Update README.md
Browse files
README.md
CHANGED
@@ -13,7 +13,43 @@ pipeline_tag: text-generation
|
|
13 |
|
14 |

|
15 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
16 |
## Score
|
|
|
17 |
LogicKor
|
18 |
| Category | Single turn | Multi turn |
|
19 |
|---|---|---|
|
|
|
13 |
|
14 |

|
15 |
|
16 |
+
## Model Description
|
17 |
+
|
18 |
+
### Model Details
|
19 |
+
- **Name**: Carrot Llama-3.2 Rabbit Ko 2412
|
20 |
+
- **Version**: 3B Instruct
|
21 |
+
- **Base Model**: CarrotAI/Llama-3.2-Rabbit-Ko-3B-Instruct-2412
|
22 |
+
- **Languages**: Korean, English
|
23 |
+
- **Model Type**: Large Language Model (Instruction-tuned)
|
24 |
+
|
25 |
+
### Training Process
|
26 |
+
๋ณธ ๋ชจ๋ธ์ ๋ค์๊ณผ ๊ฐ์ ์ฃผ์ ํ๋ จ ๋จ๊ณ๋ฅผ ๊ฑฐ์ณค์ต๋๋ค:
|
27 |
+
|
28 |
+
1. **SFT (Supervised Fine-Tuning)**
|
29 |
+
- ๊ณ ํ์ง ํ๊ตญ์ด ๋ฐ ์์ด ๋ฐ์ดํฐ์
์ ์ฌ์ฉํ์ฌ ๊ธฐ๋ณธ ๋ชจ๋ธ์ ์ธ๋ถ ์กฐ์
|
30 |
+
|
31 |
+
2. **DPO (Direct Preference Optimization)**
|
32 |
+
- ์ธ๊ฐ์ ์ ํธ๋๋ฅผ ์ง์ ์ ์ผ๋ก ๋ฐ์ํ์ฌ ๋ชจ๋ธ์ ์๋ต ํ์ง ๊ฐ์
|
33 |
+
- ๋ชจ๋ธ์ ์ผ๊ด์ฑ, ์ ํ์ฑ, ๊ทธ๋ฆฌ๊ณ ์ฌ์ฉ์ ์ ํธ๋๋ฅผ ์ต์ ํ
|
34 |
+
|
35 |
+
### Limitations
|
36 |
+
- 3B ํ๋ผ๋ฏธํฐ ๊ท๋ชจ๋ก ์ธํ ๋ณต์กํ ์์
์์์ ์ ํ์ ์ฑ๋ฅ
|
37 |
+
- ํน์ ๋๋ฉ์ธ์ ๋ํ ๊น์ด ์๋ ์ ๋ฌธ์ฑ ๋ถ์กฑ
|
38 |
+
- ํธํฅ์ฑ ๋ฐ ํ๊ฐ ๊ฐ๋ฅ์ฑ
|
39 |
+
|
40 |
+
### Ethics Statement
|
41 |
+
๋ชจ๋ธ ๊ฐ๋ฐ ๊ณผ์ ์์ ์ค๋ฆฌ์ ๊ณ ๋ ค์ฌํญ์ ์ต๋ํ ๋ฐ์ํ์์ผ๋, ์ฌ์ฉ์๋ ํญ์ ๊ฒฐ๊ณผ๋ฅผ ๋นํ์ ์ผ๋ก ๊ฒํ ํด์ผ ํฉ๋๋ค.
|
42 |
+
|
43 |
+
### How to Use
|
44 |
+
```python
|
45 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
46 |
+
|
47 |
+
model = AutoModelForCausalLM.from_pretrained("CarrotAI/Llama-3.2-Rabbit-Ko-3B-Instruct-2412")
|
48 |
+
tokenizer = AutoTokenizer.from_pretrained("CarrotAI/Llama-3.2-Rabbit-Ko-3B-Instruct-2412")
|
49 |
+
```
|
50 |
+
|
51 |
## Score
|
52 |
+
### Performance Metrics
|
53 |
LogicKor
|
54 |
| Category | Single turn | Multi turn |
|
55 |
|---|---|---|
|