silk-road/Haruhi-Zero

---
license: cc-by-4.0
---

# 用于ChatHaruhi-Zero Extend的训练数据

目前还不知道数据规模 知道的话回头会更名为Haruhi-Zero-XXX K

目前只放出每个source的sample，完整的数据将在1.0 模型放出之后发布

主项目链接 https://github.com/LC1332/Chat-Haruhi-Suzumiya

如果有兴趣加入我们的训练请联系[e-mail]

计划加入的数据源

数据源
- [x] 中文小说数据
- [x] erotics小说数据
- [x] ChatHaruhi 52K, （转了message格式）
- [x] Chinese 13.2k, 转了message格式）
- [x] Waifu-extended 0.2K, 看看方不方便转成message格式，不行就简单的user-AI
- [x] Claude-Baize数据 7.2K
- [x] PIPPA数据 1.68K
- [x] JanitorAI数据
- [ ] PIPPA翻译数据
- [x] RoleLLM 1.6K, 看看方不方便转成message格式，不行就简单的user-AI

# 0.2

进一步去掉AI助理的相关数据

# 0.3

增加身份认知数据

# 0.4

增加小说抽取数据

# 0.5

增加PIPPA翻译，小说数据增加profile

## 赞助

求捐助Claude API 求捐助OpenAI企业API

求赞助资源计算资源中。。。

---
license: CC-BY-4.0
---

# Training Data for ChatHaruhi-Zero Extend

Currently, the total scale of this dataset has not been confirmed. Once confirmed, the dataset will be renamed to Haruhi-Zero-XXX K.

Currently, only samples for each data source are publicly released. The full dataset will be made available after the 1.0 version model is launched.

Main project link: https://github.com/LC1332/Chat-Haruhi-Suzumiya

If you are interested in joining our training effort, please contact [e-mail].

### Planned Data Sources
- [x] Chinese fiction dataset
- [x] Erotic fiction dataset
- [x] ChatHaruhi 52K (converted to message format)
- [x] Chinese 13.2K (converted to message format)
- [x] Waifu-extended 0.2K: will convert to message format if feasible; otherwise adopt simple user-AI format
- [x] Claude-Baize dataset 7.2K
- [x] PIPPA dataset 1.68K
- [x] JanitorAI dataset
- [ ] PIPPA translated dataset
- [x] RoleLLM 1.6K: will convert to message format if feasible; otherwise adopt simple user-AI format

## Version 0.2
Further removed data related to AI assistants.

## Version 0.3
Added identity awareness training data.

## Version 0.4
Added novel extraction training data.

## Version 0.5
Added PIPPA translated dataset, and added profile information to the novel dataset.

## Sponsorship
We are seeking donations of Claude API and OpenAI Enterprise API.
We are currently seeking sponsorship for computing resources.

原始信息汇总

数据集概述

数据集名称

• 当前名称：Haruhi-Zero-XXX K
• 待更新名称：待定（将根据数据规模更新）