metadata
dataset_info:
features:
- name: original_index
dtype: int64
- name: text
dtype: string
- name: group
dtype: string
- name: language
dtype: string
- name: dataset_source
dtype: string
- name: scenario_category
dtype: string
- name: scenario
dtype: string
- name: jb_strategy
dtype: string
- name: jb_prompt
dtype: string
- name: response_model_name
dtype: string
- name: model_response
dtype: string
- name: label_oss120B
dtype: int64
- name: label_llama70B
dtype: int64
- name: label_qwe235B
dtype: int64
- name: label_majority_vote
dtype: int64
splits:
- name: train
num_bytes: 3035116032
num_examples: 527532
download_size: 1061243719
dataset_size: 3035116032
configs:
- config_name: default
data_files:
- split: train
path: data/train-*