distilabel-internal-testing/SystemChat-1.1-Preferences-PairRM

---
dataset_info:
features:
- name: messages
list:
- name: content
dtype: string
- name: role
dtype: string
- name: generations
sequence: string
- name: generation_models
sequence: string
- name: reorder_indices
sequence: int64
- name: reorder_generations
list:
list:
- name: content
dtype: string
- name: role
dtype: string
- name: reorder_generation_models
sequence: string
- name: pair_rm_inputs
dtype: string
- name: pair_rm_outputs
sequence: string
- name: rewards
sequence: float32
splits:
- name: train
num_bytes: 436471323
num_examples: 20193
download_size: 135940765
dataset_size: 436471323
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---

原始信息汇总

数据集概述

数据特征

• messages:
  • content: 字符串类型
  • role: 字符串类型
• generations: 字符串序列
• generation_models: 字符串序列
• reorder_indices: 整数序列
• reorder_generations:
  • content: 字符串类型
  • role: 字符串类型
• reorder_generation_models: 字符串序列
• pair_rm_inputs: 字符串类型
• pair_rm_outputs: 字符串序列
• rewards: 浮点数序列

数据分割

• train:
  • num_bytes: 436471323 字节
  • num_examples: 20193 个样本

数据集大小

• download_size: 135940765 字节
• dataset_size: 436471323 字节

配置

• default:
  • data_files:
    • split: train
    • path: data/train-*