FeaturePageSetup
- 
goto src/templates create a json file 
- 
goto src/type/template.ts create add templates: key is path(also is task_type), value is json file 
json key
- 
input: - name: request backend key
- show_name: ui show name for user
- type: image / video / text / zip / aspect / select / boolean / int / float / single_text
 
- 
task_base_url: request backend url 
- 
tips: ui show tips 
- 
output_format: .txt / .png / .mp4 / .safetensors 
- 
create_page_view: title: page title description: page description 
example:
{
    "input_type": [
        {
            "name": "LoadImage_image_4",
            "type": "image",
            "show_name": "image"
        },
        {
            "name": "easy_string_value_9",
            "type": "text",
            "show_name": "question"
        }
    ],
    "task_base_url": "/api/create_image_understanding_task",
    "tips": "cost 15 credits",
    "output_format": ".txt",
    "create_page_view": {
        "title": "",
        "description": "Transform your text prompts into vivid, dynamic videos with our AI-powered text-to-video generator. "
    }
}