mirror of
https://git.datalinker.icu/vllm-project/vllm.git
synced 2025-12-09 21:55:50 +08:00
[Kernel] Add punica dimensions for Granite 3b and 8b (#5930)
Signed-off-by: Joe Runde <joe@joerun.de>
This commit is contained in:
parent
906a19cdb0
commit
ba4994443a
@ -86,6 +86,7 @@ void bgmv_kernel(out_T *__restrict__ Y, const in_T *__restrict__ X,
|
||||
f(in_T, out_T, W_T, narrow, 36864) \
|
||||
f(in_T, out_T, W_T, narrow, 43264) \
|
||||
f(in_T, out_T, W_T, narrow, 49152) \
|
||||
f(in_T, out_T, W_T, narrow, 49408) \
|
||||
f(in_T, out_T, W_T, narrow, 60544) \
|
||||
f(in_T, out_T, W_T, narrow, 60672) \
|
||||
f(in_T, out_T, W_T, narrow, 64000) \
|
||||
@ -182,6 +183,7 @@ void bgmv_kernel(out_T *__restrict__ Y, const in_T *__restrict__ X,
|
||||
f(in_T, out_T, W_T, 36864, narrow) \
|
||||
f(in_T, out_T, W_T, 43264, narrow) \
|
||||
f(in_T, out_T, W_T, 49152, narrow) \
|
||||
f(in_T, out_T, W_T, 49408, narrow) \
|
||||
f(in_T, out_T, W_T, 60544, narrow) \
|
||||
f(in_T, out_T, W_T, 60672, narrow) \
|
||||
f(in_T, out_T, W_T, 64000, narrow) \
|
||||
|
||||
@ -111,6 +111,7 @@ H1 = H2 = [
|
||||
36864,
|
||||
43264,
|
||||
49152,
|
||||
49408,
|
||||
60544,
|
||||
60672,
|
||||
64000,
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user