JS10k

该数据集是一个JavaScript数据集，包含了训练数据中最常出现的10,000个值。此外，它还包含了扁平化的AST节点序列，用于预测特定AST节点的值或类型。该数据集的规模达到了150,000个程序文件，其任务是代码补全。

This is a JavaScript dataset containing the 10,000 most frequently occurring values in the training data. Additionally, it includes flattened AST node sequences for predicting the values or types of specific AST nodes. The dataset consists of 150,000 program files, and its targeted task is code completion.