xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-07-09 11:27:14 +08:00
Code Issues Packages Projects Releases Wiki Activity

precommit

Browse Source 
Signed-off-by: NickLucche <nlucches@redhat.com>
This commit is contained in:
NickLucche 2025-12-19 12:45:12 +00:00
parent 7c95fd8279
commit 1429a5e9a8
1 changed files with 6 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
vllm/attention/ops/vit_attn_wrappers.py
View File

@ -110,7 +110,12 @@ def vit_flash_attn_wrapper(
) )
def apply_sdpa(q: torch.Tensor, k: torch.Tensor, v: torch.Tensor, softmax_scale: float | None = None) -> torch.Tensor: def apply_sdpa(
q: torch.Tensor,
k: torch.Tensor,
v: torch.Tensor,
softmax_scale: float | None = None,
) -> torch.Tensor:
""" """
Input shape: Input shape:
(batch_size x seq_len x num_heads x head_size) (batch_size x seq_len x num_heads x head_size)