[MM] Pass prefix parameter to MMEncoderAttention (#33674)
Signed-off-by: shen-shanshan <467638484@qq.com>
This commit is contained in:
@@ -127,7 +127,10 @@ class AIMv2Attention(nn.Module):
|
||||
self.num_heads_per_partition = divide(self.num_heads, self.tp_size)
|
||||
|
||||
self.attn = MMEncoderAttention(
|
||||
self.num_heads_per_partition, self.head_dim, self.scale
|
||||
self.num_heads_per_partition,
|
||||
self.head_dim,
|
||||
self.scale,
|
||||
prefix=prefix,
|
||||
)
|
||||
|
||||
def forward(self, x: torch.Tensor) -> torch.Tensor:
|
||||
|
||||
Reference in New Issue
Block a user