This cross-encoder model is fine-tuned on NLI datasets (MNLI, SNLI) using the DeBERTa-v3-base architecture. It predicts whether a premise entails, contradicts, or is neutral to a hypothesis. It is commonly used for zero-shot classification, semantic textual similarity, and as a component in retrieval or re-ranking systems via the sentence-transformers library.
Nli Deberta V3 Base sits in PulseGate's Other AI category. It focuses on determining semantic relationships (entailment, contradiction, neutral) between sentence pairs without task-specific training data. It is built as an open-source project for developers. Nli Deberta V3 Base is open source under the Open Source license. It runs on the web and API.
Behind Nli Deberta V3 Base is sentence-transformers, and it first shipped in 2022.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do