metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: requirements
dtype: string
- name: selection_response_1
dtype: string
- name: selection_response_2
dtype: string
- name: selection_response_3
dtype: string
- name: base_response_1
dtype: string
- name: base_response_2
dtype: string
- name: current_response_1
dtype: string
- name: current_response_2
dtype: string
- name: selection_1_base_1_mean
list: float64
- name: selection_1_base_1_majority
list: float64
- name: selection_1_base_2_mean
list: float64
- name: selection_1_base_2_majority
list: float64
- name: selection_2_base_1_mean
list: float64
- name: selection_2_base_1_majority
list: float64
- name: selection_2_base_2_mean
list: float64
- name: selection_2_base_2_majority
list: float64
- name: selection_3_base_1_mean
list: float64
- name: selection_3_base_1_majority
list: float64
- name: selection_3_base_2_mean
list: float64
- name: selection_3_base_2_majority
list: float64
- name: current_1_base_1_mean
list: float64
- name: current_1_base_1_majority
list: float64
- name: current_1_base_2_mean
list: float64
- name: current_1_base_2_majority
list: float64
- name: current_2_base_1_mean
list: float64
- name: current_2_base_1_majority
list: float64
- name: current_2_base_2_mean
list: float64
- name: current_2_base_2_majority
list: float64
- name: qwen_prompt
dtype: string
- name: qwen_prompt_tokens
sequence: int64
- name: chosen
dtype: string
- name: chosen_reward
dtype: float64
- name: qwen_chosen
dtype: string
- name: qwen_chosen_tokens
sequence: int64
- name: reject
dtype: string
- name: reject_reward
dtype: float64
- name: qwen_reject
dtype: string
- name: qwen_reject_tokens
sequence: int64
- name: g_chosen
dtype: float64
- name: g_reject
dtype: float64
- name: chosen_logprob
dtype: float64
- name: reject_logprob
dtype: float64
splits:
- name: train
num_bytes: 3600010674
num_examples: 39685
- name: test
num_bytes: 90733691
num_examples: 1000
download_size: 815857776
dataset_size: 3690744365
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*