drlee1/

glm-5.1-reasoning-ko-sample

A small test sample (10 items) translated into Korean from Jackrong/GLM-5.1-Reasoning-1M-Cleaned, which is itself a cleaned derivative of Kassadin88/GLM-5.1-1000000x (reasoning traces distilled from GLM-5.1). This sample exists as an early quality check before scaling translation to a larger subset.

Released under Apache 2.0, inheriting the license of both upstream sources. You must retain attribution… See the full description on the dataset page: https://huggingface.co/datasets/drlee1/glm-5.1-reasoning-ko-sample.