mirror of
https://github.com/huggingface/text-generation-inference.git
synced 2025-09-12 12:54:52 +00:00
fix in regression in ipex flashattention
Signed-off-by: Wang, Yi A <yi.a.wang@intel.com>
This commit is contained in:
parent
30e70f2ceb
commit
6abcab843d
@ -21,7 +21,7 @@ def attention(
|
||||
out = torch.empty_like(q)
|
||||
|
||||
# We do not need to check window_size_left (not supported) here, so it is already checked ahead of time at model load.
|
||||
return ipex.llm.functional.varlen_attention(
|
||||
ipex.llm.functional.varlen_attention(
|
||||
q,
|
||||
k,
|
||||
v,
|
||||
@ -38,6 +38,8 @@ def attention(
|
||||
None,
|
||||
)
|
||||
|
||||
return out
|
||||
|
||||
|
||||
def reshape_and_cache(
|
||||
key: torch.Tensor,
|
||||
|
Loading…
Reference in New Issue
Block a user