From 12fe2002a54bc90fe5cb758e80ecb36e236d96a4 Mon Sep 17 00:00:00 2001
From: PENG Bo <33809201+BlinkDL@users.noreply.github.com>
Date: Wed, 13 Apr 2022 14:51:50 +0800
Subject: [PATCH] Update README.md

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

diff --git a/README.md b/README.md
index 435355c..fd4a1fd 100644
--- a/README.md
+++ b/README.md
@@ -34,6 +34,8 @@ I propose a simple new method to find better LR schedules. The method is cost-ef
 
 ![better_lr_schedule](Research/better_lr_schedule.png)
 
+UPDATE: In "Conclusion 1.", remember to use the best-fitting regime (ignore the initial steps where our approximations break down) to fit the parameters.
+
 # The top-p-x sampling method
 
 We propose a new sampling method called top-p-x: