yucheng4/

tts-scratch

	loader.py




	


	Dataset Summary

A astronomy dataset with sensor fusion modality, stored in parquet format.

Preprocessing: curriculum Augmentation: heavy

Split strategy: leave one out Sampling: weighted

Quality filtering: lenient Labeling: pseudo label

loader.py — main artifact of this repository

See the… See the full description on the dataset page: https://huggingface.co/datasets/yucheng4/tts-scratch.