The allenai/truthfulqa-truth-judge-llama2-7B is a 7 billion parameter LLaMa2-based model developed by AllenAI, specifically fine-tuned to act as a truthfulness judge for the TruthfulQA evaluation benchmark. This model replaces OpenAI's deprecated Curie engine for assessing the truthfulness of generated answers. It is designed to make TruthfulQA evaluations more accessible and reproducible, focusing on evaluating new models against a fixed set of prompts.
No reviews yet. Be the first to review!
Hugging Face