mirror of
https://github.com/huggingface/text-generation-inference.git
synced 2025-09-11 12:24:53 +00:00
add contiguous
This commit is contained in:
parent
5088005908
commit
4a02d3505f
@ -222,8 +222,8 @@ class FlashCohereAttention(torch.nn.Module):
|
|||||||
dim=1,
|
dim=1,
|
||||||
)
|
)
|
||||||
if self.use_qk_norm:
|
if self.use_qk_norm:
|
||||||
query = self.q_norm(query)
|
query = self.q_norm(query.contiguous())
|
||||||
key = self.k_norm(key)
|
key = self.k_norm(key.contiguous())
|
||||||
|
|
||||||
query = query.view(-1, self.num_heads, self.head_size)
|
query = query.view(-1, self.num_heads, self.head_size)
|
||||||
key = key.view(-1, self.num_key_value_heads, self.head_size)
|
key = key.view(-1, self.num_key_value_heads, self.head_size)
|
||||||
|
Loading…
Reference in New Issue
Block a user