diff --git a/examples/lookahead/README.md b/examples/lookahead/README.md
new file mode 100644
index 000000000..252a6689e
--- /dev/null
+++ b/examples/lookahead/README.md
@@ -0,0 +1,7 @@
+# llama.cpp/examples/lookahead
+
+Demonstartion of lookahead decoding technique:
+
+https://lmsys.org/blog/2023-11-21-lookahead-decoding/
+
+More info: https://github.com/ggerganov/llama.cpp/pull/4207
diff --git a/examples/speculative/README.md b/examples/speculative/README.md
new file mode 100644
index 000000000..d88fd3790
--- /dev/null
+++ b/examples/speculative/README.md
@@ -0,0 +1,8 @@
+# llama.cpp/examples/speculative
+
+Demonstartion of speculative decoding and tree-based speculative decoding techniques
+
+More info:
+
+- https://github.com/ggerganov/llama.cpp/pull/2926
+- https://github.com/ggerganov/llama.cpp/pull/3624
diff --git a/ggml.h b/ggml.h
index e27a52ba8..59352541e 100644
--- a/ggml.h
+++ b/ggml.h
@@ -251,11 +251,10 @@
 #define GGML_ASSERT(x) \
     do { \
         if (!(x)) { \
-            fprintf(stderr, "GGML_ASSERT: %s:%d: %s\n", __FILE__, __LINE__, #x); \
-            fflush(stderr); \
             fflush(stdout); \
+            fprintf(stderr, "GGML_ASSERT: %s:%d: %s\n", __FILE__, __LINE__, #x); \
             ggml_print_backtrace(); \
-            exit(1); \
+            abort(); \
         } \
     } while (0)