初始化项目,由ModelHub XC社区提供模型

Model: linjh1118/Llama3-Chinese-pro-8.4B-sft-1M
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-06-29 16:06:13 +08:00
commit f437b0e601
20 changed files with 427708 additions and 0 deletions

98
README.md Normal file
View File

@@ -0,0 +1,98 @@
# Llama3-Chinese-ORPO
基于Llama3通过进一步CPTSFTORPO得到的中文版Llama3
=log: 2024/04/21 12:00=: 本仓库是出于 *学习目的* 而进行的一个实践项目。
## News
[2024/04/22] We release the first version of Llama3-Chinese-ORPO.
## 模型列表
| 模型名称 | 模型简介 | ModelScope下载链接 | Huggingface下载链接 |
|--------------------------------|---------------------------------------------------------------------------------------------------------------------|---|---|
| Llama3-Chinese-pro-8.4B-sft-1M | 在对llama3-8b的base版本meta-llama/Meta-Llama-3-8B进行llama-pro拓展了两层后进而通过一百万条SFT数据训练而得到。 | [Llama3-Chinese-pro-8.4B-sft-1M](https://modelscope.cn/models/linjh1118/Llama3-Chinese-pro-8.4B-sft-1M) | |
| Llama3-Chinese-pro-8.4B-sft-1M-ORPO | 基于上一行产出的模型继续做ORPO进而得到 | | |
| | | | |
## 测试表现
<details>
<summary>展开查看 测试表现</summary>
1. **随意的五个问题**
### 先插入放两张没有训练完毕的Llama3-Chinese-pro-8.4B-sft-1M的回复哈哈
![Alt text](image-2.png)
### 回到正经测评
![Alt text](image.png)
![Alt text](image-1.png)
</details>
## 主要贡献者 Contributors
**排名不分*先后***
<table>
<tr>
<td align='center'>
<img src="https://avatars.githubusercontent.com/u/67041238?v=4" alt="Contributor 1" height="150">
<br>
<b>linjh1118 @ Northeastern University</b>
<br>
<a href='https://github.com/linjh1118'>Llama3-Chinese-ORPO项目的核心贡献者</a>
</td>
<td align='center'>
<img src="https://avatars.githubusercontent.com/u/59789526?v=4" alt="Contributor 2" height="150">
<br>
<b>Charon-ops @ Tsinghua University</b>
<br>
<a href='https://github.com/AIR-hl'>Llama3-Chinese-ORPO项目的核心贡献者</a>
</td>
<td align='center'>
<img src="https://avatars.githubusercontent.com/u/94281210?v=4" alt="Contributor 3" height="150">
<br>
<b>zzlf @ Northeastern University</b>
<br>
<a href='https://github.com/zzlf'>Llama3-Chinese-ORPO项目的核心贡献者</a>
</td>
</tr>
</table>
<table>
<tr>
<td align='center'>
<img src="https://avatars.githubusercontent.com/u/66808901?v=4" alt="Contributor 2" height="150">
<br>
<b>AIR-hl @ Northeastern University</b>
<br>
<a href='https://github.com/AIR-hl'>Llama3-Chinese-ORPO项目的核心贡献者</a>
</td>
<td align='center'>
<img src="https://avatars.githubusercontent.com/u/116535337?v=4" alt="Contributor 3" height="150">
<br>
<b>yanyi74 @ Northeastern University</b>
<br>
<a href='https://github.com/yanyi74'>Llama3-Chinese-ORPO项目的核心贡献者</a>
</td>
<td align='center'>
<img src="https://avatars.githubusercontent.com/u/84018897?v=4" alt="Contributor 1" height="150">
<br>
<b>MrChen-NEU @ Northeastern University</b>
<br>
<a href='https://github.com/MrChen-NEU'>Llama3-Chinese-ORPO项目的核心贡献者</a>
</td>
</tr>
</table>
非常非常感谢前人的优秀工作https://github.com/CrazyBoyM/llama3-Chinese-chat (全球第一个中文llama3)