Personal Dataset Preprocessing
Lobskodax opened this issue · comments
If I want to use my own dataset to train the gpt-2 model, the format is TXT, with one sentence per line, how can I modify the data preprocessing code to make it match and run normally.
Hi, did you figure out how?
Hi, I will close this issue for now. If you have difficulty build your own dataset, welcome to re-open this issue. Thanks~