Way to take split from the HF dataset

[dipta007]

Is there any way to take the train or val split directly from the HF dataset?

Something like this:

--dataset hf_data:train
--val_dataset hf_dataset:val