Datasets:

Modalities:
Audio
Text
Formats:
parquet
Languages:
Malayalam
Libraries:
Datasets
Dask
License:
openslr63 / README.md
kavyamanohar's picture
Upload dataset
8615e3b
|
raw
history blame
507 Bytes
metadata
dataset_info:
  features:
    - name: audio
      dtype: audio
    - name: speech_id
      dtype: string
    - name: speaker_id
      dtype: string
    - name: transcript
      dtype: string
  splits:
    - name: train
      num_bytes: 1585742876.38
      num_examples: 3446
    - name: test
      num_bytes: 276840471
      num_examples: 680
  download_size: 1668931471
  dataset_size: 1862583347.38
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*