Nick Sorros

The ability to add a different head would be a nice addition to the documentation actually

> Currently, we support any classification metric from the `evaluate` library: https://huggingface.co/docs/evaluate/index @lewtun one limitation though seems to be that in case of non binary labels there is no way...

Lately, I am leaning towards the idea of not randomness so that the results are reproducible. I know that this introduces a bias but it sounds minimal. This could be...

I don't think the sentence in italic translates to the sentence in bold. What is says is that document id is used a last resort for sorting. Document id is...

I am not sure how controversial this would be but it would definitely eliminate the need to merge tokens after as the algorithm would extract start and end for each...

> @ivyleavedtoadflax ok that makes sense re outputs. > > In terms of the instantiation of the model, is it not true that > > ``` > splitter_parser = SplitParser(config_file=MULTITASK_CFG)...

This problem is a bit annoying. I don't understand why would data miss from subsequent scrapes, that is a bit problematic. Whatever we decide it would be good to have...

Actually this is still relevant and something to explore.

that sounds super weird and important to resolve.

I think it is about time we do that and this tighs nicely with Matt's work on the splitter. How best to proceed on this? @jdu @SamDepardieu @ivyleavedtoadflax