File size: 1,007 Bytes
325eaaa
 
96a449f
 
 
325eaaa
96a449f
35cfa97
 
 
 
 
96a449f
 
 
 
 
 
f5a7c49
 
96a449f
 
 
 
f5a7c49
d0da481
a24fe43
 
 
d0da481
 
 
 
 
 
 
 
35cfa97
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
---
license: llama2
datasets:
- wikimedia/wikipedia
- stingning/ultrachat
---

結合兩個數據庫來做微調模型來達到知識問答和聊天的機器人
- wikimedia/wikipedia
- stingning/ultrachat


1.效率:透過使用GPU加速、LoRA、梯度累積和混合精度訓練(FP16),最大化運算資源和訓練速度。

2.適應性:透過LoRA對模型的特定組件進行微調,它可以以減少參數達到(30%)以更新更有效地適應目標任務的預訓練模型。


## api使用方法:

```python
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("j40pl7lly/fine-tuning-chat-liu")
model = AutoModelForCausalLM.from_pretrained("j40pl7lly/fine-tuning-chat-liu")
```

## Reference
If you use this model and love it, use this to cite it 🤗

## Citation

```
@misc{privacy_faceemotionrecognition_system,
      title={Fine-tuned LLM model based on open source mistral-7B},
      author={Liu Hsin Kuo},
      year={2024},
}
```