search

WebNLG / GenerationEval

WebNLG+ Challenge 2020: Scripts to evaluate the RDF-to-text task with automatic metrics (BLEU, METEOR, chrF++, TER and BERT-Score)
MIT License
16 stars 5 forks source link

Wrong naming #2

Closed PierreColombo closed 3 years ago

PierreColombo commented 3 years ago

Hello, Thanks a lot for your work. Just a minor detail submissions/rdf2text/en/DANGNT-SGU/human is named human_eval_id_DANGNT-SGU.json instead of primary.json . Just for consistency for futur parsers :) Cheers

PierreColombo commented 3 years ago

Similar pb with 'submissions/rdf2text/en/Huawei Noahs Ark Lab/automatic/' name primary.en.txt instead of primary.en and ORANGE-NLG in submissions/rdf2text/en/ORANGE-NLG/automatic and Facebook Conv ai and OSU Neural NLG

Shimorina commented 3 years ago

We addressed your suggestions. Please see the new repo for the challenge results: https://github.com/WebNLG/challenge-2020