Dataset · nyu-mll
Sentence pairs for NLI
nyu-mll/multi_nli
433K sentence pairs with textual entailment annotations for training and evaluating NLI models.
What's inside
A crowdsourced corpus of 433K sentence pairs with logical entailment annotations. It covers various genres of spoken and written text and is intended, among other things, for cross-genre evaluation. The language is English.
What you can do with it
- Train an NLI classifier on sentence pairs
- Evaluate cross-genre generalization of an encoder model
- Use as a benchmark for entailment tasks
- Tasks
- Text classification
- Languages
- English
- Size
- 100K–1M rows
- Downloads this month
- 23.6K
- Likes
- 125
Models trained on it
- cross-encoder/nli-deberta-v3-smallZero shot classification 299.1K
- cross-encoder/nli-MiniLM2-L6-H768Zero shot classification 214K
- cross-encoder/nli-deberta-v3-baseZero shot classification 213.5K
- cross-encoder/nli-deberta-v3-largeZero shot classification 90.2K
- cross-encoder/nli-deberta-v3-xsmallZero shot classification 49.4K
- cross-encoder/nli-distilroberta-baseZero shot classification 19.4K
- jaredpalmer/kev-0.8bText classification 12.5K
- jaredpalmer/kev-4bText classification 10K
How to load it
from datasets import load_dataset
ds = load_dataset("nyu-mll/multi_nli")