---
configs:
  - config_name: dz_ar
    data_files:
      - split: train
        path: "data/dz_ar/train-*.parquet"
  - config_name: eg_ar
    data_files:
      - split: train
        path: "data/eg_ar/train-*.parquet"
  - config_name: et_om
    data_files:
      - split: train
        path: "data/et_om/train-*.parquet"
  - config_name: et_ti
    data_files:
      - split: train
        path: "data/et_ti/train-*.parquet"
  - config_name: gm_en
    data_files:
      - split: train
        path: "data/gm_en/train-*.parquet"
  - config_name: gm_ff
    data_files:
      - split: train
        path: "data/gm_ff/train-*.parquet"
  - config_name: gm_mnk
    data_files:
      - split: train
        path: "data/gm_mnk/train-*.parquet"
  - config_name: gm_wo
    data_files:
      - split: train
        path: "data/gm_wo/train-*.parquet"
  - config_name: iq_ar
    data_files:
      - split: train
        path: "data/iq_ar/train-*.parquet"
  - config_name: jo_ar
    data_files:
      - split: train
        path: "data/jo_ar/train-*.parquet"
  - config_name: kh_km
    data_files:
      - split: train
        path: "data/kh_km/train-*.parquet"
  - config_name: kw_ar
    data_files:
      - split: train
        path: "data/kw_ar/train-*.parquet"
  - config_name: ln_voa
    data_files:
      - split: train
        path: "data/ln_voa/train-*.parquet"
      - split: test
        path: "data/ln_voa/test-*.parquet"
  - config_name: ma_ar
    data_files:
      - split: train
        path: "data/ma_ar/train-*.parquet"
  - config_name: ma_ar_lower
    data_files:
      - split: train
        path: "data/ma_ar_lower/train-*.parquet"
  - config_name: ma_rif
    data_files:
      - split: train
        path: "data/ma_rif/train-*.parquet"
  - config_name: ma_shi
    data_files:
      - split: train
        path: "data/ma_shi/train-*.parquet"
  - config_name: ma_zgh
    data_files:
      - split: train
        path: "data/ma_zgh/train-*.parquet"
  - config_name: ml_bm
    data_files:
      - split: train
        path: "data/ml_bm/train-*.parquet"
      - split: test
        path: "data/ml_bm/test-*.parquet"
  - config_name: mm_my
    data_files:
      - split: train
        path: "data/mm_my/train-*.parquet"
  - config_name: ng_ha
    data_files:
      - split: train
        path: "data/ng_ha/train-*.parquet"
  - config_name: sn_ff
    data_files:
      - split: train
        path: "data/sn_ff/train-*.parquet"
  - config_name: so_so
    data_files:
      - split: train
        path: "data/so_so/train-*.parquet"
  - config_name: ti_voa
    data_files:
      - split: train
        path: "data/ti_voa/train-*.parquet"
      - split: test
        path: "data/ti_voa/test-*.parquet"
  - config_name: tj_tg
    data_files:
      - split: train
        path: "data/tj_tg/train-*.parquet"
  - config_name: tn_ar
    data_files:
      - split: train
        path: "data/tn_ar/train-*.parquet"
license: other
task_categories:
  - automatic-speech-recognition
language:
  - tg
---

# WS2-audio — Unlabeled Speech Segments

Audio-only dataset of VAD-segmented speech clips. No transcripts.
See AUDIO_DATASET_SCHEMA.md in the WorldSpeech repo for the full spec.