Update README.md
Browse files
README.md
CHANGED
@@ -5,11 +5,16 @@
|
|
5 |
<p>基于mt5的中文问题生成任务</p>
|
6 |
</h3>
|
7 |
|
|
|
|
|
|
|
|
|
8 |
使用方法请参考github项目:https://github.com/algolet/question_generation
|
9 |
|
|
|
10 |
可以直接在线使用我们的模型:https://www.algolet.com/applications/qg
|
11 |
|
12 |
-
通过transformers调用
|
13 |
``` python
|
14 |
import torch
|
15 |
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
|
@@ -36,6 +41,14 @@ with torch.no_grad():
|
|
36 |
question = tokenizer.decode(outs[0], skip_special_tokens=True)
|
37 |
questions = [q.strip() for q in question.split("<sep>") if len(q.strip()) > 0]
|
38 |
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
39 |
---
|
40 |
language:
|
41 |
- zh
|
|
|
5 |
<p>基于mt5的中文问题生成任务</p>
|
6 |
</h3>
|
7 |
|
8 |
+
#### 可以通过安装question-generation包开始用
|
9 |
+
```
|
10 |
+
pip install question-generation
|
11 |
+
```
|
12 |
使用方法请参考github项目:https://github.com/algolet/question_generation
|
13 |
|
14 |
+
#### 在线使用
|
15 |
可以直接在线使用我们的模型:https://www.algolet.com/applications/qg
|
16 |
|
17 |
+
#### 通过transformers调用
|
18 |
``` python
|
19 |
import torch
|
20 |
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
|
|
|
41 |
question = tokenizer.decode(outs[0], skip_special_tokens=True)
|
42 |
questions = [q.strip() for q in question.split("<sep>") if len(q.strip()) > 0]
|
43 |
```
|
44 |
+
|
45 |
+
#### 指标
|
46 |
+
rouge-1: 0.4041
|
47 |
+
|
48 |
+
rouge-2: 0.2104
|
49 |
+
|
50 |
+
rouge-l: 0.3843
|
51 |
+
|
52 |
---
|
53 |
language:
|
54 |
- zh
|