add e2e test for ep and etp

Signed-off-by: MengqingCao <cmq0113@163.com>

Author: MengqingCao

tests/e2e/multicard/test_ep_etp.py

@@ -0,0 +1,38 @@
+import os
+
+import pytest
+
+from tests.conftest import VllmRunner
+from tests.model_utils import check_outputs_equal
+
+
+@pytest.mark.skipif(os.getenv("VLLM_USE_V1") == "0",
+                    reason="ep is not supported on v0")
+@pytest.mark.parametrize("model_name", ["deepseek-ai/DeepSeek-V2-Lite-Chat"])
+def test_e2e_ep_etp_correctness(model_name):
+    example_prompts = [
+        "Hello, my name is",
+        "The president of the United States is",
+        "The capital of France is",
+        "The future of AI is",
+    ]
+    max_tokens = 5
+
+    with VllmRunner(model_name,
+                    tensor_parallel_size=2,
+                    additional_config={
+                        "expert_tensor_parallel_size": 2,
+                    }) as vllm_model:
+        etp_output = vllm_model.generate_greedy(example_prompts, max_tokens)
+
+    with VllmRunner(model_name,
+                    tensor_parallel_size=2,
+                    enable_expert_parallel=True) as vllm_model:
+        ep_output = vllm_model.generate_greedy(example_prompts, max_tokens)
+
+    check_outputs_equal(
+        outputs_0_lst=ep_output,
+        outputs_1_lst=etp_output,
+        name_0="ep_output",
+        name_1="etp_output",
+    )