From 37d73f4aaaece167992bfdf3c24a44bba44824e6 Mon Sep 17 00:00:00 2001
From: Fiona Ryan <31247608+fkryan@users.noreply.github.com>
Date: Wed, 11 Dec 2024 15:53:25 -0500
Subject: [PATCH] Update README.md

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

diff --git a/README.md b/README.md
index 41a91ff..4a6b9bf 100644
--- a/README.md
+++ b/README.md
@@ -35,9 +35,9 @@ pip3 install -U xformers --index-url https://download.pytorch.org/whl/cu118
 We provide the following pretrained models for download.
 | Name | Backbone type | Backbone name | Training data | Checkpoint |
 | ---- | ------------- | ------------- |-------------- | ---------- |
-| ```gazelle_dinov2_vitb14``` | DINOv2 ViT-B | ```dinov2_vitb14 ```| GazeFollow | [Download](https://github.com/fkryan/gazelle/releases/download/v1.0.0/gazelle_dinov2_vitb14.pt) |
+| ```gazelle_dinov2_vitb14``` | DINOv2 ViT-B | ```dinov2_vitb14```| GazeFollow | [Download](https://github.com/fkryan/gazelle/releases/download/v1.0.0/gazelle_dinov2_vitb14.pt) |
 | ```gazelle_dinov2_vitl14``` | DINOv2 ViT-L | ```dinov2_vitl14``` | GazeFollow | [Download](https://github.com/fkryan/gazelle/releases/download/v1.0.0/gazelle_dinov2_vitl14.pt) |
-| ```gazelle_dinov2_vitb14_inout``` | DINOv2 ViT-B | ```dinov2_vitb14 ``` | Gazefollow -> VideoAttentionTarget | [Download](https://github.com/fkryan/gazelle/releases/download/v1.0.0/gazelle_dinov2_vitb14_inout.pt) |
+| ```gazelle_dinov2_vitb14_inout``` | DINOv2 ViT-B | ```dinov2_vitb14``` | Gazefollow -> VideoAttentionTarget | [Download](https://github.com/fkryan/gazelle/releases/download/v1.0.0/gazelle_dinov2_vitb14_inout.pt) |
 | ```gazelle_large_vitl14_inout``` | DINOv2-ViT-L | ```dinov2_vitl14```  | GazeFollow -> VideoAttentionTarget | [Download](https://github.com/fkryan/gazelle/releases/download/v1.0.0/gazelle_dinov2_vitl14_inout.pt) |
 
 Note that our Gaze-LLE checkpoints contain only the gaze decoder weights - the DINOv2 backbone weights are downloaded from ```facebookresearch/dinov2``` on PyTorch Hub when the Gaze-LLE model is created in our code.