Bump FlashInfer to 0.6.15 and revert regressions (#31502)

Co-authored-by: Brayden Zhong <b8zhong@uwaterloo.ca>
2026-07-17 00:50:12 -07:00
parent e835512303
commit d67aa05697
7 changed files with 20 additions and 15 deletions
@@ -21,7 +21,7 @@ class TestGLM52NVFP4TPMTP(
    model = "nvidia/GLM-5.2-NVFP4"
    tp_size = 4
    mem_fraction_static = 0.8
-    bs_1_speed_thres = 250
+    bs_1_speed_thres = 280
    extra_server_args = [
        "--moe-runner-backend",
        "flashinfer_trtllm",