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
b2fd0b81
Unverified
Commit
b2fd0b81
authored
Aug 18, 2025
by
Andy Lo
Committed by
GitHub
Aug 17, 2025
Browse files
[Bugfix][CI] Machete kernels: deterministic ordering for more cache hits (#23055)
Signed-off-by:
Andy Lo
<
andy@mistral.ai
>
parent
9f1c6422
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
6 additions
and
3 deletions
+6
-3
csrc/quantization/machete/generate.py
csrc/quantization/machete/generate.py
+6
-3
No files found.
csrc/quantization/machete/generate.py
View file @
b2fd0b81
...
@@ -349,9 +349,12 @@ def to_cute_constant(value: list[int]):
...
@@ -349,9 +349,12 @@ def to_cute_constant(value: list[int]):
def
unique_schedules
(
impl_configs
:
list
[
ImplConfig
]):
def
unique_schedules
(
impl_configs
:
list
[
ImplConfig
]):
return
list
(
# Use dict over set for deterministic ordering
set
(
sch
for
impl_config
in
impl_configs
return
list
({
for
sch
in
impl_config
.
schedules
))
sch
:
None
for
impl_config
in
impl_configs
for
sch
in
impl_config
.
schedules
}.
keys
())
def
unsigned_type_with_bitwidth
(
num_bits
):
def
unsigned_type_with_bitwidth
(
num_bits
):
...
...
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