jiahaowangdale/

roberta-classifier

	preprocess.py




	


	Dataset Summary

A nlp qa dataset with video text modality, stored in jsonl format.

Preprocessing: adaptive Augmentation: heavy

Split strategy: kfold 5 Sampling: random

Quality filtering: adaptive Labeling: pseudo label

preprocess.py — main artifact of this repository

See the license field… See the full description on the dataset page: https://huggingface.co/datasets/jiahaowangdale/roberta-classifier.