From 4b61ab9de5acffa297dacb324447fedf1d9a2b5a Mon Sep 17 00:00:00 2001
From: Kai Zhang <kaizh@fb.com>
Date: Fri, 22 Oct 2021 16:22:21 -0700
Subject: [PATCH] [easy] Expand usage logging to video model

Summary: As titled.

Reviewed By: sallysyw

Differential Revision: D31813675

fbshipit-source-id: 8ff879be172530bcc41aae5d71c67e1e90f09cbf
---
 torchvision/models/video/resnet.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/torchvision/models/video/resnet.py b/torchvision/models/video/resnet.py
index 5cfbbaeb559..e2e4eb420e7 100644
--- a/torchvision/models/video/resnet.py
+++ b/torchvision/models/video/resnet.py
@@ -4,7 +4,7 @@
 from torch import Tensor
 
 from ..._internally_replaced_utils import load_state_dict_from_url
-
+from ...utils import _log_api_usage_once
 
 __all__ = ["r3d_18", "mc3_18", "r2plus1d_18"]
 
@@ -209,6 +209,7 @@ def __init__(
             zero_init_residual (bool, optional): Zero init bottleneck residual BN. Defaults to False.
         """
         super(VideoResNet, self).__init__()
+        _log_api_usage_once(self)
         self.inplanes = 64
 
         self.stem = stem()