diff --git a/ggml-cuda.cu b/ggml-cuda.cu
index 96f51a04c..856ebf1c1 100644
--- a/ggml-cuda.cu
+++ b/ggml-cuda.cu
@@ -4332,7 +4332,7 @@ static void * ggml_cuda_pool_malloc(size_t size, size_t * actual_size) {
     }
     void * ptr;
 
-    size_t look_ahead_size = (size_t) (1.02 * size);
+    size_t look_ahead_size = (size_t) (1.05 * size);
     look_ahead_size = 256 * ((look_ahead_size + 255)/256);
     CUDA_CHECK(cudaMalloc((void **) &ptr, look_ahead_size));
     *actual_size = look_ahead_size;
diff --git a/koboldcpp.py b/koboldcpp.py
index 34a619cd7..40441cffa 100755
--- a/koboldcpp.py
+++ b/koboldcpp.py
@@ -303,7 +303,7 @@ maxhordectx = 1024
 maxhordelen = 256
 modelbusy = threading.Lock()
 defaultport = 5001
-KcppVersion = "1.39.1"
+KcppVersion = "1.39.2"
 showdebug = True
 showsamplerwarning = True
 showmaxctxwarning = True