Skip to main content

​
Selecting An LLM

Adaline’s Editor’s LLM setting menu lets you choose which LLM processes your prompts. It displays all supported LLMs corresponding to the Providers you have configured in Adaline’s settings: Selecting an LLM in Adaline Each LLM name is displayed in the format "::". This prefix helps you choose which specific provider account to use. This is useful when you have multiple keys for the same provider, such as “OpenAI-dev” and “OpenAI-prod” setups.

​
Configuring an LLM Settings

Configure an LLM’s settings by clicking on the meatballs menu: Changing LLMs' settings in Adaline's editor
NOTE: The settings displayed depend on the selected LLM. The interface automatically shows only relevant parameters for the LLM you selected.

​
Configure Response Format

Among the settings, you can choose the LLM’s response format by clicking on Response format: Configure an LLM's response format Choose among the following:
  • text: Instructs the LLM to return the output in textual format.
  • json_object: Instructs the LLM to return the output in the JSON format in any schema as it deems fit. To work properly, the prompt must contain the term “json” (case insensitive).
  • json_schema: Instructs the LLM to return the output in the JSON format that adheres to a precise schema you define.
Below is an example of how to set up a JSON schema: 
{
  "strict": true,
  "name": "users_data",
  "description": "A schema for creating user data.",
  "schema": {
    "type": "object",
    "properties": {
      "user_city": {
        "type": "string",
        "description": "The city where the user resides."
      },
      "user_name": {
        "type": "string",
        "description": "The full name of the user."
      },
      "user_age": {
        "type": "number",
        "description": "The age of the user in years."
      }
    },
    "required": [
      "user_name",
      "user_age",
      "user_city"
    ],
    "additionalProperties": false
  }
}
Note that:
  • The "strict": true field is mandatory.
  • The name field is mandatory. If you need to use more than one words in it choose between the following options:
    • Use the underscore. For example: users_data.
    • Use the camel case notation. For example: usersData.
  • You must use the "schema" field to define the JSON schema and its subfields.
    WARNING: The schema that you provide to the LLM must adhere to the OpenAI JSON schema structure.
The image below shows an implementation example with the LLM’s response: Schema-based responses in Adaline