Open crushonu1 opened 1 month ago
Hi, you can download the corresponding CSV format file from readme.md in the HC3 folder of the dataset, or directly access it this link. And in the generation of the corresponding MGT stage, a field of 'machine_text' will be added to the original dataset and saved as' updated_file. csv '
This is a very interesting method. I want to try running my own dataset using your method. I looked at your code, but I didn't find a CSV format dataset for HC3. I only found a JSON format dataset, so I'm not sure how to organize the dataset format. Could you please explain in detail how to organize the dataset?