Conrad Dobberstein

I think this PR is ready for review now. I have some more ideas to incorporate but they are beyond the basic functionality and may be added later within smaller...

Hey @helpmefindaname, thank you for your advice. These are the results I got on CONLL04 (same hyperparameters and training script as in the PR description): | Masks | Additional Special...

> @dobbersc could you do some training runs to evaluate the impact on the accuracy of whether `mask_remainder` is set to True or False? I don't have any concrete numbers...

Sorry that I only post now. I did some more masking experimentation similar to @helpmefindaname but nothing that got better scores. Here are some scores for the `mask_remainder` setting: |...

Hey @geheim01, 1. In your training script, you have used the same values for the `entity_pair_filters` in the `RelationExtractor` as in the example script: ```python # Step 4: Initialize relation...

> Ok Update: Now that I have once renamed all NER tags in our conll file from, for example, PER's to I-PER's/LOC to I-LOC, I am now not directly getting...