metadata
dataset_info:
features:
- name: id
dtype: string
- name: sequence
struct:
- name: main_story
dtype: bool
- name: order
dtype: int64
- name: timeline
dtype: string
- name: type
dtype: string
- name: visible
dtype: bool
- name: source
struct:
- name: act
dtype: int64
- name: action_seq_key
dtype: string
- name: chapter
dtype: int64
- name: event_id
dtype: string
- name: label
dtype: string
- name: line_index
dtype: int64
- name: role
dtype: string
- name: script_id
dtype: string
- name: speaker_id
dtype: string
- name: action_seq
struct:
- name: ja
list:
- name: action
dtype: string
- name: character
dtype: string
- name: zh_hans
list:
- name: action
dtype: string
- name: character
dtype: string
- name: mcq
struct:
- name: accepted_answers
list:
list: string
- name: answer
list: string
- name: choices
list:
- name: action
struct:
- name: code
dtype: string
- name: ja
dtype: string
- name: zh_hans
dtype: string
- name: id
dtype: string
- name: label
dtype: string
- name: statement
struct:
- name: span_ja
dtype: string
- name: span_zh_hans
dtype: string
- name: speaker_ja
dtype: string
- name: speaker_zh_hans
dtype: string
- name: text
struct:
- name: ja
dtype: string
- name: zh_hans
dtype: string
- name: id
dtype: string
- name: question
struct:
- name: ja
dtype: string
- name: zh_hans
dtype: string
- name: source_type
dtype: string
- name: subtype
dtype: string
- name: type
dtype: string
- name: state
struct:
- name: benchmark_id
dtype: string
- name: branch_choice_id
dtype: string
- name: fork_from_order
dtype: int64
- name: guard_expression
dtype: string
- name: guard_result
dtype: bool
- name: parent_case_id
dtype: string
- name: parent_row_id
dtype: string
- name: parent_timeline
dtype: string
- name: selected_choice_id
dtype: string
- name: state_before_json
dtype: string
- name: state_change_after_json
dtype: string
- name: state_change_before_json
dtype: string
- name: state_change_key
dtype: string
- name: target_label
dtype: string
splits:
- name: train
num_bytes: 11251438
num_examples: 22820
download_size: 8589533
dataset_size: 11251438
configs:
- config_name: default
data_files:
- split: train
path: data/train-*