keras-team/keras · error · ValueError

The `padding` argument must be one of 'valid', 'same'. Recei

Error message

The `padding` argument must be one of 'valid', 'same'. Received: padding={padding}

What it means

Error "The `padding` argument must be one of 'valid', 'same'. Received: padding={padding}" thrown in keras-team/keras.

Source

Thrown at keras/src/backend/common/backend_utils.py:20

import math
import operator
import re
import warnings


def _convert_conv_transpose_padding_args_from_keras_to_jax(
    kernel_size, stride, dilation_rate, padding, output_padding
):
    """Convert the padding arguments from Keras to the ones used by JAX.
    JAX starts with an shape of size `(input-1) * stride - kernel_size + 2`,
    then adds `left_pad` on the left, and `right_pad` on the right.
    In Keras, the `padding` argument determines a base shape, to which
    `output_padding` is added on the right. If `output_padding` is None, it will
    be given a default value.
    """

    if padding.lower() not in {"valid", "same"}:
        raise ValueError(
            f"The `padding` argument must be one of 'valid', 'same'. "
            f"Received: padding={padding}"
        )
    kernel_size = (kernel_size - 1) * dilation_rate + 1

    if padding.lower() == "valid":
        # If output_padding is None, we fill it so that the shape of the output
        # is `(input-1)*s + max(kernel_size, stride)`
        output_padding = (
            max(kernel_size, stride) - kernel_size
            if output_padding is None
            else output_padding
        )
        left_pad = kernel_size - 1
        right_pad = kernel_size - 1 + output_padding

    else:
        if output_padding is None:

View on GitHub (pinned to 7a34a03db6)

When it happens

Trigger: Thrown at keras/src/backend/common/backend_utils.py:20 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of keras-team/keras@7a34a03db6 (2026-08-25). Data as JSON: /api/errors/ae314a6dfa6dd6da. Report an issue: GitHub.