[CI] migrate ascend-gptq from test/srt to test/registered (#19628)

test/registered/ascend/llm_models/test_ascend_qwen3_1_7b_gptq_int8.py

@@ -0,0 +1,32 @@
+import unittest
+
+from sglang.test.ascend.gsm8k_ascend_mixin import GSM8KAscendMixin
+from sglang.test.ascend.test_ascend_utils import QWEN3_1_7B_GPTQ_INT8_WEIGHTS_PATH
+from sglang.test.ci.ci_register import register_npu_ci
+from sglang.test.test_utils import CustomTestCase
+
+register_npu_ci(est_time=400, suite="per-commit-1-npu-a2")
+
+
+class TestQwen317BGPTQInt8(GSM8KAscendMixin, CustomTestCase):
+    """Testcase: Verify that the inference accuracy of the Qwen/Qwen3-1.7B-GPTQ-Int8 model on the GSM8K dataset is no less than 0.65.
+
+    [Test Category] Model
+    [Test Target] Qwen/Qwen3-1.7B-GPTQ-Int8
+    """
+
+    model = QWEN3_1_7B_GPTQ_INT8_WEIGHTS_PATH
+    accuracy = 0.65
+    other_args = [
+        "--trust-remote-code",
+        "--mem-fraction-static",
+        "0.8",
+        "--attention-backend",
+        "ascend",
+        "--quantization",
+        "gptq",
+    ]
+
+
+if __name__ == "__main__":
+    unittest.main()

test/registered/ascend/llm_models/test_ascend_qwen3_30b.py

@@ -2,7 +2,7 @@ import unittest
 
 from sglang.test.ascend.gsm8k_ascend_mixin import GSM8KAscendMixin
 from sglang.test.ascend.test_ascend_utils import (
-    Qwen3_30B_A3B_Instruct_2507_WEIGHTS_PATH,
+    QWEN3_30B_A3B_INSTRUCT_2507_WEIGHTS_PATH,
 )
 from sglang.test.ci.ci_register import register_npu_ci
 from sglang.test.test_utils import CustomTestCase
@@ -17,7 +17,7 @@ class TestQwen330B(GSM8KAscendMixin, CustomTestCase):
     [Test Target] Qwen/Qwen3-30B-A3B-Instruct-2507
     """
 
-    model = Qwen3_30B_A3B_Instruct_2507_WEIGHTS_PATH
+    model = QWEN3_30B_A3B_INSTRUCT_2507_WEIGHTS_PATH
     accuracy = 0.90
     other_args = [
         "--trust-remote-code",

test/srt/ascend/test_ascend_gptq.py

@@ -1,73 +0,0 @@
-import unittest
-from types import SimpleNamespace
-from urllib.parse import urlparse
-
-from sglang.srt.utils import kill_process_tree
-from sglang.test.few_shot_gsm8k import run_eval as run_eval_few_shot_gsm8k
-from sglang.test.test_utils import (
-    DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
-    DEFAULT_URL_FOR_TEST,
-    CustomTestCase,
-    popen_launch_server,
-)
-
-TEST_MODEL_MATRIX = {
-    "Qwen/Qwen3-1.7B-GPTQ-Int8": {
-        "accuracy": 0.65,
-    },
-}
-
-
-class TestAscendGPTQInt8(CustomTestCase):
-
-    @classmethod
-    def setUpClass(cls):
-        cls.models = TEST_MODEL_MATRIX.keys()
-        cls.base_url = DEFAULT_URL_FOR_TEST
-        cls.url = urlparse(DEFAULT_URL_FOR_TEST)
-        cls.common_args = [
-            "--trust-remote-code",
-            "--mem-fraction-static",
-            0.8,
-            "--attention-backend",
-            "ascend",
-            "--quantization",
-            "gptq",
-        ]
-
-    def test_a_gsm8k(self):
-        for model in self.models:
-            with self.subTest(model=model):
-                print(f"##=== Testing accuracy: {model} ===##")
-
-                process = popen_launch_server(
-                    model,
-                    self.base_url,
-                    timeout=DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
-                    other_args=[
-                        *self.common_args,
-                    ],
-                )
-
-                try:
-                    args = SimpleNamespace(
-                        num_shots=5,
-                        data_path=None,
-                        num_questions=1319,
-                        max_new_tokens=512,
-                        parallel=128,
-                        host=f"http://{self.url.hostname}",
-                        port=int(self.url.port),
-                    )
-
-                    metrics = run_eval_few_shot_gsm8k(args)
-                    self.assertGreaterEqual(
-                        metrics["accuracy"],
-                        TEST_MODEL_MATRIX[model]["accuracy"],
-                    )
-                finally:
-                    kill_process_tree(process.pid)
-
-
-if __name__ == "__main__":
-    unittest.main()

test/srt/run_suite.py

@@ -85,7 +85,6 @@ suite_ascend = {
     "per-commit-1-npu-a2": [
         TestFile("ascend/test_ascend_autoround_dense.py", 400),
         TestFile("ascend/test_ascend_autoround_moe.py", 400),
-        TestFile("ascend/test_ascend_gptq.py", 400),
         TestFile("ascend/test_ascend_gptq_moe.py", 400),
         TestFile("ascend/test_ascend_graph_tp1_bf16.py", 400),
         TestFile("ascend/test_ascend_piecewise_graph_prefill.py", 400),