summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorslaren <slarengh@gmail.com>2024-03-25 15:43:22 +0100
committerGitHub <noreply@github.com>2024-03-25 16:43:22 +0200
commit2f34b865b62b1d2b5eb8a27885e4de220deeacbd (patch)
treeedb1342a9622cd3bbd596f02b420d4f8c2751485
parentae1f211ce2138448b47ebb148e25c58406845278 (diff)
cuda : fix LLAMA_CUDA_F16 build (#6298)
-rw-r--r--ggml-cuda/dmmv.cu1
1 files changed, 1 insertions, 0 deletions
diff --git a/ggml-cuda/dmmv.cu b/ggml-cuda/dmmv.cu
index f91732df..7156c921 100644
--- a/ggml-cuda/dmmv.cu
+++ b/ggml-cuda/dmmv.cu
@@ -1,5 +1,6 @@
#include "dmmv.cuh"
#include "dequantize.cuh"
+#include "convert.cuh"
// dmmv = dequantize_mul_mat_vec
#ifndef GGML_CUDA_DMMV_X