[NVIDIA] [GDN] Add FlashInfer prefill support for SM100+ (Blackwell) (#22921)
This commit is contained in:
@@ -45,12 +45,6 @@ class TestQwen35FP4(CustomTestCase):
|
||||
extra_args=base_args,
|
||||
variant="Triton",
|
||||
),
|
||||
# TODO: Fix this and re-enable it
|
||||
# ModelLaunchSettings(
|
||||
# QWEN35_FP4_MODEL,
|
||||
# extra_args=base_args + ["--linear-attn-decode-backend", "flashinfer"],
|
||||
# variant="FlashInfer",
|
||||
# ),
|
||||
]
|
||||
|
||||
run_combined_tests(
|
||||
|
||||
Reference in New Issue
Block a user