metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: dialogue
list:
- name: Rec_Id
dtype: int64
- name: annotation
sequence: string
- name: end_role
dtype: string
- name: end_time_in_seconds
dtype: int64
- name: speaker
dtype: string
- name: start_role
dtype: string
- name: start_time_in_seconds
dtype: int64
- name: target
dtype: string
- name: timestamp
dtype: string
- name: utterance
dtype: string
- name: start
dtype: int64
- name: end
dtype: int64
- name: idx
dtype: int64
- name: Game_ID
dtype: string
- name: file_name
dtype: string
- name: video_name
dtype: string
- name: startRoles
sequence: string
- name: startTime
dtype: string
- name: endRoles
sequence: string
- name: playerNames
sequence: string
- name: prompt
dtype: string
- name: completion
dtype: string
splits:
- name: train
num_bytes: 6085053775
num_examples: 12159
- name: test
num_bytes: 1766535619
num_examples: 3542
- name: validation
num_bytes: 1094969844
num_examples: 2230
download_size: 8409733834
dataset_size: 8946559238
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*