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
9d5187eb
Commit
9d5187eb
authored
Jun 30, 2025
by
zhuwenwen
Browse files
解决PA 大batch,32k,崩溃问题
parent
95d58c0d
Changes
1
Show whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
1 addition
and
1 deletion
+1
-1
csrc/attention/attention_kernels_opt_tc.cu
csrc/attention/attention_kernels_opt_tc.cu
+1
-1
No files found.
csrc/attention/attention_kernels_opt_tc.cu
View file @
9d5187eb
...
...
@@ -965,7 +965,7 @@ void get_numberthread_and_reuse_kv_v2(int& num_thread,int& reusekv,int& PARTITIO
)
max_num_partitions
=
1
;
int
blocks
=
max_num_partitions
*
batchsize
*
qheads
;
if
(
device_name
==
"gfx928"
){
if
(
device_name
==
"gfx928"
||
batchsize
>
100
){
if
(
batchsize
*
qheads
>
1024
&&
max_seq_len
>=
2000
){
max_num_partitions
=
1
;
if
(
max_seq_len
<
2000
)
reusekv
=
8
;
...
...
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