KNipun/

ai-humanizer

This dataset is designed for fine-tuning instruction-following LLMs to rewrite AI-generated text into more natural, human-like language.

train.jsonl – training split validation.jsonl – validation split

Each line is a JSON object: { "prompt": "Rewrite the following text to sound natural, human-like, and conversational...", "completion": "Humanized output text here", "attribution": "Original… See the full description on the dataset page: https://huggingface.co/datasets/KNipun/ai-humanizer.