This dataset contains Direct Preference Optimization (DPO) pairs translated and culturally adapted into Swedish from the original argilla/ultrafeedback-binarized-preferences-cleaned dataset. The dataset contains the following fields: - prompt: The culturally adapted Swedish prompt. - chosen: The high-quality, authentic Swedish response. - rejected: The poorly translated, uncalibrated response. - language: The target language of the row. This dataset inherits the license from its source dataset argilla/ultrafeedback-binarized-preferences-cleaned (MIT).
Publicly accessible
mit
This dataset contains Direct Preference Optimization (DPO) pairs translated and culturally adapted into Icelandic from the original argilla/ultrafeedback-binarized-preferences-cleaned dataset. The dataset contains the following fields: - prompt: The culturally adapted Icelandic prompt. - chosen: The high-quality, authentic Icelandic response. - rejected: The poorly translated, uncalibrated response. - language: The target language of the row. This dataset inherits the license from its source dataset argilla/ultrafeedback-binarized-preferences-cleaned (MIT).
Publicly accessible
mit
This dataset contains Direct Preference Optimization (DPO) pairs translated and culturally adapted into Finnish from an English source dataset (e.g. ultrafeedback or safe-rlhf). The dataset contains the following fields: - prompt: The culturally adapted Finnish prompt. - chosen: The high-quality, authentic Finnish response. - rejected: The poorly translated, uncalibrated response. - targetlanguage: The target language of the row. - datasetsource: The source dataset name. - issafetyfocused: Boolean indicating if this is a safety alignment sample. This dataset inherits the license from its source dataset (MIT or Apache 2.0).
Publicly accessible
mit
This dataset contains Direct Preference Optimization (DPO) pairs translated and culturally adapted into German from an English source dataset (e.g. ultrafeedback or safe-rlhf). The dataset contains the following fields: - prompt: The culturally adapted German prompt. - chosen: The high-quality, authentic German response. - rejected: The poorly translated, uncalibrated response. - targetlanguage: The target language of the row. - datasetsource: The source dataset name. - issafetyfocused: Boolean indicating if this is a safety alignment sample. This dataset inherits the license from its source dataset (MIT or Apache 2.0).
Publicly accessible
mit
This dataset contains Direct Preference Optimization (DPO) pairs translated and culturally adapted into French from an English source dataset (e.g. ultrafeedback or safe-rlhf). The dataset contains the following fields: - prompt: The culturally adapted French prompt. - chosen: The high-quality, authentic French response. - rejected: The poorly translated, uncalibrated response. - targetlanguage: The target language of the row. - datasetsource: The source dataset name. - issafetyfocused: Boolean indicating if this is a safety alignment sample. This dataset inherits the license from its source dataset (MIT or Apache 2.0).
Publicly accessible
mit
This dataset contains Direct Preference Optimization (DPO) pairs translated and culturally adapted into Norwegian from an English source dataset (e.g. ultrafeedback or safe-rlhf). The dataset contains the following fields: - prompt: The culturally adapted Norwegian prompt. - chosen: The high-quality, authentic Norwegian response. - rejected: The poorly translated, uncalibrated response. - targetlanguage: The target language of the row. - datasetsource: The source dataset name. - issafetyfocused: Boolean indicating if this is a safety alignment sample. This dataset inherits the license from its source dataset (MIT or Apache 2.0).
Publicly accessible
mit
This dataset contains Direct Preference Optimization (DPO) pairs translated and culturally adapted into Polish from an English source dataset (e.g. ultrafeedback or safe-rlhf). The dataset contains the following fields: - prompt: The culturally adapted Polish prompt. - chosen: The high-quality, authentic Polish response. - rejected: The poorly translated, uncalibrated response. - targetlanguage: The target language of the row. - datasetsource: The source dataset name. - issafetyfocused: Boolean indicating if this is a safety alignment sample. This dataset inherits the license from its source dataset (MIT or Apache 2.0).
Publicly accessible
mit
This dataset contains Direct Preference Optimization (DPO) pairs translated and culturally adapted into Spanish from an English source dataset (e.g. ultrafeedback or safe-rlhf). The dataset contains the following fields: - prompt: The culturally adapted Spanish prompt. - chosen: The high-quality, authentic Spanish response. - rejected: The poorly translated, uncalibrated response. - targetlanguage: The target language of the row. - datasetsource: The source dataset name. - issafetyfocused: Boolean indicating if this is a safety alignment sample. This dataset inherits the license from its source dataset (MIT or Apache 2.0).
Publicly accessible
mit
