Datasets:
| license: other | |
| task_categories: | |
| - automatic-speech-recognition | |
| - text-to-speech | |
| language: | |
| - ko | |
| # WTForbes/ko | |
| Corpus JSONL files are in `corpus_dir_jsonl/`. | |
| Audio payloads are stored in `audio_parquet/` shards. | |
| WAV source audio is converted to FLAC before Parquet writing. | |
| Only audio referenced by JSONL rows is included. | |