apsora/finetuning_text_model

YouTube Music Comments (Original & Augmented)

Summary: User comments from YouTube music videos labeled with one of 7 genres
(pop, Jazz, rock, electronic, R&B, Metal, Classical). Two splits are provided: original (human-labeled) and augmented (synthetic additions), both with columns:

• text (string)
• label (string)

Load

from datasets import load_dataset
ds = load_dataset("Iris314/Youtube_music_comments")
original  = ds["original"]
augmented = ds["augmented"]

@misc{youtube_music_comments_2025,
  title        = {YouTube Music Comments (Original & Augmented)},
  author       = {Iris314},
  year         = {2025},
  howpublished = {\url{https://huggingface.co/datasets/Iris314/Youtube_music_comments}}
}