Attempts to suppress Warnings produced by Tensorflow 2 fail

Viewed 629

When I do inference (i.e. predictions) using a Keras model that is based on Tensorflow 2 and uses Tensorflow Probability as well I am getting voluminous Warnings of the type:

WARNING:tensorflow:Importing a function (__inference_adj_close_min_10_q10_layer_call_and_return_conditional_losses_45193) with ops with custom gradients. Will likely fail if a gradient is requested.

I tried to suppress these Warnings by running this command -- but had no effect:

os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'

Likewise with the command:

import warnings
warnings.filterwarnings('ignore')

What you would advise me to do?

0 Answers
Related