metadata
dataset_info:
features:
- name: text
dtype: string
- name: qwen
dtype: string
- name: llama
dtype: string
- name: qwen_score
dtype: float64
- name: llama_score
dtype: float64
- name: winner
dtype: string
- name: eng_count
dtype: string
- name: qwen_count
dtype: string
- name: llama_count
dtype: string
splits:
- name: train
num_bytes: 27475643
num_examples: 13377
download_size: 11760565
dataset_size: 27475643
configs:
- config_name: default
data_files:
- split: train
path: data/train-*