snli, snli

  • תיאור :

קורפוס SNLI (גרסה 1.0) הוא אוסף של 570,000 צמדי משפטים באנגלית שנכתבו על ידי אדם, מתויגים באופן ידני לסיווג מאוזן עם התוויות הכרחי, סתירה וניטרלית, התומכים במשימה של מסקנות שפה טבעית (NLI), הידועה גם כזיהוי השפעות טקסטואלית. (RTE).

לְפַצֵל דוגמאות
'test' 10,000
'train' 550,152
'validation' 10,000
  • מבנה תכונה :
FeaturesDict({
    'hypothesis': Text(shape=(), dtype=string),
    'label': ClassLabel(shape=(), dtype=int64, num_classes=3),
    'premise': Text(shape=(), dtype=string),
})
  • תיעוד תכונה :
תכונה מעמד צוּרָה Dtype תיאור
FeaturesDict
הַשׁעָרָה טֶקסט חוּט
תווית ClassLabel int64
הַנָחַת יְסוֹד טֶקסט חוּט
  • ציטוט :
@inproceedings{snli:emnlp2015,
    Author = {Bowman, Samuel R. and Angeli, Gabor and Potts, Christopher, and Manning, Christopher D.},
    Booktitle = {Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing (EMNLP)},
    Publisher = {Association for Computational Linguistics},
    Title = {A large annotated corpus for learning natural language inference},
    Year = {2015}
}