huggingface/transformers · error · ValueError

max_size = {max_size} must be strictly greater than the requ

Error message

max_size = {max_size} must be strictly greater than the requested size for the smaller edge size = {size}

What it means

Error "max_size = {max_size} must be strictly greater than the requested size for the smaller edge size = {size}" thrown in huggingface/transformers.

Source

Thrown at src/transformers/image_utils.py:851

            size = tuple(size)

        if isinstance(size, int) or len(size) == 1:
            if default_to_square:
                size = (size, size) if isinstance(size, int) else (size[0], size[0])
            else:
                width, height = image.size
                # specified size only for the smallest edge
                short, long = (width, height) if width <= height else (height, width)
                requested_new_short = size if isinstance(size, int) else size[0]

                if short == requested_new_short:
                    return image

                new_short, new_long = requested_new_short, int(requested_new_short * long / short)

                if max_size is not None:
                    if max_size <= requested_new_short:
                        raise ValueError(
                            f"max_size = {max_size} must be strictly greater than the requested "
                            f"size for the smaller edge size = {size}"
                        )
                    if new_long > max_size:
                        new_short, new_long = int(max_size * new_short / new_long), max_size

                size = (new_short, new_long) if width <= height else (new_long, new_short)

        return image.resize(size, resample=resample)

    def center_crop(self, image, size):
        """
        Crops `image` to the given size using a center crop. Note that if the image is too small to be cropped to the
        size given, it will be padded (so the returned result has the size asked).

        Args:
            image (`PIL.Image.Image` or `np.ndarray` or `torch.Tensor` of shape (n_channels, height, width) or (height, width, n_channels)):
                The image to resize.

View on GitHub (pinned to a597f97485)

Solutions

1. Set `max_size` strictly greater than the requested smaller-edge size.
2. Remove `max_size` to disable the cap.

When it happens

Trigger: Raised in resize size computation when max_size is not strictly greater than the requested shorter-edge size.

Common situations: size and max_size configured such that max_size <= size for the smaller edge in resize preprocessing.

AI-assisted analysis of huggingface/transformers@a597f97485 (2026-08-14). Data as JSON: /api/errors/fe8cedf26e5c23db. Report an issue: GitHub.