Update model zoo.

Summary: Update model zoo. Reviewed By: lyttonhao Differential Revision: D30641908 fbshipit-source-id: dc75971188936fdfe97d637c3a95e5c4ff1d8e55
facebookresearch · Aug 30, 2021 · f5148e2 · f5148e2
1 parent d9ea0fa
commit f5148e2
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/pytorchvideo/models/hub/vision_transformers.py b/pytorchvideo/models/hub/vision_transformers.py
@@ -62,7 +62,7 @@ def mvit_base_16x4(
     """
     Multiscale Vision Transformers model architecture [1] trained with an 16x4
     setting on the Kinetics400 dataset. Model with pretrained weights has top1
-    accuracy of 79.0%.
+    accuracy of 78.9%.
 
     [1] Haoqi Fan, Bo Xiong, Karttikeya Mangalam, Yanghao Li, Zhicheng Yan, Jitendra
     Malik, Christoph Feichtenhofer, "Multiscale Vision Transformers"