refactor(dataset): 重构数据集处理流程并添加新功能

This commit is contained in:
2025-07-16 09:56:31 +08:00
parent 6edc31ce2e
commit b69ba2c5ff
8 changed files with 30 additions and 13 deletions
+10
View File
@@ -0,0 +1,10 @@
from datasets import load_dataset
from utils import pre_tarin_process
if __name__ == "__main__":
dataset = load_dataset("BelleGroup/train_3.5M_CN")
# pre_tarin_process(
# dataset_dict=dataset,
# output_subdir="belle_sft",
# max_chunk_size=5,
# )