refactor(pretrain): 更新预训练数据集输出目录名称
This commit is contained in:
@@ -36,6 +36,6 @@ if __name__ == "__main__":
|
||||
|
||||
process_dataset(
|
||||
dataset_dict=DatasetDict({"train": combined_dataset}),
|
||||
output_subdir="chinese-instruct",
|
||||
output_subdir="chinese-instruct-sft",
|
||||
process_func=process_func,
|
||||
)
|
||||
@@ -15,7 +15,7 @@ if __name__ == "__main__":
|
||||
|
||||
process_dataset(
|
||||
dataset_dict=dataset,
|
||||
output_subdir="Ling-Coder-SFT",
|
||||
output_subdir="Ling-Coder-sft",
|
||||
process_func=process_func
|
||||
)
|
||||
|
||||
@@ -24,7 +24,7 @@ if __name__ == "__main__":
|
||||
dataset = load_dataset("HuggingFaceTB/Magpie-Pro-300K-Filtered-H4")
|
||||
process_dataset(
|
||||
dataset_dict=dataset,
|
||||
output_subdir="belle-sft",
|
||||
output_subdir="Magpie-Pro-300K-sft",
|
||||
process_func=process_func,
|
||||
split_name="train-sft"
|
||||
split_name="train_sft",
|
||||
)
|
||||
Reference in New Issue
Block a user