bambara-audio / README.md
sudoping01's picture
Upload dataset
88e6811 verified
metadata
dataset_info:
  - config_name: bm-fr-en
    features:
      - name: audio
        dtype:
          audio:
            sampling_rate: 16000
      - name: bambara
        dtype: string
      - name: french
        dtype: string
      - name: duration
        dtype: float64
      - name: english
        dtype: string
    splits:
      - name: train
        num_bytes: 3736144225.68
        num_examples: 33964
      - name: test
        num_bytes: 135062391.97
        num_examples: 1231
    download_size: 3731790647
    dataset_size: 3871206617.6499996
  - config_name: bm-fr-en-2
    features:
      - name: audio
        dtype:
          audio:
            sampling_rate: 16000
      - name: bambara
        dtype: string
      - name: french
        dtype: string
      - name: duration
        dtype: float64
      - name: english
        dtype: string
    splits:
      - name: train
        num_bytes: 3736202782.68
        num_examples: 33964
      - name: test
        num_bytes: 135064226.97
        num_examples: 1231
    download_size: 3731842370
    dataset_size: 3871267009.6499996
  - config_name: bm-fr-en-3
    features:
      - name: audio
        dtype:
          audio:
            sampling_rate: 16000
      - name: bambara
        dtype: string
      - name: duration
        dtype: float64
      - name: english
        dtype: string
      - name: french
        dtype: string
    splits:
      - name: train
        num_bytes: 6329483860.6
        num_examples: 61880
      - name: test
        num_bytes: 135064226.97
        num_examples: 1231
    download_size: 6269924862
    dataset_size: 6464548087.570001
  - config_name: clean
    features:
      - name: audio
        dtype:
          audio:
            sampling_rate: 16000
      - name: bambara
        dtype: string
      - name: french
        dtype: string
      - name: duration
        dtype: float64
      - name: speaker_emb
        sequence:
          sequence: float32
      - name: speaker_id
        dtype: int64
    splits:
      - name: train
        num_bytes: 2702292943.98
        num_examples: 25554
      - name: test
        num_bytes: 674321558.43
        num_examples: 6389
    download_size: 3260864744
    dataset_size: 3376614502.41
  - config_name: clean-combined
    features:
      - name: audio
        dtype:
          audio:
            sampling_rate: 16000
      - name: text
        dtype: string
      - name: duration
        dtype: float64
      - name: source
        dtype: string
    splits:
      - name: train
        num_bytes: 4260675884.644
        num_examples: 39646
      - name: test
        num_bytes: 223586375.36
        num_examples: 2088
    download_size: 4348307332
    dataset_size: 4484262260.004
  - config_name: clean-en
    features:
      - name: audio
        dtype:
          audio:
            sampling_rate: 16000
      - name: bambara
        dtype: string
      - name: french
        dtype: string
      - name: duration
        dtype: float64
      - name: speaker_emb
        sequence:
          sequence: float32
      - name: speaker_id
        dtype: int64
      - name: english
        dtype: string
    splits:
      - name: train
        num_bytes: 2541943414.94
        num_examples: 25062
      - name: test
        num_bytes: 637249483.73
        num_examples: 6279
    download_size: 3074016073
    dataset_size: 3179192898.67
  - config_name: clean-multi-combined
    features:
      - name: audio
        dtype:
          audio:
            sampling_rate: 16000
      - name: text
        dtype: string
      - name: duration
        dtype: float64
      - name: source
        dtype: string
      - name: language
        dtype: string
    splits:
      - name: train
        num_bytes: 6242611905.592
        num_examples: 48208
      - name: test
        num_bytes: 333597770.114
        num_examples: 2838
    download_size: 6381397243
    dataset_size: 6576209675.706
  - config_name: default
    features:
      - name: audio
        dtype:
          audio:
            sampling_rate: 16000
      - name: bambara
        dtype: string
      - name: french
        dtype: string
      - name: duration
        dtype: float64
    splits:
      - name: train
        num_bytes: 3734503380.68
        num_examples: 33964
      - name: test
        num_bytes: 135000460.97
        num_examples: 1231
    download_size: 3730394900
    dataset_size: 3869503841.6499996
  - config_name: with_speaker
    features:
      - name: audio
        dtype:
          audio:
            sampling_rate: 16000
      - name: bambara
        dtype: string
      - name: french
        dtype: string
      - name: duration
        dtype: float64
      - name: speaker_emb
        sequence:
          sequence: float32
      - name: speaker_id
        dtype: int64
    splits:
      - name: train
        num_bytes: 3501158714.68
        num_examples: 33964
      - name: test
        num_bytes: 130803294.97
        num_examples: 1231
    download_size: 3502563220
    dataset_size: 3631962009.6499996
configs:
  - config_name: bm-fr-en
    data_files:
      - split: train
        path: bm-fr-en/train-*
      - split: test
        path: bm-fr-en/test-*
  - config_name: bm-fr-en-2
    data_files:
      - split: train
        path: bm-fr-en-2/train-*
      - split: test
        path: bm-fr-en-2/test-*
  - config_name: bm-fr-en-3
    data_files:
      - split: train
        path: bm-fr-en-3/train-*
      - split: test
        path: bm-fr-en-3/test-*
  - config_name: clean
    data_files:
      - split: train
        path: clean/train-*
      - split: test
        path: clean/test-*
  - config_name: clean-combined
    data_files:
      - split: train
        path: clean-combined/train-*
      - split: test
        path: clean-combined/test-*
  - config_name: clean-en
    data_files:
      - split: train
        path: clean-en/train-*
      - split: test
        path: clean-en/test-*
  - config_name: clean-multi-combined
    data_files:
      - split: train
        path: clean-multi-combined/train-*
      - split: test
        path: clean-multi-combined/test-*
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
  - config_name: with_speaker
    data_files:
      - split: train
        path: with_speaker/train-*
      - split: test
        path: with_speaker/test-*