metadata
dataset_info:
features:
- name: original_instruction
dtype: string
- name: 1st_instruction_Nemo
dtype: string
- name: 2nd_instruction_Qwen
dtype: string
- name: 1st_instruction_Qwen
dtype: string
- name: usability_2nd_Qwen
dtype: string
- name: usability_1st_Qwen
dtype: string
splits:
- name: train
num_bytes: 18797581
num_examples: 13469
download_size: 9740489
dataset_size: 18797581
configs:
- config_name: default
data_files:
- split: train
path: data/train-*