shibing624
commited on
Commit
•
3f167e2
1
Parent(s):
456d588
Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,111 @@
|
|
1 |
---
|
2 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
+
language:
|
3 |
+
- zh
|
4 |
+
tags:
|
5 |
+
- chatglm
|
6 |
+
- pytorch
|
7 |
+
- zh
|
8 |
+
- Text2Text-Generation
|
9 |
+
license: "apache-2.0"
|
10 |
+
widget:
|
11 |
+
- text: "对下面中文拼写纠错:\n少先队员因该为老人让坐。\n答:"
|
12 |
+
|
13 |
---
|
14 |
+
|
15 |
+
# Chinese Spelling Correction LoRA Model
|
16 |
+
ChatGLM中文纠错LoRA模型
|
17 |
+
|
18 |
+
`chatglm-6b-csc-zh-lora` evaluate test data:
|
19 |
+
|
20 |
+
The overall performance of chatglm-6b-csc-zh-lora on CSC **test**:
|
21 |
+
|
22 |
+
|prefix|input_text|target_text|pred|
|
23 |
+
|:-- |:--- |:--- |:-- |
|
24 |
+
|对下面中文拼写纠错:|少先队员因该为老人让坐。|少先队员应该为老人让座。|少先队员应该为老人让座。\n错误字:因,坐|
|
25 |
+
|
26 |
+
在CSC测试集上生成结果纠错准确率高,由于是基于大模型,结果常常能带给人惊喜,不仅能纠错,还带有句子润色和改写功能。
|
27 |
+
|
28 |
+
|
29 |
+
## Usage
|
30 |
+
|
31 |
+
本项目开源在lmft项目:[textgen](https://github.com/shibing624/lmft),可支持ChatGLM模型,通过如下命令调用:
|
32 |
+
|
33 |
+
Install package:
|
34 |
+
```shell
|
35 |
+
pip install -U lmft
|
36 |
+
```
|
37 |
+
|
38 |
+
```python
|
39 |
+
from lmft import ChatGlmModel
|
40 |
+
model = ChatGlmModel("chatglm", "THUDM/chatglm-6b", lora_name="shibing624/chatglm-6b-csc-zh-lora")
|
41 |
+
r = model.predict(["对下面中文拼写纠错:\n少先队员因该为老人让坐。\n答:"])
|
42 |
+
print(r) # ['少先队员应该为老人让座。\n错误字:因,坐']
|
43 |
+
```
|
44 |
+
|
45 |
+
## Usage (HuggingFace Transformers)
|
46 |
+
Without [lmft](https://github.com/shibing624/lmft), you can use the model like this:
|
47 |
+
|
48 |
+
First, you pass your input through the transformer model, then you get the generated sentence.
|
49 |
+
|
50 |
+
Install package:
|
51 |
+
```
|
52 |
+
pip install transformers
|
53 |
+
```
|
54 |
+
|
55 |
+
```python
|
56 |
+
import sys
|
57 |
+
from peft import PeftModel
|
58 |
+
from transformers import AutoModel, AutoTokenizer
|
59 |
+
|
60 |
+
sys.path.append('..')
|
61 |
+
|
62 |
+
model = AutoModel.from_pretrained("THUDM/chatglm-6b", trust_remote_code=True, device_map='auto')
|
63 |
+
model = PeftModel.from_pretrained(model, "shibing624/chatglm-6b-csc-zh-lora")
|
64 |
+
model = model.half().cuda() # fp16
|
65 |
+
tokenizer = AutoTokenizer.from_pretrained("THUDM/chatglm-6b", trust_remote_code=True)
|
66 |
+
|
67 |
+
sents = ['对下面中文拼写纠错:\n少先队员因该为老人让坐。\n答:',
|
68 |
+
'对下面中文拼写纠错:\n下个星期,我跟我朋唷打算去法国玩儿。\n答:']
|
69 |
+
for s in sents:
|
70 |
+
response = model.chat(tokenizer, s, max_length=128, eos_token_id=tokenizer.eos_token_id)
|
71 |
+
print(response)
|
72 |
+
```
|
73 |
+
|
74 |
+
output:
|
75 |
+
```shell
|
76 |
+
('少先队员应该为老人让座。\n错误字:因,坐', [('对下面中文拼写纠错:\n少先队员因该为老人让坐。\n答:', '少先队员应该为老人让座。\n错误字:因,坐')])
|
77 |
+
('下个星期,我跟我朋友打算去法国玩儿。\n错误字:唷', [('对下面中文拼写纠错:\n下个星期,我跟我朋唷打算去法国玩儿。\n答:', '下个星期,我跟我朋友打算去法国玩儿。\n错误字:唷')])
|
78 |
+
```
|
79 |
+
|
80 |
+
|
81 |
+
模型文件组成:
|
82 |
+
```
|
83 |
+
chatglm-6b-csc-zh-lora
|
84 |
+
├── adapter_config.json
|
85 |
+
└── adapter_model.bin
|
86 |
+
```
|
87 |
+
|
88 |
+
|
89 |
+
### 训练数据集
|
90 |
+
#### 中文纠错数据集
|
91 |
+
|
92 |
+
- 数据:[shibing624/CSC](https://huggingface.co/datasets/shibing624/CSC)
|
93 |
+
|
94 |
+
|
95 |
+
|
96 |
+
如果需要训练ChatGLM模型,请参考[https://github.com/shibing624/lmft](https://github.com/shibing624/lmft)
|
97 |
+
|
98 |
+
|
99 |
+
## Citation
|
100 |
+
|
101 |
+
```latex
|
102 |
+
@software{lmft,
|
103 |
+
author = {Xu Ming},
|
104 |
+
title = {lmft: Implementation of language model finetune},
|
105 |
+
year = {2023},
|
106 |
+
url = {https://github.com/shibing624/lmft},
|
107 |
+
}
|
108 |
+
```
|
109 |
+
|
110 |
+
|
111 |
+
|