metadata
dataset_info:
config_name: polish
features:
- name: audio
dtype: audio
- name: wav_filesize
dtype: int64
- name: text
dtype: string
- name: transcript_wav2vec
dtype: string
- name: levenshtein
dtype: float64
- name: duration
dtype: float64
- name: num_words
dtype: int64
- name: speaker_id
dtype: int64
splits:
- name: train
num_bytes: 11127461686.356
num_examples: 18719
- name: dev
num_bytes: 356048249
num_examples: 853
- name: test
num_bytes: 367796887
num_examples: 814
download_size: 8114633186
dataset_size: 11851306822.356
configs:
- config_name: polish
data_files:
- split: train
path: polish/train-*
- split: dev
path: polish/dev-*
- split: test
path: polish/test-*