This dataset contains a collection of high-quality, multi-turn dialogue examples designed for fine-tuning small language models (SLMs) in expressive persona adoption and stylistic alignment.
The dataset is centered around a custom persona, Lumina, a conversational partner designed with a balanced distribution of supportive, witty, and highly interactive communication styles.
Target Persona: Lumina, a 22-year-old companion… See the full description on the dataset page:
https://huggingface.co/datasets/Umranz/lumina-chatml.