lllyasviel/Fooocus · error · ValueError

Unknown type: {self.type}. Please choose from: 'numpy', 'pil

Error message

Unknown type: {self.type}. Please choose from: 'numpy', 'pil', 'filepath'.

What it means

postprocess() of the vendored gradio Image converts the model output back to a UI value according to self.type; only 'pil', 'numpy', and 'filepath' have branches, and anything else falls to else and raises ValueError. This is the runtime echo of the constructor's type check — it means self.type was changed after construction or bypassed it.

Source

Thrown at modules/gradio_hijack.py:253

    def _format_image(
        self, im: _Image.Image | None
    ) -> np.ndarray | _Image.Image | str | None:
        """Helper method to format an image based on self.type"""
        if im is None:
            return im
        fmt = im.format
        if self.type == "pil":
            return im
        elif self.type == "numpy":
            return np.array(im)
        elif self.type == "filepath":
            path = self.pil_to_temp_file(
                im, dir=self.DEFAULT_TEMP_DIR, format=fmt or "png"
            )
            self.temp_files.add(path)
            return path
        else:
            raise ValueError(
                "Unknown type: "
                + str(self.type)
                + ". Please choose from: 'numpy', 'pil', 'filepath'."
            )

    def preprocess(
        self, x: str | dict[str, str]
    ) -> np.ndarray | _Image.Image | str | dict | None:
        """
        Parameters:
            x: base64 url data, or (if tool == "sketch") a dict of image and mask base64 url data
        Returns:
            image in requested format, or (if tool == "sketch") a dict of image and mask in requested format
        """
        if x is None:
            return x

        mask = None

View on GitHub (pinned to ae05379cc9)

Solutions

1. Never assign to .type after construction — rebuild the component with the desired type.
2. If you must support an extra format, subclass postprocess and add the branch before the else.
3. Re-run the constructor validation invariant: keep self.type within {'numpy','pil','filepath'} at all times.

Example fix

# before
img.type = 'tensor'
# after
img = Image(type='numpy')  # rebuild with supported type

assert img.type in ('numpy', 'pil', 'filepath'), f'Image.type corrupted: {img.type!r}'

def has_valid_image_type(component) -> bool:
    return getattr(component, 'type', None) in ('numpy', 'pil', 'filepath')

When it happens

Trigger: Mutating img.type = 'tensor' (or 'cv2', 'str') on a live component between construction and the callback returning; subclassing and skipping __init__ validation; deserializing a saved component config with a foreign type.

Common situations: Custom tabs that dynamically reconfigure components; deepcopy/pickle round-trips of UI state that carry an old or edited type field.

Related errors

• Invalid value for parameter `type`: {type}. Please choose fr
• Invalid value for parameter `source`: {source}. Please choos
• Image streaming only available if source is 'webcam'.
• Cannot process this value as an Image
• Unsupported image type in input

AI-assisted analysis of lllyasviel/Fooocus@ae05379cc9 (2026-08-15). Data as JSON: /api/errors/293753d1f488d609. Report an issue: GitHub.