armingh2000/FactScoreLite
FactScoreLite is an implementation of the FactScore metric, designed for detailed accuracy assessment in text generation. This package builds upon the framework provided by the original FactScore repository, which is no longer maintained and contains outdated functions.
This tool helps you evaluate the factual accuracy of text generated by AI models. You provide the AI-generated text and its original knowledge source, and it breaks down the text into individual facts, then checks each fact against the source. It's designed for anyone who needs to ensure their AI-generated content is truthful and consistent with reference materials.
No commits in the last 6 months. Available on PyPI.
Use this if you need a reliable and detailed way to assess the factual correctness of AI-generated text, particularly when you have a specific knowledge source to compare against.
Not ideal if you don't have a clear knowledge source for the AI-generated text or if you are looking for subjective quality assessments beyond factual accuracy.
Stars
13
Forks
2
Language
Python
License
MIT
Category
Last pushed
Apr 25, 2024
Commits (30d)
0
Dependencies
5
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/nlp/armingh2000/FactScoreLite"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
openfactcheck-research/openfactcheck
An Open-source Factuality Evaluation Demo for LLMs
several27/FakeNewsCorpus
A dataset of millions of news articles scraped from a curated list of data sources.
lilakk/BooookScore
A package to generate summaries of long-form text and evaluate the coherence of these summaries....
Cartus/Automated-Fact-Checking-Resources
Links to conference/journal publications in automated fact-checking (resources for the...
manideep2510/siamese-BERT-fake-news-detection-LIAR
Triple Branch BERT Siamese Network for fake news classification on LIAR-PLUS dataset in PyTorch