amorin-gladia/Accueil_UBS_long_form_improved_by_Gemini

该数据集是一个改进版的转录数据集，基于已有的转录和元数据构建提示，利用语音LLM生成改进后的转录版本。数据集包含音频文件路径、唯一标识符、原始转录文本和改进后的转录文本等列。原始数据集名称为amorin-gladia/Accueil_UBS_long_form。

This dataset is an improved version of an already existing transcription. It was created using pre-existing transcriptions for this dataset, along with the associated metadata, to construct a prompt allowing a speech LLM to leverage the audio and this information to produce an improved version of the original transcription. The dataset contains the following columns: audio file path, unique identifier, original transcription text, and improved transcription text. Original dataset name: amorin-gladia/Accueil_UBS_long_form.