Default Values

Default Values#

Default values can be set in code, and this is often done in PyRITInitializers.

How Default Values Work#

When an initializer calls set_default_value, it registers a default value for a specific class and parameter combination. These defaults are stored in a global registry and are automatically applied when classes are instantiated.

One of the most important things to understand is that explicitly provided values always override defaults. Defaults only apply when:

  1. A parameter is not provided at all, OR

  2. A parameter is explicitly set to None

If you pass a value (even 0, False, or ""), that value will be used instead of the default.

Using apply_defaults Decorator#

First, it’s good to be selective over which classes can use this. It is very powerful but can also make debugging more difficult.

Classes that want to participate in the default value system use the @apply_defaults decorator on their __init__ method:

from pyrit.common.apply_defaults import apply_defaults

class MyConverter(PromptConverter):
    @apply_defaults
    def __init__(self, *, converter_target: Optional[PromptChatTarget] = None, temperature: Optional[float] = None):
        self.converter_target = converter_target
        self.temperature = temperature

When you create an instance:

# Uses defaults for both parameters (if configured)
converter1 = MyConverter()

# Uses provided value for converter_target, default for temperature
converter2 = MyConverter(converter_target=my_target)

# Uses provided values for both (defaults ignored)
converter3 = MyConverter(converter_target=my_target, temperature=0.8)

Defaults can be set on parent classes and will apply to subclasses (unless include_subclasses=False is specified).

Contents