fix: enable multi-threading for h200 tests (#16413)

2026-01-04 14:43:53 -08:00
parent a1ed247fd7
commit 55d112dc79
6 changed files with 23 additions and 1 deletions
@@ -28,6 +28,8 @@ class TestDeepseekV32DP(CustomTestCase):
            "--dp",
            "8",
            "--enable-dp-attention",
+            "--model-loader-extra-config",
+            '{"enable_multithread_load": true, "num_threads": 64}',
        ]
        cls.process = popen_launch_server(
            cls.model,
@@ -83,6 +85,8 @@ class TestDeepseekV32TP(CustomTestCase):
            "--trust-remote-code",
            "--tp",
            "8",
+            "--model-loader-extra-config",
+            '{"enable_multithread_load": true, "num_threads": 64}',
        ]
        cls.process = popen_launch_server(
            cls.model,