Enable RoPE+KV cache fusion for ROCm AITER FA (non-shuffle layout) (#35786)
Signed-off-by: Rohan138 <rohanpotdar138@gmail.com>
This commit is contained in:
@@ -196,6 +196,7 @@ class QKRoPEKVCacheTestModel(torch.nn.Module):
|
||||
AttentionBackendEnum.ROCM_AITER_UNIFIED_ATTN,
|
||||
AttentionBackendEnum.TRITON_ATTN,
|
||||
AttentionBackendEnum.ROCM_ATTN,
|
||||
AttentionBackendEnum.ROCM_AITER_FA,
|
||||
],
|
||||
)
|
||||
@pytest.mark.parametrize("enable_rope_custom_op", [True]) # [True, False])
|
||||
|
||||
Reference in New Issue
Block a user