• 📌 Dataset Description • 🔑 Supported Tasks • 📊 Languages • 📝 Data Fields • 📖 Citation
The Debate SFT Dataset is curated to support supervised fine-tuning (SFT) of large language models (LLMs) for generating logical, interpretable, and structured debate responses.A key motivation for this dataset is to address instruction refusal behaviors commonly observed in open-source models, which often result from… See the full description on the dataset page:
https://huggingface.co/datasets/adreammaker/InspireDebate.
