6 models · ranked by HuggingFace downloads
BART-large fine-tuned on MultiNLI for zero-shot text classification via natural language inference. Given a text and a candidate label, it predicts entailment probability to classify without task-specific training data.
nli-deberta-v3-base is an open-source zero-shot-classification model available on HuggingFace. Details are sourced from the public model registry.
mDeBERTa-v3-base-xnli-multilingual-nli-2mil7 is an open-source zero-shot-classification model available on HuggingFace. Details are sourced from the public model registry.
DeBERTa-v3-small fine-tuned on MNLI+SNLI for natural language inference and zero-shot text classification, from the sentence-transformers team. The small DeBERTa-v3 backbone outperforms BERT-base on NLI while being much faster than DeBERTa-v3-large.
distilbert-base-uncased-mnli is an open-source zero-shot-classification model available on HuggingFace. Details are sourced from the public model registry.
nli-MiniLM2-L6-H768 is an open-source zero-shot-classification model available on HuggingFace. Details are sourced from the public model registry.