Ki-Seki/llm-failure-dataset

LLM Failure Dataset是一个支持多种任务类别的数据集，包括文本分类、表格问答、问答、零样本分类、翻译、摘要、文本生成、文本到文本生成和填充掩码。数据集支持中文和英文，包含查询、模型响应、模型、用户响应、用户和时间戳等特征。数据集分为训练集，包含2个示例，总大小为3204字节。

This is a dataset for analyzing failure cases of large language models (LLM), covering various natural language processing tasks including text classification, table question answering, question answering, zero-shot classification, translation, summarization, text generation, text-to-text generation, and fill-mask. The dataset supports both Chinese and English languages and includes features such as query, model response, model name, user response, user, and timestamp. The dataset is divided into a training set with 2 samples.