40 lines
888 B
Markdown
40 lines
888 B
Markdown
|
---
|
|||
|
|
frameworks:
|
|||
|
|
- Pytorch
|
|||
|
|
license: other
|
|||
|
|
tasks:
|
|||
|
|
- text-generation
|
|||
|
|
---
|
|||
|
|
|
|||
|
|
#### Clone with HTTP
|
|||
|
|
```bash
|
|||
|
|
git clone https://www.modelscope.cn/sakuraumi/Sakura-13B-Novel-Korean.git
|
|||
|
|
```
|
|||
|
|
|
|||
|
|
<div align="center">
|
|||
|
|
<h1>
|
|||
|
|
Sakura-13B-Novel-Korean
|
|||
|
|
</h1>
|
|||
|
|
</div>
|
|||
|
|
|
|||
|
|
# 介绍
|
|||
|
|
|
|||
|
|
基于LLaMA2-13B，OpenBuddy构建，在约360K轻小说中韩文本数据上微调2个epoch.
|
|||
|
|
|
|||
|
|
# 模型
|
|||
|
|
|
|||
|
|
- Finetuned by [SakuraUmi](https://github.com/pipixia244)
|
|||
|
|
- Finetuned on [Openbuddy-LLaMA2-13B](https://huggingface.co/OpenBuddy/openbuddy-llama2-13b-v11.1-fp16)
|
|||
|
|
- Data support: [CjangCjengh](https://github.com/CjangCjengh)
|
|||
|
|
- Base model: [LLaMA2-13B](https://huggingface.co/meta-llama/Llama-2-13b-chat-hf)
|
|||
|
|
- Languages: Chinese/Korean
|
|||
|
|
|
|||
|
|
# 推理
|
|||
|
|
|
|||
|
|
- Prompt构建
|
|||
|
|
|
|||
|
|
```python
|
|||
|
|
input_text = "" # 用户输入
|
|||
|
|
query = "将下面的韩语文本翻译成中文：" + input_text
|
|||
|
|
prompt = "User: " + query + "\nAssistant: "
|
|||
|
|
```
|