From da1a4ff01f42d058cfa59806dd5679c0fe5a8604 Mon Sep 17 00:00:00 2001
From: Georgi Gerganov <ggerganov@gmail.com>
Date: Sun, 12 Mar 2023 01:26:32 +0200
Subject: [PATCH] Update README.md

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

diff --git a/README.md b/README.md
index deb80fa29cb5b..5194f6efc7b9d 100644
--- a/README.md
+++ b/README.md
@@ -2,9 +2,9 @@
 
 Inference of [Facebook's LLaMA](https://github.com/facebookresearch/llama) model in pure C/C++
 
-**!!! IMPORTANT !!!**
+**Hot topics**
 
-Commit [007a8f6f459c6eb56678fdee4c09219ddb85b640](https://github.com/ggerganov/llama.cpp/commit/007a8f6f459c6eb56678fdee4c09219ddb85b640) added support for all LLaMA models, but introduced breaking changes. If you generated any models before that commit, you must regenerate them after updating to latest master.
+- Running on Windows: https://github.com/ggerganov/llama.cpp/issues/22
 
 ## Description