[CI] Add Mistral Large 3 Eagle basic PR test

python/sglang/srt/models/mistral_large_3_eagle.py

@@ -4,8 +4,8 @@
 from torch import nn
 from transformers import PretrainedConfig
 
-from python.sglang.srt.layers.attention.nsa.utils import is_nsa_enable_prefill_cp
 from sglang.srt.distributed import get_pp_group
+from sglang.srt.layers.attention.nsa.utils import is_nsa_enable_prefill_cp
 from sglang.srt.layers.layernorm import RMSNorm
 from sglang.srt.layers.linear import RowParallelLinear
 from sglang.srt.layers.quantization.base_config import QuantizationConfig

scripts/ci/slash_command_handler.py

@@ -161,6 +161,7 @@ def handle_rerun_stage(
         "unit-test-deepep-8-gpu",
         "unit-test-backend-4-gpu-b200",
         "unit-test-backend-4-gpu-gb200",
+        "unit-test-backend-8-gpu-b200",
     ]
 
     # Valid AMD stage names that support target_stage

test/srt/run_suite.py

@@ -182,6 +182,7 @@
     ],
     "per-commit-8-gpu-b200": [
         TestFile("test_mistral_large3_basic.py", 275),
+        TestFile("test_mistral_large3_eagle_basic.py", 275),
     ],
     "per-commit-4-gpu-gb200": [
         TestFile("test_cutedsl_moe.py", 300),

test/srt/test_mistral_large3_eagle_basic.py

@@ -0,0 +1,105 @@
+import os
+import unittest
+from types import SimpleNamespace
+
+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.send_one import BenchArgs, send_one_prompt
+from sglang.test.test_utils import (
+    DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
+    DEFAULT_URL_FOR_TEST,
+    CustomTestCase,
+    is_in_ci,
+    popen_launch_server,
+    write_github_step_summary,
+)
+
+# Base model and Eagle draft model
+MISTRAL_LARGE3_MODEL_PATH = "mistralai/Mistral-Large-3-675B-Instruct-2512"
+MISTRAL_LARGE3_EAGLE_MODEL_PATH = "mistralai/Mistral-Large-3-675B-Instruct-2512-Eagle"
+
+
+class TestMistralLarge3EagleBasic(CustomTestCase):
+    @classmethod
+    def setUpClass(cls):
+        # Set environment variable to disable JIT DeepGemm
+        os.environ["SGLANG_ENABLE_JIT_DEEPGEMM"] = "0"
+
+        cls.model = MISTRAL_LARGE3_MODEL_PATH
+        cls.base_url = DEFAULT_URL_FOR_TEST
+        # Mistral-Large-3 with Eagle speculative decoding
+        # Eagle model is used as draft model for speculative decoding
+        other_args = [
+            "--tp",
+            "8",
+            "--attention-backend",
+            "trtllm_mla",
+            "--speculative-algorithm",
+            "EAGLE",
+            "--speculative-draft-model-path",
+            MISTRAL_LARGE3_EAGLE_MODEL_PATH,
+            "--speculative-num-steps",
+            "3",
+            "--speculative-eagle-topk",
+            "1",
+            "--speculative-num-draft-tokens",
+            "4",
+            "--kv-cache-dtype",
+            "auto",
+            "--model-loader-extra-config",
+            '{"enable_multithread_load": true}',
+            "--chat-template",
+            "mistral",
+        ]
+        cls.process = popen_launch_server(
+            cls.model,
+            cls.base_url,
+            timeout=DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH * 5,
+            other_args=other_args,
+        )
+
+    @classmethod
+    def tearDownClass(cls):
+        kill_process_tree(cls.process.pid)
+        # Clean up environment variable
+        if "SGLANG_ENABLE_JIT_DEEPGEMM" in os.environ:
+            del os.environ["SGLANG_ENABLE_JIT_DEEPGEMM"]
+
+    def test_a_gsm8k(
+        self,
+    ):  # Append an "a" to make this test run first (alphabetically) to warm up the server
+        args = SimpleNamespace(
+            num_shots=8,
+            data_path=None,
+            num_questions=1400,
+            parallel=1400,
+            max_new_tokens=512,
+            host="http://127.0.0.1",
+            port=int(self.base_url.split(":")[-1]),
+        )
+        metrics = run_eval_few_shot_gsm8k(args)
+        print(f"{metrics=}")
+
+        if is_in_ci():
+            write_github_step_summary(
+                f"### test_gsm8k (mistral-large-3-eagle)\n"
+                f'{metrics["accuracy"]=:.3f}\n'
+            )
+            self.assertGreater(metrics["accuracy"], 0.90)
+
+    def test_bs_1_speed(self):
+        args = BenchArgs(port=int(self.base_url.split(":")[-1]), max_new_tokens=2048)
+        acc_length, speed = send_one_prompt(args)
+
+        print(f"{speed=:.2f}")
+
+        if is_in_ci():
+            write_github_step_summary(
+                f"### test_bs_1_speed (mistral-large-3-eagle)\n"
+                f"{speed=:.2f} token/s\n"
+            )
+            self.assertGreater(speed, 50)
+
+
+if __name__ == "__main__":
+    unittest.main()