← All notes

Short answer

my_dataset/
  metadata.csv   # file_name,transcription
  data/first.wav

Claims

• id: C1 claim: AudioFolder is a no-code way to create an audio dataset. source: S1 evidence: "Create an audio dataset repository with the AudioFolder builder."
• id: C2 claim: Datasets can also be uploaded with push_to_hub. source: S1 evidence: "Then upload the dataset to the Hugging Face Hub using Dataset.push_to_hub():"
• id: C3 claim: metadata.csv needs a file_name column linking audio files to metadata. source: S2 evidence: "Your metadata.csv file must have a file_name column"

Sources

• [S1] Create an audio dataset · Hugging Face. Hugging Face (datasets docs). https://huggingface.co/docs/datasets/audio_dataset . License: Apache-2.0 (huggingface/datasets docs). Accessed 2026-11-04.
• [S2] Load audio data · Hugging Face. Hugging Face (datasets docs). https://huggingface.co/docs/datasets/audio_load . License: Apache-2.0 (huggingface/datasets docs). Accessed 2026-11-04.

License: this note is CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Summary by iinventedlight (AI-assisted).

@misc{iil_note_ahd05aqf,
  author = {{iinventedlight}},
  title = {{Create an audio dataset with AudioFolder}},
  howpublished = {iinventedlight Library},
  year = {2026},
  date = {2026-11-04},
  url = {https://iinventedlight.com/library/notes/hf-datasets-audiofolder/},
  note = {Permanent ID iil:note:ahd05aqf. License: CC-BY-4.0.},
}

Markdown

Spotted a mistake? Use the MCP feedback tool or email hello@iinventedlight.com.

For language models and agents: start at /llms.txt. This page in Markdown: /library/notes/hf-datasets-audiofolder.md.