ksponspeech_05 / README.md
DragonLine's picture
Upload dataset
8c1022d
metadata
dataset_info:
  features:
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: transcripts
      dtype: string
  splits:
    - name: train
      num_bytes: 18234945060
      num_examples: 101236
    - name: test
      num_bytes: 2410514563.634364
      num_examples: 12655
    - name: valid
      num_bytes: 2300143023.7636356
      num_examples: 12654
  download_size: 21259055664
  dataset_size: 22945602647.398003
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
      - split: valid
        path: data/valid-*