DewiBrynJones/evals-speech-recognition-cy-en-2512

该数据集是一个多配置的语音和文本数据集，包含威尔士语和英语的语音转录和翻译任务。数据集包含五个子集：1)cym_en_arfor__lleisiau_arfor：包含威尔士语语音数据，带有口音信息；2)techiaith__banc_trawsgrifiadau_bangor：威尔士语转录数据集；3)techiaith__commonvoice_23_0_cy：CommonVoice项目的威尔士语语音数据集；4)techiaith__commonvoice_23_0_cy_en：威尔士语和英语的双语数据集；5)techiaith__commonvoice_23_0_en__GB_IE：英国和爱尔兰英语的语音数据集。每个子集都包含句子、ID、任务类型和预测结果等特征。数据集主要用于语音识别和机器翻译任务。

The dataset is a multi-configuration speech and text dataset containing Welsh and English speech transcription and translation tasks. It includes five subsets: 1) cym_en_arfor__lleisiau_arfor: Welsh speech data with accent information; 2) techiaith__banc_trawsgrifiadau_bangor: Welsh transcription dataset; 3) techiaith__commonvoice_23_0_cy: Welsh speech dataset from the CommonVoice project; 4) techiaith__commonvoice_23_0_cy_en: Bilingual dataset in Welsh and English; 5) techiaith__commonvoice_23_0_en__GB_IE: Speech dataset in British and Irish English. Each subset contains features such as sentences, IDs, task types, and prediction results. The dataset is primarily used for speech recognition and machine translation tasks.