[Test] Add GLM-4.5 nightly test (#4225)

### What this PR does / why we need it?
Add GLM-4.5 nightly test

- vLLM version: v0.11.2

Signed-off-by: hfadzxy <starmoon_zhang@163.com>

.github/workflows/vllm_ascend_test_nightly_a3.yaml

@@ -134,6 +134,9 @@ jobs:
           - name: deepseek3_2-exp-w8a8
             os: linux-aarch64-a3-16
             tests: tests/e2e/nightly/models/test_deepseek_v3_2_exp_w8a8.py
+          - name: glm-4-5
+            os: linux-aarch64-a3-16
+            tests: tests/e2e/nightly/models/test_glm4_5.py
     uses: ./.github/workflows/_e2e_nightly_single_node.yaml
     with:
       vllm: v0.11.2

tests/e2e/nightly/models/test_glm4_5.py

@@ -0,0 +1,111 @@
+# Copyright (c) 2025 Huawei Technologies Co., Ltd. All Rights Reserved.
+# Copyright 2023 The vLLM team.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# This file is a part of the vllm-ascend project.
+#
+from typing import Any
+
+import openai
+import pytest
+from vllm.utils import get_open_port
+
+from tests.e2e.conftest import RemoteOpenAIServer
+from tools.aisbench import run_aisbench_cases
+
+MODELS = [
+    "ZhipuAI/GLM-4.5",
+]
+
+TENSOR_PARALLELS = [8]
+DATA_PARALLELS = [2]
+
+prompts = [
+    "San Francisco is a",
+]
+
+api_keyword_args = {
+    "max_tokens": 10,
+}
+
+aisbench_cases = [{
+    "case_type": "accuracy",
+    "dataset_path": "vllm-ascend/gsm8k-lite",
+    "request_conf": "vllm_api_general_chat",
+    "dataset_conf": "gsm8k/gsm8k_gen_0_shot_cot_chat_prompt",
+    "max_out_len": 4096,
+    "batch_size": 8,
+    "baseline": 95,
+    "threshold": 5
+}, {
+    "case_type": "performance",
+    "dataset_path": "vllm-ascend/GSM8K-in3500-bs400",
+    "request_conf": "vllm_api_stream_chat",
+    "dataset_conf": "gsm8k/gsm8k_gen_0_shot_cot_str_perf",
+    "num_prompts": 16,
+    "max_out_len": 1500,
+    "batch_size": 8,
+    "request_rate": 0,
+    "baseline": 1,
+    "threshold": 0.97
+}]
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize("model", MODELS)
+@pytest.mark.parametrize("tp_size", TENSOR_PARALLELS)
+@pytest.mark.parametrize("dp_size", DATA_PARALLELS)
+async def test_models(
+    model: str,
+    tp_size: int,
+    dp_size: int,
+) -> None:
+    port = get_open_port()
+    env_dict = {"HCCL_BUFFSIZE": "1024"}
+    server_args = [
+        "--no-enable-prefix-caching",
+        "--enable-expert-parallel",
+        "--tensor-parallel-size",
+        str(tp_size),
+        "--data-parallel-size",
+        str(dp_size),
+        "--port",
+        str(port),
+        "--max-model-len",
+        "8192",
+        "--max-num-batched-tokens",
+        "8192",
+        "--block-size",
+        "16",
+        "--trust-remote-code",
+        "--gpu-memory-utilization",
+        "0.9",
+    ]
+    request_keyword_args: dict[str, Any] = {
+        **api_keyword_args,
+    }
+    with RemoteOpenAIServer(model,
+                            server_args,
+                            server_port=port,
+                            env_dict=env_dict,
+                            auto_port=False) as server:
+        client = server.get_async_client()
+        batch = await client.completions.create(
+            model=model,
+            prompt=prompts,
+            **request_keyword_args,
+        )
+        choices: list[openai.types.CompletionChoice] = batch.choices
+        assert choices[0].text, "empty response"
+        # aisbench test
+        run_aisbench_cases(model, port, aisbench_cases)

tests/e2e/nightly/multi_node/config/models/DeepSeek-V3_2-Exp-bf16.yaml

@@ -13,7 +13,7 @@ env_common:
 deployment:
   -
     server_cmd: >
-        vllm serve Yanguan/DeepSeek-V3.2-Exp-bf16 \
+        vllm serve "Yanguan/DeepSeek-V3.2-Exp-bf16"
         --host 0.0.0.0
         --port $SERVER_PORT
         --data-parallel-address $LOCAL_IP
@@ -33,7 +33,7 @@ deployment:
 
   -
     server_cmd: >
-        vllm serve Yanguan/DeepSeek-V3.2-Exp-bf16 \
+        vllm serve "Yanguan/DeepSeek-V3.2-Exp-bf16"
         --headless
         --data-parallel-size 2 
         --data-parallel-size-local 1 

tests/e2e/nightly/multi_node/scripts/run.sh

@@ -108,8 +108,8 @@ install_extra_components() {
     fi
     pip install custom_ops-1.0-cp311-cp311-linux_aarch64.whl
     
-    export ASCEND_CUSTOM_OPP_PATH=/usr/local/Ascend/ascend-toolkit/latest/opp/vendors/customize:${ASCEND_CUSTOM_OPP_PATH}
+    export ASCEND_CUSTOM_OPP_PATH="/usr/local/Ascend/ascend-toolkit/latest/opp/vendors/customize${ASCEND_CUSTOM_OPP_PATH:+:${ASCEND_CUSTOM_OPP_PATH}}"
-    export LD_LIBRARY_PATH=/usr/local/Ascend/ascend-toolkit/latest/opp/vendors/customize/op_api/lib/:${LD_LIBRARY_PATH}
+    export LD_LIBRARY_PATH="/usr/local/Ascend/ascend-toolkit/latest/opp/vendors/customize/op_api/lib/${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}"
     source /usr/local/Ascend/ascend-toolkit/set_env.sh
     
     rm -f CANN-custom_ops-sfa-linux.aarch64.run \