Non Deterministic Behaviour from Deeplabv3+

Prerequisites

Please answer the following questions for yourself before submitting an issue.

• I am using the latest TensorFlow 1.14.
• I am reporting the issue to the correct repository. (Model Garden official or research directory)
• I checked to make sure that this issue has not been filed already.

1. The entire URL of the file you are using

https://github.com/tensorflow/models/tree/master/official/models/research/deeplab/

2. Describe the bug

I am currently battling non determinism , and the way it has to be solved is to :

Set tf.set_random_seed(seed=0) at a global level
Set seed to individual initializers such as tf.truncated_normal(std,seed=0)

I am currently battling non determinism , and the way it has to be solved is to :

Set tf.set_random_seed(seed=0) at a global level
Set seed to individual initializers such as tf.truncated_normal(std,seed=0)
Although Since i am working with Deeplab3+ from Tensorflow Research , i found that it uses slim.conv2D extensively and i am having a hard time to track it down and set individual seed at its layer. Is there a simple way we can make this deterministic?