This dataset is a Swedish machine translation of the openeurollm/Dolci-Instruct-SFT-translated dataset, originally created as part of the OpenEuroLLM project.
Examples: 494,841 multi-turn conversations
Language: Swedish (sv-SE)
Format: Chat/messages format (id, messages)
License: Apache 2.0
All English source texts were machine-translated to Swedish using Google Gemma 3 27B-IT (w8a8_fp8… See the full description on the dataset page:
https://huggingface.co/datasets/AI-Sweden-Models/Dolci-Instruct-SFT-translated.
