Revert "drop ascend scheduler" (#4580)

Reverts vllm-project/vllm-ascend#4498
- vLLM version: v0.11.2
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.2

tests/e2e/multicard/test_expert_parallel.py

@@ -15,14 +15,23 @@ def test_e2e_ep_correctness(model_name):
     max_tokens = 5
 
     # FIXME: Really strange that chunked prefill might lead to different results, investigate further
-    with VllmRunner(model_name, tensor_parallel_size=2,
-                    enforce_eager=False) as vllm_model:
+    with VllmRunner(
+            model_name,
+            tensor_parallel_size=2,
+            additional_config={"ascend_scheduler_config": {
+                "enabled": True
+            }},
+            enforce_eager=False) as vllm_model:
         tp_output = vllm_model.generate_greedy(example_prompts, max_tokens)
 
-    with VllmRunner(model_name,
-                    tensor_parallel_size=2,
-                    enable_expert_parallel=True,
-                    enforce_eager=False) as vllm_model:
+    with VllmRunner(
+            model_name,
+            tensor_parallel_size=2,
+            enable_expert_parallel=True,
+            additional_config={"ascend_scheduler_config": {
+                "enabled": True
+            }},
+            enforce_eager=False) as vllm_model:
         ep_output = vllm_model.generate_greedy(example_prompts, max_tokens)
 
     check_outputs_equal(