train_on_batch() outputs progress bar

[aedavids]

I am training a trival GAN using train_on_batch() it does not have a verbose argument. it produced a mount of out put of the form

$ head *.out
2/2 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step 
2/2 ━━━━━━━━━━━━━━━━━━━━ 0s 916us/step
2/2 ━━━━━━━━━━━━━━━━━━━━ 0s 771us/step
2/2 ━━━━━━━━━━━━━━━━━━━━ 0s 878us/step
2/2 ━━━━━━━━━━━━━━━━━━━━ 0s 731us/step
2/2 ━━━━━━━━━━━━━━━━━━━━ 0s 836us/step
2/2 ━━━━━━━━━━━━━━━━━━━━ 0s 770us/step
2/2 ━━━━━━━━━━━━━━━━━━━━ 0s 726us/step
2/2 ━━━━━━━━━━━━━━━━━━━━ 0s 724us/step
2/2 ━━━━━━━━━━━━━━━━━━━━ 0s 867us/step

setting the following did not resolve the issue

tf.get_logger().setLevel('ERROR')
tf.autograph.set_verbosity(0)
tf.config.set_soft_device_placement(False)

Python 3.11.10
$
conda list tensorflow
tensorflow 2.17.0 cpu_py311h6ac8430_2 conda-forge
tensorflow-base 2.17.0 cpu_py311h0e3e7eb_2 conda-forge
tensorflow-estimator 2.17.0 cpu_py311h331965b_2 conda-forge

$ conda list keras

Name Version Build Channel

keras 3.4.1 pyhd8ed1ab_2 conda-forge

please see attached test script

Kind regards

Andy

08_complete_example.py.txt

[Grvzard]

line 68:

X = generator.predict(x_input, verbose=0)