KrzTyb/fim-dataset

fim-dataset是一个用于训练代码自动补全模型的代码片段数据集，采用Fill-in-the-Middle (FIM)方法。数据集包含格式化的代码片段，其中包括前缀、后缀和需要完成的代码部分，使用特殊令牌进行标识。数据集分为训练集和验证集，分别包含42,922和2,259个示例。The fim-dataset is a dataset of code snippets for training code autocompletion models using the Fill-in-the-Middle (FIM) approach. It contains formatted code snippets with a prefix, suffix, and the part of the code to be completed, identified by special tokens. The dataset is split into a training set and a validation set, containing 42,922 and 2,259 examples respectively.