refactor: 重构数据管道模块
This commit is contained in:
@@ -1,9 +1,11 @@
|
||||
from datasets import load_dataset
|
||||
from modules.utils import process_dataset
|
||||
from modules.datapipeline import DataPipeline
|
||||
|
||||
if __name__ == "__main__":
|
||||
dataset = load_dataset("shjwudp/chinese-c4")
|
||||
process_dataset(
|
||||
|
||||
pipeline = DataPipeline()
|
||||
pipeline.process_dataset(
|
||||
dataset_dict=dataset,
|
||||
output_subdir="chinese-c4-pretrain"
|
||||
)
|
||||
Reference in New Issue
Block a user