File size: 683 Bytes
20d0a39
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
prompt:
  messages:
    - role: user
      content: 
        - type: image_url
          image_url:
            url: "{image_url}"
        - type: text
          text: "Please detect and label all objects in the following image and mark their positions."

  # Optional: You can list the input variables here to enable validation of the template 
  # This follows the LangChain PromptTemplate convention: https://api.python.langchain.com/en/latest/core/prompts/langchain_core.prompts.prompt.PromptTemplate.html
  input_variables:
      - image_url
      
  # Additional metadata
  metadata:
    model: https://huggingface.co/collections/OpenGVLab/internvl-20-667d3961ab5eb12c7ed1463e