Washedashore/thepower

该数据集是一个多任务数据集，包含文本分类、问答、总结、文本生成、表格问答、标记分类、零样本分类、翻译、填空、深度估计、自动语音识别等多种任务。它支持多种语言，包括但不限于阿拉伯语、英语、汉语、法语、德语等。数据集的标签涵盖化学、生物学、医学、金融、艺术、代码和合成数据等。数据集的大小在100B到1T之间。不过，具体的数据集描述、来源、用途、结构、创建过程等详细信息未提供。

This dataset is a multi-task dataset that includes tasks such as text classification, question answering, summarization, text generation, table question answering, token classification, zero-shot classification, translation, fill-mask, depth estimation, and automatic speech recognition. It supports multiple languages including but not limited to Arabic, English, Chinese, French, German, etc. The dataset tags cover fields such as chemistry, biology, medicine, finance, art, code, and synthetic data. The size of the dataset is between 100B and 1T. However, detailed information about the datasets description, sources, uses, structure, creation process, etc. is not provided.