cornfieldrm/pair-preference-dataset-700K_subset-4-of-4_gemma-2b_1of4_iter1_bs128_lr1e-5_conf-0.9

---
dataset_info:
features:
- name: rejected
list:
- name: content
dtype: string
- name: role
dtype: string
- name: rejected_score
dtype: float64
- name: chosen_score
dtype: float64
- name: chosen
list:
- name: content
dtype: string
- name: role
dtype: string
- name: length
dtype: int64
- name: chosen_prob
dtype: float64
- name: messages
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 169620288.73388383
num_examples: 25014
download_size: 117217827
dataset_size: 169620288.73388383
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---

The dataset includes multiple features such as rejected content (rejected), rejected score (rejected_score), chosen score (chosen_score), chosen content (chosen), length (length), chosen probability (chosen_prob), and messages. Each feature has detailed data type and structure descriptions. The dataset is divided into a training set (train) with 25014 samples, with a total size of 169620288.73388383 bytes. The download size of the dataset is 117217827 bytes.

原始信息汇总

数据集概述

数据集信息

特征

• rejected
  • content: 数据类型为字符串
  • role: 数据类型为字符串
• rejected_score: 数据类型为浮点数 (float64)
• chosen_score: 数据类型为浮点数 (float64)
• chosen
  • content: 数据类型为字符串
  • role: 数据类型为字符串
• length: 数据类型为整数 (int64)
• chosen_prob: 数据类型为浮点数 (float64)
• messages
  • content: 数据类型为字符串
  • role: 数据类型为字符串

分割

• train
  • 字节数: 169620288.73388383
  • 样本数: 25014

大小

• 下载大小: 117217827
• 数据集大小: 169620288.73388383

配置

• default
  • 数据文件:
    • train: 路径为 data/train-*