DeBERTa-v3-large -- Hallucination Detection

Fine-tuned for hallucination detection as part of Director-AI.

Training

  • Base: MoritzLaurer/DeBERTa-v3-large-mnli-fever-anli-ling-wanli
  • Data: ~100K examples from HaluEval, FEVER, VitaminC, ANLI R3
  • Epochs: 3, lr 2e-5, batch 32 (effective), class-weighted CE loss
  • Labels: 0 = entailment, 1 = neutral, 2 = contradiction

Usage

from director_ai.core import NLIScorer

scorer = NLIScorer(model_name="anulum/deberta-v3-large-hallucination")
score = scorer.score("The capital of France is Paris.", "Paris is in Germany.")

License

AGPL-3.0 | Commercial licensing: anulum.li

Downloads last month
17
Safetensors
Model size
0.4B params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for anulum/deberta-v3-large-hallucination

Datasets used to train anulum/deberta-v3-large-hallucination