Upload README.md with huggingface_hub

README.md

@@ -0,0 +1,15 @@
+# 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.