[PD & HiSparse] Add DeepSeek V4 support for HiSparse direct Prefill-to-Decode DRAM (#24880)

test/registered/disaggregation/test_disaggregation_dsv4.py

@@ -11,11 +11,15 @@ from sglang.test.test_utils import (
     try_cached_model,
 )
 
-register_cuda_ci(est_time=250, stage="base-c", runner_config="deepep-8-gpu-h200")
+register_cuda_ci(est_time=500, stage="base-c", runner_config="deepep-8-gpu-h200")
 
 DSV4_FLASH_MODEL = "sgl-project/DeepSeek-V4-Flash-FP8"
 
 DEEPEP_CONFIG = '{"normal_dispatch":{"num_sms":96},"normal_combine":{"num_sms":96}}'
+DSV4_FLASH_LOADER_CONFIG = '{"enable_multithread_load": true, "num_threads": 64}'
+DSV4_HISPARSE_CONFIG = (
+    '{"top_k":512,"device_buffer_size":4096,"host_to_device_ratio":2}'
+)
 
 DSV4_FLASH_ENV = {
     "SGLANG_DSV4_FP4_EXPERTS": "0",
@@ -123,5 +127,105 @@ class TestDisaggregationDSV4(PDDisaggregationServerBase, GSM8KMixin):
         )
 
 
+class TestDisaggregationDSV4HiSparseMooncake(PDDisaggregationServerBase, GSM8KMixin):
+
+    gsm8k_accuracy_thres = 0.93
+    gsm8k_num_questions = 200
+    gsm8k_num_shots = 20
+
+    @classmethod
+    def setUpClass(cls):
+        super().setUpClass()
+
+        cls.model = try_cached_model(DSV4_FLASH_MODEL)
+
+        cls.start_prefill()
+        cls.start_decode()
+
+        cls.wait_server_ready(cls.prefill_url + "/health", process=cls.process_prefill)
+        cls.wait_server_ready(cls.decode_url + "/health", process=cls.process_decode)
+
+        cls.launch_lb()
+
+    @classmethod
+    def start_prefill(cls):
+        prefill_args = [
+            "--trust-remote-code",
+            "--disaggregation-mode",
+            "prefill",
+            "--disaggregation-bootstrap-port",
+            cls.bootstrap_port,
+            "--tp",
+            4,
+            "--page-size",
+            256,
+            "--chunked-prefill-size",
+            8192,
+            "--max-running-requests",
+            16,
+            "--mem-fraction-static",
+            0.9,
+            "--skip-server-warmup",
+            "--reasoning-parser",
+            "deepseek-v4",
+            "--tool-call-parser",
+            "deepseekv4",
+            "--model-loader-extra-config",
+            DSV4_FLASH_LOADER_CONFIG,
+            "--watchdog-timeout",
+            "900",
+        ]
+        prefill_args += cls.transfer_backend + cls.rdma_devices
+        cls.process_prefill = popen_launch_pd_server(
+            cls.model,
+            cls.prefill_url,
+            timeout=DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
+            other_args=prefill_args,
+            env=DSV4_FLASH_ENV,
+        )
+
+    @classmethod
+    def start_decode(cls):
+        decode_args = [
+            "--trust-remote-code",
+            "--disaggregation-mode",
+            "decode",
+            "--disaggregation-bootstrap-port",
+            cls.bootstrap_port,
+            "--tp",
+            4,
+            "--base-gpu-id",
+            4,
+            "--page-size",
+            256,
+            "--chunked-prefill-size",
+            8192,
+            "--max-running-requests",
+            16,
+            "--mem-fraction-static",
+            0.9,
+            "--skip-server-warmup",
+            "--reasoning-parser",
+            "deepseek-v4",
+            "--tool-call-parser",
+            "deepseekv4",
+            "--model-loader-extra-config",
+            DSV4_FLASH_LOADER_CONFIG,
+            "--enable-hisparse",
+            "--hisparse-config",
+            DSV4_HISPARSE_CONFIG,
+            "--watchdog-timeout",
+            "900",
+        ]
+        decode_args += cls.transfer_backend + cls.rdma_devices
+        cls.process_decode = popen_launch_pd_server(
+            cls.model,
+            cls.decode_url,
+            timeout=DEFAULT_TIMEOUT_FOR_SERVER_LAUNCH,
+            other_args=decode_args,
+            env=DSV4_FLASH_ENV,
+        )
+
+
 if __name__ == "__main__":
     unittest.main()