datasets A superset of the highest quality datasets, with the highest quality versions, and active deduplication to train on the entire collection. kaiokendev/SuperCOT-dataset Viewer β’ Updated May 26, 2023 β’ 58.3k β’ 72 β’ 46 kaist-ai/CoT-Collection Viewer β’ Updated Oct 14, 2023 β’ 1.84M β’ 928 β’ 144 TIGER-Lab/MathInstruct Viewer β’ Updated May 15, 2024 β’ 262k β’ 1.37k β’ 280 Open-Orca/OpenOrca Viewer β’ Updated Feb 19 β’ 2.94M β’ 9.29k β’ 1.41k