https://github.com/lumenvestige/json2dslformatmodel
Fine-tuned from the T5 model, this project parses complex JSON structures and converts them into a Vue-like template DSL
https://github.com/lumenvestige/json2dslformatmodel
ai dsl json t5-model
Last synced: 11 months ago
JSON representation
Fine-tuned from the T5 model, this project parses complex JSON structures and converts them into a Vue-like template DSL
- Host: GitHub
- URL: https://github.com/lumenvestige/json2dslformatmodel
- Owner: LumenVestige
- Created: 2025-06-18T01:26:52.000Z (about 1 year ago)
- Default Branch: master
- Last Pushed: 2025-06-19T00:30:47.000Z (about 1 year ago)
- Last Synced: 2025-06-19T01:29:34.472Z (about 1 year ago)
- Topics: ai, dsl, json, t5-model
- Language: Python
- Homepage:
- Size: 1.98 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
👀
+ html访问环境：http://127.0.0.1:8888/dsl/test
+ API接口:http://127.0.0.1:8099/dsl2
+ 请求方式：post提交，参数为form-data格式，参数名为data,值为jsonstring
+ curl url 示范例子(注意jsonstring转义问题)：
```
curl -X POST http://127.0.0.1:8099/dsl2 -d "data={\"a\":{\"b\":\"图片\",\"c\":[{\"d\":\"https://example.com/image1.jpg\",\"e\":\"风景\",\"f\":{\"g\":\"自然\",\"h\":\"山景\"}},{\"d\":\"https://example.com/image2.jpg\",\"e\":\"人物\",\"f\":{\"g\":\"肖像\",\"h\":\"微笑\"},\"i\":{\"j\":\"摄影师\",\"k\":\"张三\"}}]},\"l\":{\"m\":\"新闻\",\"n\":[{\"o\":\"国际\",\"p\":\" 俄乌冲突\",\"q\":\"最新进展\",\"r\":\"紧张局势升级\"},{\"o\":\"科技\",\"p\":\"AI发展\",\"q\":\"突破\",\"s\":\"自动驾驶\"}],\"t\":{\"u\":\"国内\",\"v\":\"经济\",\"w\":\"政策调整\",\"x\":\"促进增长\"}},\"y\":{\"z\":\"关于我们\",\"aa\":\"公司简介\",\"ab\":\"我们是一家专注于科技创新的公司\",\"ac\":{\"ad\":\"团队\",\"ae\":\"精英\",\"af\":\"专业\",\"ag\":\"热情\"},\"ah\":{\"ai\":\"联系方式\",\"aj\":\"电话\",\"ak\":\"123-456-7890\",\"al\":\"邮箱\",\"am\":\"info@example.com\"}},\"an\":{\"ao\":\"产品列表\",\"ap\":[{\"aq\":\"产品1\",\"ar\":\"智能音箱\",\"as\":\"https://example.com/product1.jpg\",\"at\":\"语音助手\"},{\"aq\":\"产品2\",\"ar\":\"智能手表\",\"as\":\"https://example.com/product2.jpg\",\"at\":\"健康监测\"},{\"aq\":\"产品3\",\"ar\":\"智能眼镜\",\"as\":\"https://example.com/product3.jpg\",\"at\":\"AR体验\",\"au\":{\"av\":\"特色\",\"aw\":\"高清显示\"}}]}}"
```
+ python
```
import json
import requests
url = "http://127.0.0.1:8099/dsl2"
data = {
"a": {
"b": "图片",
"c": [{
"d": "https://example.com/image1.jpg",
"e": "风景",
"f": {
"g": "自然",
"h": "山景"
}
}, {
"d": "https://example.com/image2.jpg",
"e": "人物",
"f": {
"g": "肖像",
"h": "微笑"
},
"i": {
"j": "摄影师",
"k": "张三"
}
}]
},
"l": {
"m": "新闻",
"n": [{
"o": "国际",
"p": " 俄乌冲突",
"q": "最新进展",
"r": "紧张局势升级"
}, {
"o": "科技",
"p": "AI发展",
"q": "突破",
"s": "自动驾驶"
}],
"t": {
"u": "国内",
"v": "经济",
"w": "政策调整",
"x": "促进增长"
}
},
"y": {
"z": "关于我们",
"aa": "公司简介",
"ab": "我们是一家专注于科技创新的公司",
"ac": {
"ad": "团队",
"ae": "精英",
"af": "专业",
"ag": "热情"
},
"ah": {
"ai": "联系方式",
"aj": "电话",
"ak": "123-456-7890",
"al": "邮箱",
"am": "info@example.com"
}
},
"an": {
"ao": "产品列表",
"ap": [{
"aq": "产品1",
"ar": "智能音箱",
"as": "https://example.com/product1.jpg",
"at": "语音助手"
}, {
"aq": "产品2",
"ar": "智能手表",
"as": "https://example.com/product2.jpg",
"at": "健康监测"
}, {
"aq": "产品3",
"ar": "智能眼镜",
"as": "https://example.com/product3.jpg",
"at": "AR体验",
"au": {
"av": "特色",
"aw": "高清显示"
}
}]
}
}
headers = {"Content-Type": "application/x-www-form-urlencoded"}
response = requests.post(url, data={"data": json.dumps(data, ensure_ascii=True)})
print(response.status_code)
print(response.text)
```
+ 页面预览：
+ 简介
+ 基于 T5 模型微调，可以解析有一定复杂层次的 JSON 数据结构，并转为类VUE模板DSL描述语言，适用于列表页、落地页等 JSON 格式。
+ 使用说明
+ 试试手气:随机抽取一条样本数据（列表页或落地页 JSON 数据）进行模型推理测试。
+ 开始推理:使用模型推理。
+ 支持标签：
```
布局类型：block，flex，list，li_item，card
文本类型：title，paragraph，text，bold_text，small_text，italic_text，address
资源类型：img，radian_img，video，media，
链接类型：link
```
+ 支持属性：
```
布局属性：flex-justify-content，flex-align-items
内容属性：content，src，href
逻辑属性：for，if，else-if，else
其他属性：border-radius
```
+ 服务架构：
+ 依赖库安装：
```
pip install requests
pip install transformers
pip install beautifulsoup4
pip install jsonfixer
pip install psutil gputil websockets
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121
npm install pm2 -g
```
+ 启动服务：
```
cd ./workspace
sh ./start.sh
```