From 42737dac04d8c922c524a51fd451ea4c0b932bf2 Mon Sep 17 00:00:00 2001
From: Mihail Eric <mihaeric@amazon.com>
Date: Fri, 9 Oct 2020 14:58:45 -0700
Subject: [PATCH] add note about checkpoints

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

diff --git a/README.md b/README.md
index f512fb9..874eaeb 100644
--- a/README.md
+++ b/README.md
@@ -118,6 +118,10 @@ Almost all of the models can be trained/evaluated using the `run.py` script. Mul
 
 The commands for training/evaluating models are as follows. If you want to *only* run inference/evaluation, simply change `--num_epochs` to 0.
 
+### Checkpoints
+
+The relevant *convbert* and *convbert-dg* models can be found [here](https://registry.opendata.aws/dialoglue/).
+
 **HWU64**
 
 ```