metadata
dataset_info:
features:
- name: sample_index
dtype: int32
- name: phonemes_sequence_ipa
dtype: string
- name: phonemes_annotated_ipa
dtype: string
- name: phonemes_annotated_hex
dtype: string
- name: transcription_full
dtype: string
- name: transcription_original
dtype: string
- name: removed_words
dtype: string
- name: left_tokens_count
dtype: int32
- name: right_tokens_count
dtype: int32
- name: removed_tokens_count
dtype: int32
- name: left_duration_s
dtype: float32
- name: right_duration_s
dtype: float32
- name: removed_duration_s
dtype: float32
- name: full_duration_s
dtype: float32
- name: audio_left
dtype:
audio:
sampling_rate: 16000
- name: audio_right
dtype:
audio:
sampling_rate: 16000
- name: audio_removed
dtype:
audio:
sampling_rate: 16000
- name: audio_full
dtype:
audio:
sampling_rate: 16000
splits:
- name: train
num_bytes: 89296098
num_examples: 100
download_size: 86615386
dataset_size: 89296098
configs:
- config_name: default
data_files:
- split: train
path: data/train-*