[CI] Disable the prefill CUDA graph on the P worker of test_kimi_linear_pd_dcp4 (#34779)

Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-13 15:26:21 -07:00
parent 81fe452810
commit f3beb2c529
1 changed files with 2 additions and 0 deletions
@@ -306,6 +306,8 @@ class TestKimiLinearPDDCP4(GSM8KMixin, PDDisaggregationServerBase):
            str(PHYSICAL_PAGE_SIZE),
            "--chunked-prefill-size",
            str(CHUNKED_PREFILL_SIZE),
+            "--cuda-graph-backend-prefill",
+            "disabled",
            "--mem-fraction-static",
            "0.80",
        ]