Error due to a negative dimension in C1W4_Assignment

Course Q&A TensorFlow: Advanced Techniques Specialization Custom Models, Layers and Loss Functions with TF
1

In the (optional) training part of C1W4_Assignment, I get an error due to some…

Negative dimension size caused by subtracting 2 from 1.

What is this due to? All my test passed just fine. I even tried to account for the case where the number of classes was two and thus encoded as booleans, in which case I used a sigmoid activation and a binary cross-entropy.

The full error is as follows:

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-61-a97d1c255550> in <module>
     17 
     18 # Train the custom VGG model
---> 19 vgg.fit(dataset, epochs=10)

/opt/conda/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training.py in fit(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, validation_freq, max_queue_size, workers, use_multiprocessing, **kwargs)
    817         max_queue_size=max_queue_size,
    818         workers=workers,
--> 819         use_multiprocessing=use_multiprocessing)
    820 
    821   def evaluate(self,

/opt/conda/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training_v2.py in fit(self, model, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, validation_freq, max_queue_size, workers, use_multiprocessing, **kwargs)
    233           max_queue_size=max_queue_size,
    234           workers=workers,
--> 235           use_multiprocessing=use_multiprocessing)
    236 
    237       total_samples = _get_total_number_of_samples(training_data_adapter)

/opt/conda/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training_v2.py in _process_training_inputs(model, x, y, batch_size, epochs, sample_weights, class_weights, steps_per_epoch, validation_split, validation_data, validation_steps, shuffle, distribution_strategy, max_queue_size, workers, use_multiprocessing)
    591         max_queue_size=max_queue_size,
    592         workers=workers,
--> 593         use_multiprocessing=use_multiprocessing)
    594     val_adapter = None
    595     if validation_data:

/opt/conda/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training_v2.py in _process_inputs(model, mode, x, y, batch_size, epochs, sample_weights, class_weights, shuffle, steps, distribution_strategy, max_queue_size, workers, use_multiprocessing)
    704       max_queue_size=max_queue_size,
    705       workers=workers,
--> 706       use_multiprocessing=use_multiprocessing)
    707 
    708   return adapter

/opt/conda/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/data_adapter.py in __init__(self, x, y, sample_weights, standardize_function, **kwargs)
    700 
    701     if standardize_function is not None:
--> 702       x = standardize_function(x)
    703 
    704     # Note that the dataset instance is immutable, its fine to reusing the user

/opt/conda/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training_v2.py in standardize_function(dataset)
    658         model.sample_weight_mode = getattr(model, 'sample_weight_mode', None)
    659 
--> 660       standardize(dataset, extract_tensors_from_dataset=False)
    661 
    662       # Then we map using only the tensor standardization portion.

/opt/conda/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training.py in _standardize_user_data(self, x, y, sample_weight, class_weight, batch_size, check_steps, steps_name, steps, validation_split, shuffle, extract_tensors_from_dataset)
   2344     # First, we build the model on the fly if necessary.
   2345     if not self.inputs:
-> 2346       all_inputs, y_input, dict_inputs = self._build_model_with_inputs(x, y)
   2347       is_build_called = True
   2348     else:

/opt/conda/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training.py in _build_model_with_inputs(self, inputs, targets)
   2570     else:
   2571       cast_inputs = inputs
-> 2572     self._set_inputs(cast_inputs)
   2573     return processed_inputs, targets, is_dict_inputs
   2574 

/opt/conda/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training.py in _set_inputs(self, inputs, outputs, training)
   2657           kwargs['training'] = training
   2658       try:
-> 2659         outputs = self(inputs, **kwargs)
   2660       except NotImplementedError:
   2661         # This Model or a submodel is dynamic and hasn't overridden

/opt/conda/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/base_layer.py in __call__(self, inputs, *args, **kwargs)
    771                     not base_layer_utils.is_in_eager_or_tf_function()):
    772                   with auto_control_deps.AutomaticControlDependencies() as acd:
--> 773                     outputs = call_fn(cast_inputs, *args, **kwargs)
    774                     # Wrap Tensors in `outputs` in `tf.identity` to avoid
    775                     # circular dependencies.

