Inside TensorFlow: TF Debugging

In this episode of Inside TensorFlow, Software Engineer Shanqing Cai demonstrates to us TensorFlow Debugging for TF 2 and TF 1. Let us know what you think about this presentation in the comments below!

[leixun]

*My takeaways:*
1. Debugging tensor values 4:15
2. Debugging device placement 9:06
3. Debugging graph structures 12:08
4. Step debugging 22:35
5. Debugging high-level API (tf.keras) 25:26
6. Numerical issues (NaN/Infinity) 27:59
7. TensorFlow Debugger (tfdbg) 32:50

[ramzykaram]

Where is the Google Collab notebook he mentioned for the code ?

[redsun]

Very useful information. Would be handy to have a link to collab notebook, which is mentioned in the presentation.

[ramzykaram]

Is there any resources for the missing topics like Performance Bugs, Post-training model evaluation and Fairness issues ?

[jony]

22:58 ok and what do you do if the error disappears when you enable eager execution? what then?

[motbus]

content seems amazing, but cant hear nothing. could reupload with better audio ?

[nvsabhilash]

Waiting for tfdbg to roll out. Really awesome UI.