From 35b2800eef2436e71724f59b2788b57eeae9554a Mon Sep 17 00:00:00 2001
From: rainlizard <15337628+rainlizard@users.noreply.github.com>
Date: Fri, 25 Oct 2024 21:00:31 +1100
Subject: [PATCH] llama.cpp version of gguf-py no longer needed

---
 tools/README.md | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/tools/README.md b/tools/README.md
index 9c6fb15..e39e1d5 100644
--- a/tools/README.md
+++ b/tools/README.md
@@ -1,11 +1,9 @@
-This needs the llama.cpp version of gguf-py to work at the moment, not the pip one as that one does not have the python quantization code yet.
-
+# Quantizing Models
 ```
+pip install gguf
 git clone https://github.com/ggerganov/llama.cpp
-pip install llama.cpp/gguf-py
 ```
 
-
 To convert your initial source model to FP16 (or BF16), run the following command:
 ```
 python convert.py --src E:\models\unet\flux1-dev.safetensors