Transformer Network Application: Named-Entity Recognition lab - Kernel keeps dying before training can finish

The last cell of code which trains the TFDistilBertForTokenClassification pre-trained model on new data keeps causing the kernel to die.

What to do?

Hi @davidaguilaratx,

Try getting a fresh copy of this notebook using the instructions provided in the “How to refresh your workspace” in the first week of the course. Use option 2.

Try again and let me know if it keeps occurring.


1 Like

Thanks, just needed to refresh the notebook.