Split local attention test from fa3 test

test/srt/run_suite.py

@@ -30,6 +30,7 @@ class TestFile:
         TestFile("test_chunked_prefill.py", 336),
         TestFile("test_eagle_infer.py", 500),
         TestFile("test_ebnf_constrained.py"),
+        TestFile("test_fa3.py", 30),
         TestFile("test_fp8_kernel.py", 8),
         TestFile("test_embedding_openai_server.py", 36),
         TestFile("test_hidden_states.py", 55),
@@ -91,7 +92,7 @@ class TestFile:
         TestFile("test_verl_engine.py", 100),
     ],
     "per-commit-8-gpu": [
-        TestFile("test_fa3.py", 30),
+        TestFile("test_local_attn.py", 10),
     ],
     "nightly": [
         TestFile("test_nightly_gsm8k_eval.py"),

test/srt/test_fa3.py

@@ -10,7 +10,6 @@
 from sglang.test.test_utils import (
     DEFAULT_MODEL_NAME_FOR_TEST,
     DEFAULT_MODEL_NAME_FOR_TEST_EAGLE3,
-    DEFAULT_MODEL_NAME_FOR_TEST_LOCAL_ATTENTION,
     DEFAULT_MODEL_NAME_FOR_TEST_MLA,
     DEFAULT_MODEL_NAME_FOR_TEST_MLA_NEXTN,
     DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
@@ -127,22 +126,6 @@ def get_server_args(cls):
         return DEFAULT_SERVER_ARGS
 
 
-class TestFlashAttention3LocalAttn(BaseFlashAttentionTest):
-    """Test FlashAttention3 with Model with local attention, e.g. Llama 4."""
-
-    accuracy_threshold = 0.70
-    model = DEFAULT_MODEL_NAME_FOR_TEST_LOCAL_ATTENTION
-
-    @classmethod
-    def get_server_args(cls):
-        cloned_args = DEFAULT_SERVER_ARGS.copy()
-        # remove --enable-torch-compile from cloned_args since llama4 does not support it for now
-        cloned_args.remove("--enable-torch-compile")
-        # we cannot use scout's 10m context due to this bug: https://github.com/sgl-project/sglang/issues/5755
-        cloned_args.extend(["--tp", "4", "--context-length", "1000000"])
-        return cloned_args
-
-
 class TestFlashAttention3SpeculativeDecode(BaseFlashAttentionTest):
     """Test FlashAttention3 with speculative decode enabled with Llama 3.1 8B and its eagle3 model"""
 

test/srt/test_local_attn.py

@@ -0,0 +1,72 @@
+import os
+import unittest
+from types import SimpleNamespace
+
+import requests
+
+from sglang.srt.utils import get_device_sm, 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_MODEL_NAME_FOR_TEST_LOCAL_ATTENTION,
+    DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
+    DEFAULT_URL_FOR_TEST,
+    popen_launch_server,
+)
+
+
+@unittest.skipIf(get_device_sm() < 90, "Test requires CUDA SM 90 or higher")
+class TestFlashAttention3LocalAttn(unittest.TestCase):
+    model = DEFAULT_MODEL_NAME_FOR_TEST_LOCAL_ATTENTION
+    base_url = DEFAULT_URL_FOR_TEST
+    accuracy_threshold = 0.90
+
+    @classmethod
+    def get_server_args(cls):
+        return [
+            "--trust-remote-code",
+            "--cuda-graph-max-bs",
+            "2",
+            "--attention-backend",
+            "fa3",
+            "--tp",
+            "4",
+            "--context-length",
+            "1000000",
+        ]
+
+    @classmethod
+    def setUpClass(cls):
+        # disable deep gemm precompile to make launch server faster
+        # please don't do this if you want to make your inference workload faster
+        cls.process = popen_launch_server(
+            cls.model,
+            cls.base_url,
+            timeout=DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
+            other_args=cls.get_server_args(),
+            env=os.environ,
+        )
+
+    @classmethod
+    def tearDownClass(cls):
+        kill_process_tree(cls.process.pid)
+
+    def test_gsm8k(self):
+        args = SimpleNamespace(
+            num_shots=4,
+            num_questions=100,
+            max_new_tokens=512,
+            parallel=128,
+            host="http://127.0.0.1",
+            port=int(self.base_url.split(":")[-1]),
+            data_path=None,
+        )
+        metrics = run_eval_few_shot_gsm8k(args)
+        print(f"{metrics=}")
+
+        # Use the appropriate metric key based on the test class
+        metric_key = "accuracy"
+        self.assertGreater(metrics[metric_key], self.accuracy_threshold)
+
+
+if __name__ == "__main__":
+    unittest.main()