hinglish / README.md
nikhilhyperneuron's picture
Update README.md
ff7608c verified
|
Raw
History Blame Contribute Delete
946 Bytes
metadata
license: other
license_name: private-dataset-hyperneuron
license_link: LICENSE
dataset_info:
  features:
    - name: audio
      dtype:
        audio:
          sampling_rate: 44000
    - name: transcript
      dtype: string
    - name: audio_duration
      dtype: float64
    - name: audio_array
      list: float32
    - name: sample_rate
      dtype: int64
  splits:
    - name: hinglish
      num_bytes: 2427577577.326
      num_examples: 4982
    - name: hindi
      num_bytes: 791265298
      num_examples: 409
  download_size: 1646863265
  dataset_size: 3218842875.326
configs:
  - config_name: default
    data_files:
      - split: hinglish
        path: data/train1to4982-*
      - split: hindi
        path: data/hindi1to409-*
size_categories:
  - 1K<n<10K

This dataset contains hindi and hinglish. There are ~5000 rows which aggregates to 3+hrs of audio data.

This is a private dataset and a properiatory data of HyperneuronAI ,using this without permission will lead to law suit.