customdataset / README.md
GGarri's picture
Upload dataset
e0adbda verified
metadata
dataset_info:
  features:
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: transcription
      dtype: string
  splits:
    - name: train
      num_bytes: 621015746.8685446
      num_examples: 894
    - name: test
      num_bytes: 187135363.08763692
      num_examples: 256
    - name: valid
      num_bytes: 98381996.04381846
      num_examples: 128
  download_size: 775819937
  dataset_size: 906533106
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
      - split: valid
        path: data/valid-*