MultiSynt is an open multilingual synthetic dataset.
The MT Reasoning subset of MultiSynt is made of automatic translations into 2 languages of Glaive AI reasoning dataset containing 22mil+ general reasoning questions, reasoning traces and responses.
fra_Latn
17_354_716
1_802_885_115
25_224_272_259… See the full description on the dataset page:
https://huggingface.co/datasets/MultiSynt/MT-Reasoning.