南开大学计算机学院自然语言处理课程大作业，含Python实现的代码及数据集

This folder contains the datasets used for X-Class. Due to size constraints, we uploaded the data to google drive, [here](https://drive.google.com/drive/folders/1w0g3c0z9eoV-IYHCcA54tBKiNTYJy-3J?usp=sharing) is the download link. After download, you can unzip the zipped dataset through `unzip -o`. ## Data format We also describe the dataset format for potential use of new datasets. All files should be placed in a folder with the dataset's name, in this directory. The files to include are - dataset.txt - A text file containing documents, one per line. We will use BERT's tokenizer for tokenization. - classes.txt - A text file containing the class names, one per line. - labels.txt - A text file containing the class (index) of each document in `dataset.txt`, one label per line. All the files should have the exact same names.