prli/uspto_draft_bf_4-0_0-01_0-02_qwen_falcon_perturb
收藏Hugging Face2026-04-24 更新2026-04-26 收录
下载链接：
https://hf-mirror.com/datasets/prli/uspto_draft_bf_4-0_0-01_0-02_qwen_falcon_perturb
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是一个文本变换增强数据集，包含原始文本和经过六种不同变换方式处理后的文本。变换方式包括：同义词替换、butter_fingers变换(模拟打字错误)、随机删除部分文本、改变字符大小写、添加空格扰动以及使用下划线技巧。数据集仅包含验证集，共4000个文本样本。
This dataset is a text transformation augmentation dataset containing original texts and their versions processed by six different transformation methods. The transformations include: synonym substitution, butter_fingers (simulating typing errors), random deletion, changing character cases, whitespace perturbation, and underscore trick. The dataset only contains a validation set with 4,000 text samples.
提供机构：
prli
