Skip to content

prompt_image_generation_task

PromptImageGenerationTask

Bases: BaseImageGenerationTask

Used to generate an image from a text prompt.

Accepts prompt as input in one of the following formats: - template string - TextArtifact - Callable that returns a TextArtifact.

Attributes:

Name Type Description
image_generation_driver BaseImageGenerationDriver

The engine used to generate the image.
negative_rulesets list[Ruleset]

List of negatively-weighted rulesets applied to the text prompt, if supported by the driver.
negative_rules list[Rule]

List of negatively-weighted rules applied to the text prompt, if supported by the driver.
output_dir Optional[str]

If provided, the generated image will be written to disk in output_dir.
output_file Optional[str]

If provided, the generated image will be written to disk as output_file.
Source code in griptape/tasks/prompt_image_generation_task.py 
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
@define
class PromptImageGenerationTask(BaseImageGenerationTask):
    """Used to generate an image from a text prompt.

    Accepts prompt as input in one of the following formats:
    - template string
    - TextArtifact
    - Callable that returns a TextArtifact.

    Attributes:
        image_generation_driver: The engine used to generate the image.
        negative_rulesets: List of negatively-weighted rulesets applied to the text prompt, if supported by the driver.
        negative_rules: List of negatively-weighted rules applied to the text prompt, if supported by the driver.
        output_dir: If provided, the generated image will be written to disk in output_dir.
        output_file: If provided, the generated image will be written to disk as output_file.
    """

    DEFAULT_INPUT_TEMPLATE = "{{ args[0] }}"

    _input: Union[str, TextArtifact, Callable[[BaseTask], TextArtifact]] = field(
        default=DEFAULT_INPUT_TEMPLATE, alias="input"
    )

    @property
    def input(self) -> TextArtifact:
        if isinstance(self._input, TextArtifact):
            return self._input
        if isinstance(self._input, Callable):
            return self._input(self)
        return TextArtifact(J2().render_from_string(self._input, **self.full_context))

    @input.setter
    def input(self, value: TextArtifact) -> None:
        self._input = value

    def try_run(self) -> ImageArtifact:
        image_artifact = self.image_generation_driver.run_text_to_image(
            prompts=self._get_prompts(self.input.to_text()),
            negative_prompts=self._get_negative_prompts(),
        )

        if self.output_dir or self.output_file:
            self._write_to_file(image_artifact)

        return image_artifact

DEFAULT_INPUT_TEMPLATE = '{{ args[0] }}' class-attribute instance-attribute

_input = field(default=DEFAULT_INPUT_TEMPLATE, alias='input') class-attribute instance-attribute

input property writable

try_run()

Source code in griptape/tasks/prompt_image_generation_task.py 
47
48
49
50
51
52
53
54
55
56
def try_run(self) -> ImageArtifact:
    image_artifact = self.image_generation_driver.run_text_to_image(
        prompts=self._get_prompts(self.input.to_text()),
        negative_prompts=self._get_negative_prompts(),
    )

    if self.output_dir or self.output_file:
        self._write_to_file(image_artifact)

    return image_artifact