metadata
dataset_info:
features:
- name: sample_index
dtype: int64
- name: sample_name
dtype: string
- name: num_generated_tokens
dtype: int64
- name: generation_time
dtype: float64
- name: transcription_full
dtype: string
- name: removed_words
dtype: string
- name: generated_audio
dtype:
audio:
sampling_rate: 16000
- name: full_audio
dtype:
audio:
sampling_rate: 16000
splits:
- name: inference_results
num_bytes: 12850068
num_examples: 100
download_size: 12155039
dataset_size: 12850068
configs:
- config_name: default
data_files:
- split: inference_results
path: data/inference_results-*