metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: answer
dtype: string
- name: sound_convo
list:
- name: content
dtype: string
- name: role
dtype: string
- name: text_convo
list:
- name: content
dtype: string
- name: role
dtype: string
- name: sound_transcribe
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 5021897
num_examples: 100
download_size: 1127840
dataset_size: 5021897
configs:
- config_name: default
data_files:
- split: train
path: data/train-*