cuda : fix compile warning (llama/7454)

diff --git a/src/ggml-cuda/fattn-tile-f32.cu b/src/ggml-cuda/fattn-tile-f32.cu
index 54db765e2f8eebc57119b859ad61c2e28ddbf020..b8b2f69e19edb86c237be2cc99823cc1df8a2bc0 100644 (file)
--- a/src/ggml-cuda/fattn-tile-f32.cu
+++ b/src/ggml-cuda/fattn-tile-f32.cu
@@ -283,8 +283,7 @@ void launch_fattn_tile_f32_64_128(ggml_backend_cuda_context & ctx, ggml_tensor *
 }
 
 void ggml_cuda_flash_attn_ext_tile_f32(ggml_backend_cuda_context & ctx, ggml_tensor * dst) {
-    const ggml_tensor * KQV = dst;
-    const ggml_tensor * Q   = dst->src[0];
+    const ggml_tensor * Q = dst->src[0];
 
     if (Q->ne[1] <= 16) {
         constexpr int cols_per_block = 16;