saurabh1043/noisy_english

这是一个多模态数据集，包含文本和音频特征，适用于音频-文本相关任务（如语音识别或音频生成）。数据集分为训练集（train）和开发集（dev），其中训练集有271,018个示例，开发集有14,264个示例。总数据集大小约为143GB，下载大小约为133GB。每个示例包括id（标识符）、text（文本内容）和audio（音频数据）字段。

This is a multimodal dataset containing text and audio features, suitable for audio-text related tasks such as speech recognition or audio generation. The dataset is divided into a training set (train) and a development set (dev), with 271,018 examples in the training set and 14,264 examples in the development set. The total dataset size is approximately 143GB, with a download size of about 133GB. Each example includes fields for id (identifier), text (text content), and audio (audio data).