HuggingFaceH4/deita-10k-v0-sft

---
license: mit
language:
- en
size_categories:
- 1K<n<10K
dataset_info:
features:
- name: prompt
dtype: string
- name: prompt_id
dtype: string
- name: messages
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train_sft
num_bytes: 349335841.1
num_examples: 9500
- name: test_sft
num_bytes: 18386096.9
num_examples: 500
- name: train_gen
num_bytes: 336873383
num_examples: 9500
- name: test_gen
num_bytes: 16979716
num_examples: 500
download_size: 289754284
dataset_size: 721575037.0
configs:
- config_name: default
data_files:
- split: train_sft
path: data/train_sft-*
- split: test_sft
path: data/test_sft-*
- split: train_gen
path: data/train_gen-*
- split: test_gen
path: data/test_gen-*
---

# Dataset Card for Deita 10k v0

This is a formatted version of [`hkust-nlp/deita-10k-v0`](https://huggingface.co/datasets/hkust-nlp/deita-10k-v0) to store the conversations in the same format as the OpenAI SDK.

## Citation
If you find this dataset useful, please cite the original dataset:

```
@misc{liu2023what,
title={What Makes Good Data for Alignment? A Comprehensive Study of Automatic Data Selection in Instruction Tuning},
author={Wei Liu and Weihao Zeng and Keqing He and Yong Jiang and Junxian He},
year={2023},
eprint={2312.15685},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```

原始信息汇总

数据集卡片 Deita 10k v0

数据集信息

特征

• prompt: 数据类型为字符串。
• prompt_id: 数据类型为字符串。
• messages: 列表类型，包含以下子特征：
  • content: 数据类型为字符串。
  • role: 数据类型为字符串。

数据分割

• train_sft:
  • 字节数: 349335841.1
  • 样本数: 9500
• test_sft:
  • 字节数: 18386096.9
  • 样本数: 500
• train_gen:
  • 字节数: 336873383
  • 样本数: 9500
• test_gen:
  • 字节数: 16979716
  • 样本数: 500

数据大小

• 下载大小: 289754284
• 数据集大小: 721575037.0

配置

• default:
  • train_sft: 数据路径为 data/train_sft-*
  • test_sft: 数据路径为 data/test_sft-*
  • train_gen: 数据路径为 data/train_gen-*
  • test_gen: 数据路径为 data/test_gen-*