metadata
dataset_info:
config_name: en-ami-ihm
features:
- name: audio_filepath
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: duration
dtype: float32
- name: text
dtype: string
- name: whisper_transcript
dtype: string
- name: text_norm
dtype: string
- name: whisper_transcript_norm
dtype: string
- name: wer
dtype: float32
- name: prev_text
dtype: string
- name: prev_whisper_transcript
dtype: string
splits:
- name: train
num_bytes: 5876394913.25
num_examples: 6775
download_size: 5470925105
dataset_size: 5876394913.25
configs:
- config_name: en-ami-ihm
data_files:
- split: train
path: en-ami-ihm/train-*