Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in
Toggle navigation
Menu
Open sidebar
OpenDAS
vllm_cscc
Commits
8026a335
Unverified
Commit
8026a335
authored
Aug 01, 2025
by
Richard Zou
Committed by
GitHub
Aug 01, 2025
Browse files
[BugFix] Update AttnFusionPass cache key (#21947)
Signed-off-by:
Richard Zou
<
zou3519@gmail.com
>
parent
a59cd9d9
Changes
2
Hide whitespace changes
Inline
Side-by-side
Showing
2 changed files
with
5 additions
and
1 deletion
+5
-1
vllm/compilation/fusion_attn.py
vllm/compilation/fusion_attn.py
+3
-0
vllm/compilation/inductor_pass.py
vllm/compilation/inductor_pass.py
+2
-1
No files found.
vllm/compilation/fusion_attn.py
View file @
8026a335
...
...
@@ -164,3 +164,6 @@ class AttnFusionPass(VllmInductorPass):
logger
.
debug
(
"Fused quantization onto %s attention nodes"
,
count
)
self
.
dump_graph
(
graph
,
"after_attn_fusion"
)
self
.
end_and_log
()
def
uuid
(
self
):
return
VllmInductorPass
.
hash_source
(
self
,
AttentionStaticQuantPattern
)
vllm/compilation/inductor_pass.py
View file @
8026a335
...
...
@@ -76,9 +76,10 @@ class InductorPass(CustomGraphPass):
for
src
in
srcs
:
if
isinstance
(
src
,
str
):
src_str
=
src
elif
isinstance
(
src
,
types
.
FunctionType
):
elif
isinstance
(
src
,
(
types
.
FunctionType
,
type
)
):
src_str
=
inspect
.
getsource
(
src
)
else
:
# object instance
src_str
=
inspect
.
getsource
(
src
.
__class__
)
hasher
.
update
(
src_str
.
encode
(
"utf-8"
))
return
hasher
.
hexdigest
()
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment