HODACHI/data-llama-3-8b-instruct-sppo-iter1_generated

---
dataset_info:
features:
- name: prompt
dtype: string
- name: generate_0
list:
- name: content
dtype: string
- name: role
dtype: string
- name: generate_1
list:
- name: content
dtype: string
- name: role
dtype: string
- name: generate_2
list:
- name: content
dtype: string
- name: role
dtype: string
- name: generate_3
list:
- name: content
dtype: string
- name: role
dtype: string
- name: generate_4
list:
- name: content
dtype: string
- name: role
dtype: string
- name: probability
sequence:
sequence: float64
- name: rm_scores
sequence: float32
splits:
- name: train
num_bytes: 1184620
num_examples: 85
download_size: 579660
dataset_size: 1184620
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---
# Dataset Card for "data-llama-3-8b-instruct-sppo-iter1_generated"

[More Information needed](https://github.com/huggingface/datasets/blob/main/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)

原始信息汇总

数据集概述

数据集信息

• 数据集名称: data-llama-3-8b-instruct-sppo-iter1_generated
• 特征:
  • prompt: 字符串类型
  • generate_0:
    • content: 字符串类型
    • role: 字符串类型
  • generate_1:
    • content: 字符串类型
    • role: 字符串类型
  • generate_2:
    • content: 字符串类型
    • role: 字符串类型
  • generate_3:
    • content: 字符串类型
    • role: 字符串类型
  • generate_4:
    • content: 字符串类型
    • role: 字符串类型
  • probability: 浮点数序列
  • rm_scores: 浮点数序列
• 分割:
  • train:
    • 样本数量: 85
    • 字节数: 1184620
• 下载大小: 579660 字节
• 数据集大小: 1184620 字节

配置

• 配置名称: default
• 数据文件:
  • 分割: train
  • 路径: data/train-*