| language: | |
| - ko | |
| dataset_info: | |
| features: | |
| - name: audio | |
| dtype: | |
| audio: | |
| sampling_rate: 16000 | |
| - name: text | |
| dtype: string | |
| splits: | |
| - name: train | |
| num_bytes: 1113459385.076 | |
| num_examples: 10226 | |
| download_size: 1016618966 | |
| dataset_size: 1113459385.076 | |
| 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 |