Expect input spec to be 4 or 5 dimensions {input_specs.shape

Error message

Expect input spec to be 4 or 5 dimensions {input_specs.shape}

What it means

Error "Expect input spec to be 4 or 5 dimensions {input_specs.shape}" thrown in tensorflow/models.

Source

Thrown at official/projects/assemblenet/modeling/assemblenet.py:809

    grouping = {-3: [], -2: [], -1: [], 0: [], 1: [], 2: [], 3: []}
    for i in range(len(structure)):
      grouping[structure[i][0]].append(i)

    stem_count = len(grouping[-3]) + len(grouping[-2]) + len(grouping[-1])

    assert stem_count != 0
    stem_filters = 128 // stem_count

    original_inputs = inputs
    if len(input_specs.shape) == 5:
      first_dim = (
          input_specs.shape[0] * input_specs.shape[1]
          if input_specs.shape[0] and input_specs.shape[1] else -1)
      reshape_inputs = tf.reshape(inputs, (first_dim,) + input_specs.shape[2:])
    elif len(input_specs.shape) == 4:
      reshape_inputs = original_inputs
    else:
      raise ValueError(
          f'Expect input spec to be 4 or 5 dimensions {input_specs.shape}')
    if grouping[-2]:
      # Instead of loading optical flows as inputs from data pipeline, we are
      # applying the "Representation Flow" to RGB frames so that we can compute
      # the flow within TPU/GPU on fly. It's essentially optical flow since we
      # do it with RGBs.
      axis = 3 if data_format == 'channels_last' else 1
      flow_inputs = rf.RepresentationFlow(
          original_num_frames,
          depth=reshape_inputs.shape.as_list()[axis],
          num_iter=40,
          bottleneck=1)(
              reshape_inputs)
    streams = []

    for i in range(len(structure)):
      with tf.name_scope('Node_' + str(i)):
        if structure[i][0] == -1:

View on GitHub (pinned to e006f5f0d5)

When it happens

Trigger: Thrown at official/projects/assemblenet/modeling/assemblenet.py:809 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/03063e4b50817155. Report an issue: GitHub.