diff --git a/README.md b/README.md
index 97fc2cb3de054be27c571c24583df13182271143..afb3bb3086b9f6b3c7b84eb031688b7d1339d3d7 100644
--- a/README.md
+++ b/README.md
@@ -1,3 +1,7 @@
+## *** For readers of "Out-of-Task Training for Dialog State Tracking Models" ***
+
+The code update to cover the content of the paper (https://arxiv.org/abs/2011.09379) will be available here after our presentation at COLING 2020
+
 ## Introduction
 
 TripPy is a new approach to dialogue state tracking (DST) which makes use of various copy mechanisms to fill slots with values. Our model has no need to maintain a list of candidate values. Instead, all values are extracted from the dialog context on-the-fly.
@@ -27,8 +31,9 @@ With a sequence length of 180, you should expect the following average JGA:
 
 ## Requirements
 
-- transformers
-- tensorboardX
+- torch (tested: 1.4.0)
+- transformers (tested: 2.0)
+- tensorboardX (tested: 2.9.1)
 
 ## Citation