Is The Token Classification Notebook Outdated?

Hi :hugs:
I was trying to run the Token Classification Example Notebook and thought it might be outdated.

The signs I have found are as follows:

  1. There are some missing dependencies, evaluate, seqeval and accelerate which need to be installed.

  2. The training loss is not being logged while training the model, and the following warning is shown: “/usr/local/lib/python3.10/dist-packages/seqeval/metrics/ UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.”

Is there anything else that needs to be updated in order to run this notebook as intended? Have there been any breaking changes that could have impacted this notebook?


1 Like