vox-pretrain / README.md
xacer's picture
Update README.md
4e5b6ab verified
|
raw
history blame
2.11 kB
metadata
dataset_info:
  - config_name: test_a
    features:
      - name: audio
        dtype:
          audio:
            sampling_rate: 22050
            decode: false
      - name: librivox_sources
        list:
          - name: reader_id
            dtype: int32
          - name: section_id
            dtype: int32
          - name: book_id
            dtype: int32
          - name: offset_seconds
            dtype: float32
          - name: speed_factor
            dtype: float32
          - name: mixing_factor
            dtype: float32
      - name: audioset_sources
        list:
          - name: tar
            dtype: string
          - name: filename
            dtype: string
          - name: mixing_factor
            dtype: float32
  - config_name: test_b
    features:
      - name: audio
        dtype:
          audio:
            sampling_rate: 22050
            decode: false
      - name: librivox_sources
        list:
          - name: reader_id
            dtype: int32
          - name: section_id
            dtype: int32
          - name: book_id
            dtype: int32
          - name: offset_seconds
            dtype: float32
          - name: speed_factor
            dtype: float32
          - name: mixing_factor
            dtype: float32
      - name: audioset_sources
        list:
          - name: tar
            dtype: string
          - name: filename
            dtype: string
          - name: mixing_factor
            dtype: float32
  - config_name: test_c
    features:
      - name: audio
        dtype:
          audio:
            sampling_rate: 22050
            decode: false
      - name: librivox_sources
        list:
          - name: reader_id
            dtype: int32
          - name: section_id
            dtype: int32
          - name: book_id
            dtype: int32
          - name: offset_seconds
            dtype: float32
          - name: speed_factor
            dtype: float32
          - name: mixing_factor
            dtype: float32
      - name: audioset_sources
        list:
          - name: tar
            dtype: string
          - name: filename
            dtype: string
          - name: mixing_factor
            dtype: float32
configs:
  - config_name: test_a
    data_files:
      - split: train
        path: data/test_a/train-*
  - config_name: test_b
    data_files:
      - split: train
        path: data/test_b/train-*
  - config_name: test_c
    data_files:
      - split: train
        path: data/test_c/train-*
license: apache-2.0
task_categories:
  - audio-to-audio
language:
  - en