laion/

llama-nemotron-science-reasoning-on-canonical-think-full

The complete reasoning:on science split of nvidia/Llama-Nemotron-Post-Training-Dataset, converted once into the canonical Delphi chat-template thinking format. 708,920 rows. Unlike the cold-start warmup slice laion/llama-nemotron-science-reasoning-on-le3000tok-100k (and its -canonical-think variant), this build applies no length cap and no subsample — every long-CoT science example is… See the full description on the dataset page: https://huggingface.co/datasets/laion/llama-nemotron-science-reasoning-on-canonical-think-full.