From 90dc8a91ae6d5ba1d334cd07f2d6af030f14caaf Mon Sep 17 00:00:00 2001
From: oobabooga <112222186+oobabooga@users.noreply.github.com>
Date: Thu, 1 Jun 2023 11:57:57 -0300
Subject: [PATCH] Update llama.cpp-models.md

---
 docs/llama.cpp-models.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/docs/llama.cpp-models.md b/docs/llama.cpp-models.md
index 153f70af..be368fd3 100644
--- a/docs/llama.cpp-models.md
+++ b/docs/llama.cpp-models.md
@@ -12,9 +12,9 @@ Follow the instructions in the llama.cpp README to generate the `ggml-model.bin`
 
 ## GPU offloading
 
-Enabled with the `--n-gpu-layers` parameter. If you have enough VRAM, use a high number like `--n-gpu-layers 200000` to offload all layers to the GPU.
+Enabled with the `--n-gpu-layers` parameter. If you have enough VRAM, use a high number like `--n-gpu-layers 200000` to offload all layers to the GPU. Otherwise, start with a low value like `--n-gpu-layers 10` and gradually increase it until you run out of memory.
 
-Note that you need to manually install `llama-cpp-python` with GPU support. To do that:
+To use this feature, you need to manually compile and install `llama-cpp-python` with GPU support.
 
 #### Linux