Author: hf-mirror · License: unspecified (community upload, no license tag — default HF terms) · Format: parquet
0 downloads · 0 stars · 0 mirrors · 3,077,653,717 · 15 files
Curated R1 reasoning traces across math/code/science (open-r1's MoT recipe).
Tags: reasoning · cot · deepseek-r1 · huggingface
OpenR1's Mixture-of-Thoughts: curated DeepSeek-R1 reasoning traces across math, code, and science subsets, selected by verifiable-answer correctness for the R1 reproduction ablations. No license tag on the repo (default HF terms). Pins the 15-shard 'all' split (~3.1 GB).
Direct download: https://huggingface.co/datasets/open-r1/Mixture-of-Thoughts/resolve/main/all/train-00000-of-00015.parquet
Or via the API: curl -L -o file "https://huggingface.co/datasets/open-r1/Mixture-of-Thoughts/resolve/main/all/train-00000-of-00015.parquet"
Added 2026-09-27. Platform: Distillio — the open source training data collective.