From: A. Unique TensorFlower <gardener@tensorflow.org>
Date: Mon, 5 Feb 2018 23:30:54 +0000 (-0800)
Subject: Assign total_loss in order not to crash if training loop exists early.
X-Git-Tag: upstream/v1.7.0~31^2~993
X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=395550bc423f6a5d9c96233f34f21fce6bd23b4e;p=platform%2Fupstream%2Ftensorflow.git

Assign total_loss in order not to crash if training loop exists early.

PiperOrigin-RevId: 184596877
---

diff --git a/tensorflow/contrib/slim/python/slim/learning.py b/tensorflow/contrib/slim/python/slim/learning.py
index 54362c8..83f3380 100644
--- a/tensorflow/contrib/slim/python/slim/learning.py
+++ b/tensorflow/contrib/slim/python/slim/learning.py
@@ -738,6 +738,7 @@ def train(train_op,
   if summary_writer is not None:
     train_step_kwargs['summary_writer'] = sv.summary_writer
 
+  total_loss = 0
   should_retry = True
   while should_retry:
     try: