tensorflow/models · error · ValueError

Strides must match in the spatial dimensions, got {}

Error message

Strides must match in the spatial dimensions, got {}

What it means

Error "Strides must match in the spatial dimensions, got {}" thrown in tensorflow/models.

Source

Thrown at official/projects/movinet/modeling/movinet.py:620

                input_shape[0],
                kernel_size[0] - 1,
                divide_resolution(input_shape[2], num_downsamples),
                divide_resolution(input_shape[3], num_downsamples),
                expand_filters,
            )

          if '3d' in self._se_type:
            states[f'{prefix}_pool_buffer'] = (
                input_shape[0], 1, 1, 1, expand_filters,
            )
            states[f'{prefix}_pool_frame_count'] = (1,)

          if use_positional_encoding:
            name = f'{prefix}_pos_enc_frame_count'
            states[name] = (1,)

          if strides[1] != strides[2]:
            raise ValueError('Strides must match in the spatial dimensions, '
                             'got {}'.format(strides))

          # If we use a 3D kernel, we apply spatial downsampling
          # after the buffer.
          if (tuple(strides[1:3]) != (1, 1) and
              self._conv_type not in ['2plus1d', '3d_2plus1d']):
            num_downsamples += 1
      elif isinstance(block, HeadSpec):
        states['state_head_pool_buffer'] = (
            input_shape[0], 1, 1, 1, block.project_filters,
        )
        states['state_head_pool_frame_count'] = (1,)

    return states

  def _get_state_dtype(self, name: str) -> str:
    """Returns the dtype associated with a state."""
    if 'frame_count' in name:

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/projects/movinet/modeling/movinet.py:620 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of tensorflow/models@e006f5f0d5 (2026-08-24). Data as JSON: /api/errors/41c93dcb0591b198. Report an issue: GitHub.