Generation Task

​
Generation Task Properties

​
Required Properties

​
prompt
string
The input prompt for generation.
​
model
string
The model to use for generation.Allowed values:
  • "gpt-4": GPT-4 model
  • "gpt-4o-mini": GPT-4O Mini model
​
api_key
string
API key for authentication.

​
Optional Properties

​
max_epochs
integer
default:"100"
Maximum number of epochs for reference free generation.
​
response_format
string
Desired format of the response. Reference the type_id present in the types section of the schema.

​
OpenAI API Properties

In addition to the above properties, the generation task supports all OpenAI API call properties. Some commonly used properties include:
​
temperature
float
Controls randomness in output (0.0 to 2.0).
​
max_tokens
integer
Maximum number of tokens to generate.
​
top_p
float
Controls diversity via nucleus sampling.
​
frequency_penalty
float
Reduces repetition of token sequences.
​
presence_penalty
float
Encourages the model to talk about new topics.
​
stop
string or array
Sequences where the API will stop generating further tokens.
For a complete list of OpenAI API properties, please refer to the OpenAI API documentation.

​
Usage Examples

types:
    customer_info: # This gets referenced in the response_format field
        type: object
        properties:
        customer_name:
            type: string
        customer_address:
            type: string
tasks:
    extract_customer_info:
        task_type: generation
        task_properties:
        model: gpt-4o-mini
        prompt: Extract customer info from the given invoice
        response_format: customer_info # Specify the response format, if needed
        api_key: $OPENAI_API_KEY
It’s currently work in progress. If there something else you’d like to see, we’d love to know.