metadata
dataset_info:
features:
- name: text
dtype: string
splits:
- name: train
num_bytes: 31705238
num_examples: 200000
download_size: 197019
dataset_size: 31705238
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
license: mit
language: - en license: mit task_categories: - text-generation - text-classification tags: - fact-checking - source-verification - anti-hallucination - citation - reality-check - reasoning size_categories: - 100K<n<1M
Source Check Dataset
Description
A dataset for training language models to cite sources, verify facts, check reliability, and reduce hallucinations.
Content
- 200,000 examples
- 5 categories: source reliability, citations, anti-hallucination, reality check, explanation
- English language
Categories
| Category | Description |
|---|---|
| Source Reliability | Model evaluates the reliability of sources |
| Citations | Model cites its sources |
| Anti-Hallucination | Model avoids making things up |
| Reality Check | Model verifies facts and myths |
| Explanation | Model explains and verifies |
Usage
from datasets import load_dataset
dataset = load_dataset("Lelonthecodeur/source-check-dataset")