# Generate train data for follow model | |
This repo will generate data of `Follow-Lang/set.mm.label` in huggingface. | |
## Format | |
- The data is located in datasets/train. | |
- Each line is formatted as: `s label arguments`. | |
- The maximum word length is 1024. | |
- All vocabulary words are listed in `words.txt` based on `Follow-Lang/set.mm.proof`. | |
- The data was generated with a depth of 2. | |
If you need additional data, feel free to reach out. | |
This version improves readability and flow while maintaining the original meaning. | |