From 6d47f02a0fc2559ffb2de4e5eac9e519d0d97def Mon Sep 17 00:00:00 2001
From: Simon Gardling <titaniumtown@proton.me>
Date: Mon, 6 Apr 2026 02:29:37 -0400
Subject: [PATCH] llama-cpp: set batch size to 4096

---
 services/llama-cpp/llama-cpp.nix | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/services/llama-cpp/llama-cpp.nix b/services/llama-cpp/llama-cpp.nix
index 8e77a3d..86d6557 100644
--- a/services/llama-cpp/llama-cpp.nix
+++ b/services/llama-cpp/llama-cpp.nix
@@ -40,6 +40,10 @@ in
       "--metrics"
       "--alias"
       modelAlias
+      "-b"
+      "4096"
+      "-ub"
+      "4096"
     ];
   };