_TOKEN_IDS: list[int] = [
+    791,
+    2217,
+    374,
+    264,
+    6573,
+    11,
+    14113,
+    6307,
+    1933,
+    449,
+    912,
+    27339,
+    11,
+    6302,
+    11,
+    477,
+    3649,
+    3118,
+    13,
+    1102,
+]
+# Decoded form, kept only for human-readable assertion messages.
+EXPECTED_PREFIX_TEXT = "The image is a solid, uniform green color with no variations, objects, or details present. It"
+
+pytestmark = [pytest.mark.full_model, pytest.mark.diffusion]
+
+
+@pytest.fixture(scope="module")
+def omni() -> Generator[Omni, None, None]:
+    with OmniRunner(
+        MODEL_NAME,
+        stage_configs_path=str(STAGE_CONFIG_PATH),
+    ) as runner:
+        yield runner.omni
+
+
+@pytest.mark.skipif(torch.accelerator.device_count() < 4, reason="Need at least 4 CUDA GPUs.")
+def test_i2t_generates_text(omni: Omni) -> None:
+    """Verify I2T output's first 20 token IDs match the HF greedy baseline."""
+    # Solid-color image keeps the input self-contained and reproducible.
+    input_image = Image.new("RGB", (256, 256), color=(128, 200, 100))
+
+    prompt = build_prompt("Describe the content of the picture.", task="i2t")
+    prompt_dict = {
+        "prompt": prompt,
+        "modalities": ["text"],
+        "multi_modal_data": {"image": input_image},
+    }
+
+    outputs = omni.generate(prompts=[prompt_dict])
+    assert outputs, "No outputs returned from Omni.generate()"
+
+    request_output = outputs[0].request_output
+    assert request_output.outputs, "No completion outputs"
+
+    completion = request_output.outputs[0]
+    finish_reason = getattr(completion, "finish_reason", None)
+    assert finish_reason is not None, "AR generation did not finish (finish_reason is None)"
+    assert str(finish_reason) != "abort", f"AR generation aborted: finish_reason={finish_reason!r}"
+
+    token_ids = list(getattr(completion, "token_ids", []) or [])
+    n = len(EXPECTED_PREFIX_TOKEN_IDS)
+    assert len(token_ids) >= n, (
+        f"AR output shorter than {n} tokens (got {len(token_ids)}): token_ids={token_ids!r} text={completion.text!r}"
+    )
+    assert token_ids[:n] == EXPECTED_PREFIX_TOKEN_IDS, (
+        f"AR prefix drift vs HF reference\n"
+        f"  expected ids : {EXPECTED_PREFIX_TOKEN_IDS!r}\n"
+        f"  actual ids   : {token_ids[:n]!r}\n"
+        f"  expected text: {EXPECTED_PREFIX_TEXT!r}\n"
+        f"  actual text  : {completion.text!r}"
+    )