Matej Gjurkovic

Happy to announce the first public release of Podium: github.com/TakeLab/podium, a framework-agnostic Python #NLProc library for data loading & preprocessing developed at #TakeLab. Install via `pip install podium-nlp`, docs available at: takelab.fer.hr/podium/

@yoavgo By starting acknowledging the efforts that go into creating those resources while educating reviewers that there's more to the scientific process besides fancy models that beat SOTA for 0.5%. We got rejected for ACL by a tiny margin due to the lack of methodological novelty (?).

my main takeaway from conversations at ACL: our datasets are crap. I am not talking about MRC, which goes without saying but also improving. I'm talking about real-world stuff: NER, IE, document classification. What we have is toyish and unrealistic. How do we get good ones?