dataset_info:
features:
- name: id
dtype: int64 - name: prompt
list:- name: content
dtype: string - name: role
dtype: string
- name: content
- name: completion
list:- name: content
dtype: string - name: role
dtype: string
- name: content
- name: task
dtype: string - name: generation_ms
dtype: float64 - name: scoring_ms
dtype: float64 - name: total_ms
dtype: float64 - name: info
struct:- name: constraint
dtype: string - name: constraint_type
dtype: string
- name: constraint
- name: answer
dtype: string - name: reward
dtype: float64 - name: follwed_instruction
dtype: float64
splits: - name: train
num_bytes: 20863
num_examples: 15
download_size: 19699
dataset_size: 20863
configs: - config_name: default
data_files:- split: train
path: data/train-*
- split: train