Skip to content

Distribute and run LLMs with a single file.

Add i386 sanity check for Windows users

Justine Tunneyfd051081c525
diff --git a/llama.cpp/ggml-cuda.cu b/llama.cpp/ggml-cuda.cu
index ff57eddb..63487ce8 100644
--- a/llama.cpp/ggml-cuda.cu
+++ b/llama.cpp/ggml-cuda.cu
@@ -84,6 +84,10 @@
#include "ggml-cuda.h"
#include "ggml.h"
+#ifdef __i386__
+#error "you need to use a 64-bit compiler for llamafile"
+#endif
+
////////////////////////////////////////////////////////////////////////////////
// BEGIN: COPIED FROM GGML.C