tingtang2/the_stack_v2_python_repos_pretraining_dataset_repo_reconstruction-dataset
收藏Hugging Face2025-09-20 更新2025-10-25 收录
下载链接：
https://hf-mirror.com/datasets/tingtang2/the_stack_v2_python_repos_pretraining_dataset_repo_reconstruction-dataset
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含三个字段：仓库名称(repo_name)，合并内容(combined_content)，和文件路径列表(file_paths)。仓库名称和合并内容为文本形式，文件路径列表则记录了相关文件的路径。数据集的训练集大小为17855275374字节，共有400000个训练样本。数据集的下载大小为5696432796字节。
The dataset includes three fields: repository name (repo_name), combined content (combined_content), and a list of file paths (file_paths). The repository name and combined content are in text form, while the file path list records the paths of related files. The training set of the dataset is 17855275374 bytes in size and contains 400,000 training samples. The download size of the dataset is 5696432796 bytes.
提供机构：
tingtang2
