sajadh/SAJADH-llama2-1k

该数据集包含多个特征，如消息ID、父消息ID、用户ID、创建日期、文本内容、角色、语言、评论次数、评论结果、是否删除、排名、是否为合成数据、模型名称、毒性分析（包括身份攻击、侮辱、淫秽、严重毒性、性暗示、威胁、毒性等）、消息树ID、树状态、表情符号（包括计数和名称）、标签（包括计数、名称和值）等。数据集分为训练集，包含85,443个样本，总大小为107,144,003字节。

The dataset includes multiple features such as message ID, parent message ID, user ID, creation date, text content, role, language, review count, review result, deletion status, rank, whether it is synthetic data, model name, toxicity analysis (including identity attack, insult, obscene, severe toxicity, sexual explicit, threat, toxicity), message tree ID, tree state, emojis (including count and name), labels (including count, name, and value), etc. The dataset is divided into a training set containing 85,443 samples with a total size of 107,144,003 bytes.