openlr

Referensi:

SLR41

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR41')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 5822
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR42

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR42')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 2906
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR43

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR43')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 2064
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR44

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR44')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 4213
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR63

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR63')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 4126
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR64

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR64')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 1569
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR65

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR65')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 4284
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR66

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR66')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 4448
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR69

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR69')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 4240
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR35

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR35')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 185076
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR36

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR36')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 219156
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR70

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR70')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 3359
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR71

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR71')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 4374
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR72

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR72')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 4903
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR73

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR73')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 5447
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR74

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR74')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 617
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR75

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR75')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 3357
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR76

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR76')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 7136
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR77

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR77')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 5587
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR78

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR78')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 4272
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR79

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR79')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 4400
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR80

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR80')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 2530
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR86

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR86')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 3583
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR32

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR32')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 9821
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "struct",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR52

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR52')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 185293
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR53

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR53')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 218703
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR54

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR54')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 157905
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

SLR83

Gunakan perintah berikut untuk memuat kumpulan data ini di TFDS:

ds = tfds.load('huggingface:openslr/SLR83')
  • Keterangan :
OpenSLR is a site devoted to hosting speech and language resources, such as training corpora for speech recognition,
and software related to speech recognition. We intend to be a convenient place for anyone to put resources that
they have created, so that they can be downloaded publicly.
  • Lisensi : Tidak ada lisensi yang diketahui
  • Versi : 0.0.0
  • Perpecahan :
Membelah Contoh
'train' 17877
  • Fitur :
{
    "path": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "audio": {
        "sampling_rate": 48000,
        "mono": true,
        "_storage_dtype": "string",
        "id": null,
        "_type": "Audio"
    },
    "sentence": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}