Add pytorch sampling backend ut (#1425)

test/srt/run_suite.py

@@ -17,6 +17,7 @@ suites = {
         "test_skip_tokenizer_init.py",
         "test_torch_compile.py",
         "test_triton_attn_backend.py",
+        "test_pytorch_sampling_backend.py",
         "test_update_weights.py",
         "test_vision_openai_server.py",
         "test_server_args.py",

test/srt/test_pytorch_sampling_backend.py

+import unittest
+from types import SimpleNamespace
+
+from sglang.srt.utils import kill_child_process
+from sglang.test.run_eval import run_eval
+from sglang.test.test_utils import (
+    DEFAULT_MODEL_NAME_FOR_TEST,
+    DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
+    DEFAULT_URL_FOR_TEST,
+    popen_launch_server,
+)
+
+
+class TestPyTorchSamplingBackend(unittest.TestCase):
+    @classmethod
+    def setUpClass(cls):
+        cls.model = DEFAULT_MODEL_NAME_FOR_TEST
+        cls.base_url = DEFAULT_URL_FOR_TEST
+        cls.process = popen_launch_server(
+            cls.model,
+            cls.base_url,
+            timeout=DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
+            other_args=["--sampling-backend", "pytorch"],
+        )
+
+    @classmethod
+    def tearDownClass(cls):
+        kill_child_process(cls.process.pid)
+
+    def test_mmlu(self):
+        args = SimpleNamespace(
+            base_url=self.base_url,
+            model=self.model,
+            eval_name="mmlu",
+            num_examples=64,
+            num_threads=32,
+        )
+
+        metrics = run_eval(args)
+        assert metrics["score"] >= 0.65
+
+
+if __name__ == "__main__":
+    unittest.main()