diff --git a/scripts/pre_train/chinese-cosmopedia.py b/scripts/pre_train/chinese-cosmopedia.py
index 8a1e769..01f64b3 100644
--- a/scripts/pre_train/chinese-cosmopedia.py
+++ b/scripts/pre_train/chinese-cosmopedia.py
@@ -4,7 +4,6 @@ from pipeline import export_dataset
 if __name__ == "__main__":
     dataset = load_dataset(
         "opencsg/chinese-cosmopedia",
-        data_files={"train": [f"data/000{i:02d}.parquet" for i in range(25)]},
     )
     export_dataset(
         dataset=dataset["train"],
diff --git a/scripts/pre_train/english-dolma.py b/scripts/pre_train/english-dolma.py
index 07ef4c7..7e82b3c 100644
--- a/scripts/pre_train/english-dolma.py
+++ b/scripts/pre_train/english-dolma.py
@@ -7,5 +7,4 @@ if __name__ == "__main__":
         dataset=dataset["train"],
         output_dir="./dataset",
         output_prefix="english-dolma-30b-pretrain",
-        max_chunks=18,
     )