Datasetpublic

skupina-7/nlp-paraphrases-20k-pruned

The dataset contains 20703 records. The dataset was created by removing all dataset items from the original 27k dataset that had a BLEU score 0 or more than 0.3388.

sourceHugging Faceupdated 3y agoView on Hugging Face

0likes7downloads

root

file.gitattributes2 KBdownload

fileREADME.md295 Bdownload

filetest-0.json3.5 MBdownload

filetest-1.json3.6 MBdownload

filetest-2.json3.5 MBdownload

filetest-3.json3.6 MBdownload

filetest-4.json3.5 MBdownload

filetest-5.json3.5 MBdownload

filetest-6.json3.5 MBdownload

filetest-7.json1.8 MBdownload

filetrain-0.json3.6 MBdownload

filetrain-1.json3.7 MBdownload

filetrain-10.json3.6 MBdownload

filetrain-11.json3.6 MBdownload

filetrain-12.json3.6 MBdownload

filetrain-13.json3.5 MBdownload

filetrain-14.json3.6 MBdownload

filetrain-15.json3.7 MBdownload

filetrain-16.json3.6 MBdownload

filetrain-17.json3.7 MBdownload

filetrain-18.json3.6 MBdownload

filetrain-19.json3.6 MBdownload

filetrain-2.json3.6 MBdownload

filetrain-20.json3.6 MBdownload

filetrain-21.json3.5 MBdownload

filetrain-22.json1.8 MBdownload

filetrain-3.json3.5 MBdownload

filetrain-4.json3.5 MBdownload

filetrain-5.json3.6 MBdownload

filetrain-6.json3.6 MBdownload

filetrain-7.json3.7 MBdownload

filetrain-8.json3.5 MBdownload

filetrain-9.json3.5 MBdownload

skupina-7/nlp-paraphrases-20k-pruned · main · files are served by the source, never re-hosted here