[npu]add chunk gdn kernel and unify ssm state layout for ascend gdn backend (#36187)
This commit is contained in:
@@ -38,8 +38,6 @@ class TestAscendMhaHicache(CustomTestCase):
|
||||
"--attention-backend",
|
||||
"ascend",
|
||||
"--enable-hierarchical-cache",
|
||||
"--hicache-ratio",
|
||||
1.2,
|
||||
]
|
||||
|
||||
def test_a_gsm8k(self):
|
||||
|
||||
@@ -40,8 +40,6 @@ class TestAscendMlaHicache(CustomTestCase):
|
||||
"--tp-size",
|
||||
4,
|
||||
"--enable-hierarchical-cache",
|
||||
"--hicache-ratio",
|
||||
1.2,
|
||||
]
|
||||
|
||||
def test_a_gsm8k(self):
|
||||
|
||||
Reference in New Issue
Block a user