[CI] drop ascend scheduler test (#4582)

let' drop ascend scheduler test first to ensure all function works
without it.


- vLLM version: v0.11.2
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.2

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

tests/e2e/nightly/models/test_deepseek_r1_0528_w8a8.py

@@ -33,7 +33,6 @@ MODES = [
     "single",
     "aclgraph",
     "aclgraph_mlapo",
-    "no_chunkprefill",
 ]
 
 prompts = [
@@ -82,9 +81,6 @@ async def test_models(model: str, mode: str) -> None:
         "method": "deepseek_mtp"
     }
     additional_config = {
-        "ascend_scheduler_config": {
-            "enabled": False
-        },
         "torchair_graph_config": {
             "enabled": True,
             "enable_multistream_moe": False,
@@ -112,10 +108,6 @@ async def test_models(model: str, mode: str) -> None:
     if mode == "aclgraph_mlapo":
         env_dict["VLLM_ASCEND_ENABLE_MLAPO"] = "1"
         additional_config["torchair_graph_config"] = {"enabled": False}
-    if mode == "no_chunkprefill":
-        additional_config["ascend_scheduler_config"] = {"enabled": True}
-        i = server_args.index("--max-num-batched-tokens") + 1
-        server_args[i] = "36864"
     server_args.extend(["--additional-config", json.dumps(additional_config)])
     request_keyword_args: dict[str, Any] = {
         **api_keyword_args,
@@ -134,7 +126,7 @@ async def test_models(model: str, mode: str) -> None:
         choices: list[openai.types.CompletionChoice] = batch.choices
         assert choices[0].text, "empty response"
         print(choices)
-        if mode in ["single", "no_chunkprefill"]:
+        if mode in ["single"]:
             return
         # aisbench test
         run_aisbench_cases(model,