shibing624
commited on
Commit
•
fdc5c15
1
Parent(s):
52164ee
Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,108 @@
|
|
1 |
---
|
2 |
-
license:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
+
license: cc-by-4.0
|
3 |
+
language:
|
4 |
+
- zh
|
5 |
+
tags:
|
6 |
+
- text-generation
|
7 |
+
- e-commerce advertise
|
8 |
+
pretty_name: AdvertiseGen
|
9 |
+
task_categories:
|
10 |
+
- text-generation
|
11 |
---
|
12 |
+
|
13 |
+
# Dataset Card for Alpaca-zh
|
14 |
+
|
15 |
+
- **formal url:** https://www.luge.ai/#/luge/dataDetail?id=9
|
16 |
+
|
17 |
+
## Dataset Description
|
18 |
+
数据集介绍
|
19 |
+
|
20 |
+
AdvertiseGen是电商广告文案生成数据集。
|
21 |
+
|
22 |
+
AdvertiseGen以商品网页的标签与文案的信息对应关系为基础构造,是典型的开放式生成任务,在模型基于key-value输入生成开放式文案时,与输入信息的事实一致性需要得到重点关注。
|
23 |
+
|
24 |
+
|
25 |
+
- 任务描述:给定商品信息的关键词和属性列表kv-list,生成适合该商品的广告文案adv;
|
26 |
+
- 数据规模:训练集114k,验证集1k,测试集3k;
|
27 |
+
- 数据来源:清华大学CoAI小组;
|
28 |
+
|
29 |
+
### Supported Tasks and Leaderboards
|
30 |
+
|
31 |
+
The Alpaca dataset designed for instruction training pretrained language models.
|
32 |
+
|
33 |
+
### Languages
|
34 |
+
|
35 |
+
The data in AdvertiseGen are in Chinese.
|
36 |
+
|
37 |
+
## Dataset Structure
|
38 |
+
|
39 |
+
### Data Instances
|
40 |
+
|
41 |
+
An example of "train" looks as follows:
|
42 |
+
|
43 |
+
```json
|
44 |
+
{
|
45 |
+
"content": "类型#上衣*材质#牛仔布*颜色#白色*风格#简约*图案#刺绣*衣样式#外套*衣款式#破洞",
|
46 |
+
"summary": "简约而不简单的牛仔外套,白色的衣身十分百搭。衣身多处有做旧破洞设计,打破单调乏味,增加一丝造型看点。衣身后背处有趣味刺绣装饰,丰富层次感,彰显别样时尚。"
|
47 |
+
}
|
48 |
+
```
|
49 |
+
|
50 |
+
## Dataset Creation
|
51 |
+
|
52 |
+
### Curation Rationale
|
53 |
+
|
54 |
+
[More Information Needed]
|
55 |
+
|
56 |
+
### Source Data
|
57 |
+
|
58 |
+
#### Initial Data Collection and Normalization
|
59 |
+
|
60 |
+
[More Information Needed]
|
61 |
+
|
62 |
+
#### Who are the source language producers?
|
63 |
+
|
64 |
+
[More Information Needed]
|
65 |
+
|
66 |
+
### Annotations
|
67 |
+
|
68 |
+
#### Annotation process
|
69 |
+
|
70 |
+
[More Information Needed]
|
71 |
+
|
72 |
+
#### Who are the annotators?
|
73 |
+
|
74 |
+
[More Information Needed]
|
75 |
+
|
76 |
+
### Personal and Sensitive Information
|
77 |
+
|
78 |
+
[More Information Needed]
|
79 |
+
|
80 |
+
## Considerations for Using the Data
|
81 |
+
|
82 |
+
|
83 |
+
### Discussion of Biases
|
84 |
+
|
85 |
+
[More Information Needed]
|
86 |
+
|
87 |
+
### Other Known Limitations
|
88 |
+
|
89 |
+
The `alpaca` data is generated by a language model (`text-davinci-003`) and inevitably contains some errors or biases. We encourage users to use this data with caution and propose new methods to filter or improve the imperfections.
|
90 |
+
|
91 |
+
|
92 |
+
## Additional Information
|
93 |
+
|
94 |
+
### Dataset Curators
|
95 |
+
|
96 |
+
[More Information Needed]
|
97 |
+
|
98 |
+
### Citation Information
|
99 |
+
数据集引用
|
100 |
+
|
101 |
+
如在学术论文中使用本数据集,请添加相关引用说明,具体如下:
|
102 |
+
```
|
103 |
+
Shao, Zhihong, et al. "Long and Diverse Text Generation with Planning-based Hierarchical Variational Model." Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.
|
104 |
+
```
|
105 |
+
|
106 |
+
### Contributions
|
107 |
+
|
108 |
+
[More Information Needed]
|