ShenaoZ/0.0001_sft_nodpo_5iters_bs256_5102lr_dataset

---
dataset_info:
features:
- name: prompt
dtype: string
- name: prompt_id
dtype: string
- name: messages
list:
- name: content
dtype: string
- name: role
dtype: string
- name: score_chosen
dtype: float64
- name: score_rejected
dtype: float64
- name: reference_response
dtype: string
- name: chosen
list:
- name: content
dtype: string
- name: role
dtype: string
- name: rejected
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: test_prefs_1
num_bytes: 15193743
num_examples: 2000
- name: train_prefs_1
num_bytes: 93282680
num_examples: 12227
- name: test_prefs_2
num_bytes: 17412805
num_examples: 2000
- name: train_prefs_2
num_bytes: 105612350
num_examples: 12227
- name: test_prefs_3
num_bytes: 17572146
num_examples: 2000
- name: train_prefs_3
num_bytes: 107649723
num_examples: 12227
- name: test_prefs_4
num_bytes: 17573740
num_examples: 2000
- name: train_prefs_4
num_bytes: 107317264
num_examples: 12227
- name: test_prefs_5
num_bytes: 17532324
num_examples: 2000
- name: train_prefs_5
num_bytes: 107844314
num_examples: 12227
download_size: 329521533
dataset_size: 606991089
configs:
- config_name: default
data_files:
- split: test_prefs_1
path: data/test_prefs_1-*
- split: train_prefs_1
path: data/train_prefs_1-*
- split: test_prefs_2
path: data/test_prefs_2-*
- split: train_prefs_2
path: data/train_prefs_2-*
- split: test_prefs_3
path: data/test_prefs_3-*
- split: train_prefs_3
path: data/train_prefs_3-*
- split: test_prefs_4
path: data/test_prefs_4-*
- split: train_prefs_4
path: data/train_prefs_4-*
- split: test_prefs_5
path: data/test_prefs_5-*
- split: train_prefs_5
path: data/train_prefs_5-*
---
# Dataset Card for "0.0001_sft_nodpo_5iters_bs256_5102lr_dataset"

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

The dataset includes multiple features such as prompt, prompt_id, messages (containing content and role), score_chosen, score_rejected, reference_response, chosen, and rejected. The dataset is divided into multiple splits, each with different names, number of bytes, and number of examples. The download size and total size of the dataset are also provided. Configuration information shows the data file paths for different splits.

原始信息汇总

数据集概述

数据集特征

• prompt: 字符串类型
• prompt_id: 字符串类型
• messages: 列表类型，包含：
  • content: 字符串类型
  • role: 字符串类型
• score_chosen: 浮点数类型
• score_rejected: 浮点数类型
• reference_response: 字符串类型
• chosen: 列表类型，包含：
  • content: 字符串类型
  • role: 字符串类型
• rejected: 列表类型，包含：
  • content: 字符串类型
  • role: 字符串类型

数据集分割

• test_prefs_1: 2000个样本，大小15193743字节
• train_prefs_1: 12227个样本，大小93282680字节
• test_prefs_2: 2000个样本，大小17412805字节
• train_prefs_2: 12227个样本，大小105612350字节
• test_prefs_3: 2000个样本，大小17572146字节
• train_prefs_3: 12227个样本，大小107649723字节
• test_prefs_4: 2000个样本，大小17573740字节
• train_prefs_4: 12227个样本，大小107317264字节
• test_prefs_5: 2000个样本，大小17532324字节
• train_prefs_5: 12227个样本，大小107844314字节

数据集大小

• 下载大小: 329521533字节
• 数据集大小: 606991089字节

配置文件

• config_name: default
• data_files:
  • split: 对应于上述分割名称
  • path: 数据文件路径，格式为data/分割名称-*