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_fused_moe_allgather_ep.py

@@ -49,7 +49,13 @@ def test_generate_with_allgather():
                     tensor_parallel_size=2,
                     max_model_len=1024,
                     dtype="auto",
-                    enable_expert_parallel=True) as vllm_model:
+                    enable_expert_parallel=True,
+                    additional_config={
+                        "ascend_scheduler_config": {
+                            "enabled": True,
+                            "chunked_prefill_enabled": False,
+                        },
+                    }) as vllm_model:
         vllm_model.generate(example_prompts, sampling_params)
 
 
@@ -70,5 +76,11 @@ def test_generate_with_alltoall():
                     tensor_parallel_size=2,
                     max_model_len=1024,
                     dtype="auto",
-                    enable_expert_parallel=True) as vllm_model:
+                    enable_expert_parallel=True,
+                    additional_config={
+                        "ascend_scheduler_config": {
+                            "enabled": True,
+                            "chunked_prefill_enabled": False,
+                        },
+                    }) as vllm_model:
         vllm_model.generate(example_prompts, sampling_params)