continuedev/instinct-data

Instinct数据集包含用于训练和评估Continue公司最先进的下一个编辑模型的数据。数据集按照编程语言进行划分，TypeScript为原始语言，其他语言则是基于TypeScript数据合成的。数据集的特征包括内容和角色，均为字符串类型。数据集分为不同编程语言的训练集和测试集，并提供了每个数据集的示例数量和字节数。同时，还提供了数据集的下载大小和总大小。

The Instinct dataset contains data used for training and evaluating Continues state-of-the-art next-edit model. The dataset is split by programming languages, with TypeScript as the original language and others synthetically derived from the TypeScript data. The features of the dataset include content and role, which are both strings. The dataset is divided into training and testing sets for different programming languages, and the number of examples and size in bytes for each set is provided. The download size and the total size of the dataset are also mentioned.