llama : fix parameter order for hybrid memory initialization (#14725)
This commit is contained in:
Diner Burger authored and 
GitHub committed
1 parent
21c021745d
commit
496957e1cb
1 file changed
+1
-1
@@ -38,9 +38,9 @@ llama_memory_hybrid::llama_memory_hybrid(
|
||||
type_v,
|
||||
v_trans,
|
||||
offload,
|
||||
1,
|
||||
kv_size,
|
||||
n_seq_max,
|
||||
1,
|
||||
n_pad,
|
||||
n_swa,
|
||||
swa_type
|
||||
|
||||
Reference in new issue
Block a user