Norwegian Wordnet Bokmål. Models

[sigdelina]

This issue provides information about the baseline and models that have been tested on the data of the Wordnet from the Norwegian National Library.

Models that have already been tested for provided data:

1. Most frequent sense baseline (with the using of Norbert2)

2. A number of tests with different input data on Norbert2, Multilingual Bert (12 models are evaluated by the moment)

Other tests are in the process of implementation.

[sigdelina]

The first version of documentation is uploaded to the gilthub

The first version of the documentation can be found by the link

What was done:

1. the POS-tagging and Sentiment analysis files were optimised.
2. all models for POS-tagging task and sentiment analysis were ran once more to check the results are relevant. The files with the results for this can be found here. The summary is here
3. The merged script for bert and roberta was run on Saga (but failed in some places).

What have to be done by 22.08:

1. The codes for POS-tagging and Sentiment analysis should be updated for distilbert as well
2. The scripts for NER task for bert and roberta should be merged (and not fail)

[akutuzov]

I believe this was for #4 ?