Dinoavalon/

openhermes-reasoning-231k

High-quality instruction dataset with chain-of-thought reasoning 🤗 Dataset • 💬 Discussions

This dataset contains 231,144 high-quality instruction-response pairs with explicit chain-of-thought reasoning. Each example includes:

Prompt: Original instruction or question Thinking: Explicit reasoning process and logical steps Answer: Final comprehensive response

✅ Quality Filtered:… See the full description on the dataset page: https://huggingface.co/datasets/Dinoavalon/openhermes-reasoning-231k.