keras-team/keras · error · ValueError

If using `weights="imagenet"` as true, `classes` should be 1

Error message

If using `weights="imagenet"` as true, `classes` should be 1000

What it means

In the cropping output spec, top_cropping (explicit or inferred as height - target_height - bottom_cropping) must be >= 0. Negative top cropping means you asked to crop more than the image height allows (or to grow via negative crop).

Source

Thrown at keras/src/applications/efficientnet.py:283

          on the "top" layer. Ignored unless `include_top=True`. Set
          `classifier_activation=None` to return the logits of the "top" layer.

    Returns:
        A model instance.
    """
    if blocks_args == "default":
        blocks_args = DEFAULT_BLOCKS_ARGS

    if not (weights in {"imagenet", None} or file_utils.exists(weights)):
        raise ValueError(
            "The `weights` argument should be either "
            "`None` (random initialization), `imagenet` "
            "(pre-training on ImageNet), "
            "or the path to the weights file to be loaded."
        )

    if weights == "imagenet" and include_top and classes != 1000:
        raise ValueError(
            'If using `weights="imagenet"` with `include_top`'
            " as true, `classes` should be 1000"
        )

    # Determine proper input shape
    input_shape = imagenet_utils.obtain_input_shape(
        input_shape,
        default_size=default_size,
        min_size=32,
        data_format=backend.image_data_format(),
        require_flatten=include_top,
        weights=weights,
    )

    if input_tensor is None:
        img_input = layers.Input(shape=input_shape)
    else:
        if not backend.is_keras_tensor(input_tensor):

View on GitHub (pinned to 7a34a03db6)

Solutions

1. Use pad_images to enlarge instead of negative cropping
2. Ensure target_height <= height - top_cropping - bottom_cropping
3. Clamp crop amounts to the available extent

Example fix

# before
out = ops.image.crop_images(img, target_height=(256, 256))  # img is 128 tall
# after
out = ops.image.pad_images(img, target_height=(256, 256))

if height is not None and target_height + (top_cropping or 0) + (bottom_cropping or 0) > height:
    raise ValueError('crop window exceeds image height')

When it happens

Trigger: Explicit negative top_cropping, or croppings left None with target_height > height.

Common situations: Upsizing attempts via negative crop values; symmetric crop arithmetic going negative on small images; off-by-one target sizes.

Related errors

• The number of repeats in `EfficientNet` must be > 0. Receive
• If using `weights="imagenet"` as true, `classes` should be 1
• The number of repeats in `EfficientNetV2` must be > 0. Recei
• weights_path undefined
• The `weights` argument should be either `None` (random initi

AI-assisted analysis of keras-team/keras@7a34a03db6 (2026-08-25). Data as JSON: /api/errors/f1cab21d36da304f. Report an issue: GitHub.