/opt/conda/lib/python3.7/site-packages/tensorflow_core/python/autograph/impl/api.py in wrapper(*args, **kwargs)
    235       except Exception as e:  # pylint:disable=broad-except
    236         if hasattr(e, 'ag_error_metadata'):
--> 237           raise e.ag_error_metadata.to_exception(e)
    238         else:
    239           raise

ValueError: in converted code:

    <ipython-input-57-f62868bf82fd>:34 call  *
        x = self.block_c(x)
    /opt/conda/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/base_layer.py:773 __call__
        outputs = call_fn(cast_inputs, *args, **kwargs)
    <ipython-input-27-648bf33d2825>:39 call  *
        max_pool = self.max_pool(x)
    /opt/conda/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/base_layer.py:773 __call__
        outputs = call_fn(cast_inputs, *args, **kwargs)
    /opt/conda/lib/python3.7/site-packages/tensorflow_core/python/keras/layers/pooling.py:248 call
        data_format=conv_utils.convert_data_format(self.data_format, 4))
    /opt/conda/lib/python3.7/site-packages/tensorflow_core/python/ops/nn_ops.py:3875 max_pool
        name=name)
    /opt/conda/lib/python3.7/site-packages/tensorflow_core/python/ops/gen_nn_ops.py:5200 max_pool
        data_format=data_format, name=name)
    /opt/conda/lib/python3.7/site-packages/tensorflow_core/python/framework/op_def_library.py:742 _apply_op_helper
        attrs=attr_protos, op_def=op_def)
    /opt/conda/lib/python3.7/site-packages/tensorflow_core/python/framework/func_graph.py:595 _create_op_internal
        compute_device)
    /opt/conda/lib/python3.7/site-packages/tensorflow_core/python/framework/ops.py:3322 _create_op_internal
        op_def=op_def)
    /opt/conda/lib/python3.7/site-packages/tensorflow_core/python/framework/ops.py:1786 __init__
        control_input_ops)
    /opt/conda/lib/python3.7/site-packages/tensorflow_core/python/framework/ops.py:1622 _create_c_op
        raise ValueError(str(e))

    ValueError: Negative dimension size caused by subtracting 2 from 1 for 'my_vgg_9/block_55/max_pooling2d_55/MaxPool' (op: 'MaxPool') with input shapes: [?,1,1,256].


​
2

Hi Amine,

You have to learn to trace the error from the error output eventually. Ok so from your error output

Blockquote
ValueError: Negative dimension size caused by subtracting 2 from 1 for ‘my_vgg_9/block_55/max_pooling2d_55/MaxPool’ (op: ‘MaxPool’) with input shapes: [?,1,1,256]. x = self.block_c(x)

So it tells you there is an issue in the x = self.block_c(x) and specifically the (op: ‘MaxPool’) with input shapes: [?,1,1,256]., there should be something wrong with that shape.

Assuming that class Block is built correctly then you need to check self.block_c parameters that they are correctly implemented.

3

hi gent,

I am also getting attribute error to the collaborative vgg network. although submission of the assignment I cleared. it says vgg network has error.

Can you please guide me through where I am going wrong. n I am sharing the screenshot of the error.
Thank you
DP

Screenshot 1945-02-21 at 1.03.56 PM
Screenshot 1945-02-21 at 1.03.56 PM1280×800 255 KB

4

Hi I think the problem here is in using multiple processing units. I dont know where it specifically could be but I would guess some changes need to be made either in the code or in the settings for the code to use many processing units.

5

ok if the coding has error how can the vgg network pass all the test in the previous block. I have followed as per the instructions given. Processing unit part I am unable to understand as I am not a IT person but healthcare related professional. can you elaborate about the processing part mistake if you know.

Thank you again
DP

6

Hi, I can not give you exact or definitive instructions but what I can tell you is maybe try and use some of the knowledge on distributive training from the courses. Maybe some one else can help in more detail.