o0dimplz0o's picture
Update Readme.md
05e657c verified
metadata
dataset_info:
  features:
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: text
      dtype: string
  splits:
    - name: train
      num_bytes: 1117294900.769
      num_examples: 10227
  download_size: 1021497255
  dataset_size: 1117294900.769
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

Zeroth-STT-Korean Dataset

Description

This is a one of ten equally divided parts of the Zeroth-STT-Korean dataset. Ten equal shards of the original dataset were created and uploaded as individual repos for fine-tuning/training needs.

Citation

Zeroth-Korean Dataset, created by [Lucas Jo(@Atlas Guide Inc.) and Wonkyum Lee(@Gridspace Inc.)], 2023. Available at https://github.com/goodatlas/zeroth under CC-BY-4.0 license.

Junhoee/STT_Korean_Dataset_80000 Dataset, created by [Junhoee], 2024. Available at https://huggingface.co/datasets/Junhoee/STT_Korean_Dataset_80000