Update README.md
Browse files
README.md
CHANGED
|
@@ -21,12 +21,33 @@ language:
|
|
| 21 |
- [豆角](https://github.com/goodnessSZW)完成了qwen-1.8B Lora和Yi-6B Lora训练，我们会在之后上传
|
| 22 |
- [米唯实](https://github.com/hhhwmws0117)测试并上传模型到hugging face
|
| 23 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
## inference code
|
| 25 |
|
| 26 |
（搭建中）
|
| 27 |
|
| 28 |
https://github.com/LC1332/Zero-Haruhi/blob/main/notebook/HaruhiZeroGradio.ipynb
|
| 29 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 30 |
## TODO
|
| 31 |
|
| 32 |
|
|
|
|
| 21 |
- [豆角](https://github.com/goodnessSZW)完成了qwen-1.8B Lora和Yi-6B Lora训练，我们会在之后上传
|
| 22 |
- [米唯实](https://github.com/hhhwmws0117)测试并上传模型到hugging face
|
| 23 |
|
| 24 |
+
# Haruhi-Zero: Zero-Shot Role-Playing Model Tuned on Yi-6B
|
| 25 |
+
|
| 26 |
+
Main project link: https://github.com/LC1332/Chat-Haruhi-Suzumiya
|
| 27 |
+
|
| 28 |
+
Previous ChatHaruhi models required a character RAG database to complete character creation. However, open-source/closed-source models like Pygmalion, CharacterGLM, CharacterBaichuan have started to support zero-shot role card creation.
|
| 29 |
+
|
| 30 |
+
We constructed and collected 105k Chinese and English conversations, resegmented them into around 120k conversations with a token length of 2500, and combined them with novel data for training.
|
| 31 |
+
|
| 32 |
## inference code
|
| 33 |
|
| 34 |
（搭建中）
|
| 35 |
|
| 36 |
https://github.com/LC1332/Zero-Haruhi/blob/main/notebook/HaruhiZeroGradio.ipynb
|
| 37 |
|
| 38 |
+
## Official Prompt
|
| 39 |
+
|
| 40 |
+
system prompt:
|
| 41 |
+
|
| 42 |
+
```
|
| 43 |
+
You are now in roleplay conversation mode. Pretend to be {bot_name} whose persona follows:
|
| 44 |
+
{persona}
|
| 45 |
+
|
| 46 |
+
You will stay in-character whenever possible, and generate responses as if you were {bot_name}
|
| 47 |
+
```
|
| 48 |
+
|
| 49 |
+
persona a.k.a. bot definition
|
| 50 |
+
|
| 51 |
## TODO
|
| 52 |
|
| 53 |
|