X-Git-Url: https://us-delaporte.pagekite.me/gitweb/dotfiles.git/blobdiff_plain/3823afa90f8688a99c1d522938687123da0324c8..4faa33c4d7bd12a90d67f64e62781e5ae9e3b502:/files/llama-cpp.ini

diff --git a/files/llama-cpp.ini b/files/llama-cpp.ini
index 3c5657a..f260af2 100644
--- a/files/llama-cpp.ini
+++ b/files/llama-cpp.ini
@@ -3,6 +3,17 @@ version = 1
 ; (Optional) This section provides global settings shared across all presets.
 ; If the same key is defined in a specific preset, it will override the value in this global section.
 [*]
-c = 8192
-n-gpu-layers = 8
+c = 8192  # works good
+; c = 16400  # works less well
+# c = 2048  # works terrible
+
+
+; When you ask for too many layers, llama.cpp fails at load time with a CUDA out of memory error. It will not silently fall back. It stops.
+
+; The fix is to drop ngl by a few and reload. Step down until it loads. If you are right at the edge, shave 2 or 3 layers and try again.
+
+; Watch your VRAM with nvidia-smi while the model loads. You want a buffer left over, not a card pinned at 100 percent. Other apps, your desktop, and the KV cache all want a slice.
+
+; 8 was fine, 16 was fine
+n-gpu-layers = 16