[Test] Remove VLLM_USE_V1 in example and tests (#1733)

V1 is enabled by default, no need to set it by hand now. This PR remove
the useless setting in example and tests

- vLLM version: v0.9.2
- vLLM main:
9ad0a4588b

Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>

tests/e2e/singlecard/test_aclgraph.py

@@ -20,14 +20,12 @@ Compare the outputs of vLLM with and without aclgraph.
 Run `pytest tests/compile/test_aclgraph.py`.
 """
 
-import os
-
 import pytest
 import torch
 from vllm import LLM, SamplingParams
 
-from tests.conftest import VllmRunner
-from tests.model_utils import check_outputs_equal
+from tests.e2e.conftest import VllmRunner
+from tests.e2e.model_utils import check_outputs_equal
 
 MODELS = [
     "Qwen/Qwen2.5-0.5B-Instruct",
@@ -36,37 +34,29 @@ MODELS = [
 ]
 
 
-@pytest.mark.skipif(os.getenv("VLLM_USE_V1") == "0",
-                    reason="aclgraph only support on v1")
 @pytest.mark.parametrize("model", MODELS)
 @pytest.mark.parametrize("max_tokens", [32])
 def test_models(
     model: str,
     max_tokens: int,
-    monkeypatch: pytest.MonkeyPatch,
 ) -> None:
-    with monkeypatch.context() as m:
-        prompts = [
-            "Hello, my name is", "The president of the United States is",
-            "The capital of France is", "The future of AI is"
-        ]
+    prompts = [
+        "Hello, my name is", "The president of the United States is",
+        "The capital of France is", "The future of AI is"
+    ]
 
-        # aclgraph only support on v1
-        m.setenv("VLLM_USE_V1", "1")
+    sampling_params = SamplingParams(max_tokens=max_tokens, temperature=0.0)
+    # TODO: change to use vllmrunner when the registry of custom op is solved
+    # while running pytest
+    vllm_model = LLM(model)
+    vllm_aclgraph_outputs = vllm_model.generate(prompts, sampling_params)
+    del vllm_model
+    torch.npu.empty_cache()
 
-        sampling_params = SamplingParams(max_tokens=max_tokens,
-                                         temperature=0.0)
-        # TODO: change to use vllmrunner when the registry of custom op is solved
-        # while running pytest
-        vllm_model = LLM(model)
-        vllm_aclgraph_outputs = vllm_model.generate(prompts, sampling_params)
-        del vllm_model
-        torch.npu.empty_cache()
-
-        vllm_model = LLM(model, enforce_eager=True)
-        vllm_eager_outputs = vllm_model.generate(prompts, sampling_params)
-        del vllm_model
-        torch.npu.empty_cache()
+    vllm_model = LLM(model, enforce_eager=True)
+    vllm_eager_outputs = vllm_model.generate(prompts, sampling_params)
+    del vllm_model
+    torch.npu.empty_cache()
 
     vllm_aclgraph_outputs_list = []
     for output in vllm_aclgraph_outputs:
@@ -86,12 +76,9 @@ def test_models(
     )
 
 
-@pytest.mark.skipif(os.getenv("VLLM_USE_V1") == "0",
-                    reason="aclgraph only support on v1")
 def test_deepseek_raises_error(monkeypatch: pytest.MonkeyPatch) -> None:
     with monkeypatch.context() as m:
         m.setenv("VLLM_USE_MODELSCOPE", "True")
-        m.setenv("VLLM_USE_V1", "1")
         with pytest.raises(NotImplementedError) as excinfo:
             VllmRunner("deepseek-ai/DeepSeek-V2-Lite-Chat",
                        max_model_len=1024,