Our dataset ranked #1 on the Hugging Face Datasets Trending leaderboard from January 20 to January 30.
The Superior-Reasoning-SFT-gpt-oss-120b dataset is a high-quality, open-source collection containing 435K samples designed to democratize the training of high-performance Long Chain-of-Thought (Long-CoT) models. Unlike standard distilled datasets that rely on random sampling or… See the full description on the dataset page:
https://huggingface.co/datasets/zoelaenskii/zinnety.
