metadata
dataset_info:
features:
- name: instruction
dtype: string
- name: original_response
dtype: string
- name: wizardlm
dtype: string
- name: gpt4o
dtype: string
- name: wizardlm_score
dtype: float64
- name: gpt4o_score
dtype: float64
splits:
- name: train
num_bytes: 498782005
num_examples: 49938
download_size: 255970853
dataset_size: 498782005
configs:
- config_name: default
data_files:
- split: train
path: data/train-*