[Feature] Qwen3-Next & FLA: Support MTP topk>1; Up to 6% faster (#11133)

Co-authored-by: Stefan He <hebiaobuaa@gmail.com>

test/srt/models/test_qwen3_next_models.py

@@ -59,11 +59,56 @@ class TestQwen3NextMTP(CustomTestCase):
                 "--speculative-algorithm",
                 "NEXTN",
                 "--speculative-num-steps",
-                "1",
+                "3",
                 "--speculative-eagle-topk",
                 "1",
                 "--speculative-num-draft-tokens",
-                "2",
+                "4",
+                "--mem-fraction-static",
+                "0.8",
+                "--tp",
+                "4",
+            ],
+        )
+
+    @classmethod
+    def tearDownClass(cls):
+        kill_process_tree(cls.process.pid)
+
+    def test_gsm8k(self):
+        args = SimpleNamespace(
+            num_shots=5,
+            data_path=None,
+            num_questions=200,
+            max_new_tokens=512,
+            parallel=128,
+            host="http://127.0.0.1",
+            port=int(self.base_url.split(":")[-1]),
+        )
+        metrics = run_eval(args)
+        print(f"{metrics=}")
+        self.assertGreater(metrics["accuracy"], 0.93)
+
+
+class TestQwen3NextMTPTopk(CustomTestCase):
+    @classmethod
+    def setUpClass(cls):
+        cls.model = "Qwen/Qwen3-Next-80B-A3B-Instruct"
+        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=[
+                "--trust-remote-code",
+                "--speculative-algorithm",
+                "NEXTN",
+                "--speculative-num-steps",
+                "5",
+                "--speculative-eagle-topk",
+                "4",
+                "--speculative-num-draft-tokens",
+                "8",
                 "--mem-fraction-static",
                 "0.8",
                 "--tp",