We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
1 parent ac5e0c0 commit 410171aCopy full SHA for 410171a
ggml/src/ggml-cuda/conv2d.cuh
@@ -1,9 +1,9 @@
1
#pragma once
2
#include "common.cuh"
3
4
-#define BS_OC 64
+#define BS_OC 16
5
#define BS_ICKHKW 16
6
-#define BS_NOHOW 64
+#define BS_NOHOW 128
7
8
#define WMMA_M 16
9
#define WMMA_N 16
0 commit comments