ValueError: Negative dimension size caused by subtracting 2 from 1 for 'max_pooling2d_1/MaxPool'

Question

ValueError: Negative dimension size caused by subtracting 2 from 1 for 'max_pooling2d_1/MaxPool'

Instassa opened this issue 7 years ago · comments

When I try to run: python dcgan.py --mode train --batch_size 100 I get the following:


Using TensorFlow backend.
dcgan.py:41: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(64, (5, 5), input_shape=(1, 28, 28..., padding="same")`
  input_shape=(1, 28, 28)))
Traceback (most recent call last):
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/common_shapes.py", line 671, in _call_cpp_shape_fn_impl
    input_tensors_as_shapes, status)
  File "/home/marija/anaconda3/envs/tensorflow-gpu-3.5/lib/python3.5/contextlib.py", line 66, in __exit__
    next(self.gen)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/errors_impl.py", line 466, in raise_exception_on_not_ok_status
    pywrap_tensorflow.TF_GetCode(status))
tensorflow.python.framework.errors_impl.InvalidArgumentError: Negative dimension size caused by subtracting 2 from 1 for 'max_pooling2d_1/MaxPool' (op: 'MaxPool') with input shapes: [?,1,28,64].

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "dcgan.py", line 169, in <module>
    train(BATCH_SIZE=args.batch_size)
  File "dcgan.py", line 82, in train
    discriminator = discriminator_model()
  File "dcgan.py", line 43, in discriminator_model
    model.add(MaxPooling2D(pool_size=(2, 2)))
  File "/home/marija/.local/lib/python3.5/site-packages/keras/models.py", line 466, in add
    output_tensor = layer(self.outputs[0])
  File "/home/marija/.local/lib/python3.5/site-packages/keras/engine/topology.py", line 585, in __call__
    output = self.call(inputs, **kwargs)
  File "/home/marija/.local/lib/python3.5/site-packages/keras/layers/pooling.py", line 154, in call
    data_format=self.data_format)
  File "/home/marija/.local/lib/python3.5/site-packages/keras/layers/pooling.py", line 217, in _pooling_function
    pool_mode='max')
  File "/home/marija/.local/lib/python3.5/site-packages/keras/backend/tensorflow_backend.py", line 3245, in pool2d
    x = tf.nn.max_pool(x, pool_size, strides, padding=padding)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/ops/nn_ops.py", line 1821, in max_pool
    name=name)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/ops/gen_nn_ops.py", line 1638, in _max_pool
    data_format=data_format, name=name)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/op_def_library.py", line 768, in apply_op
    op_def=op_def)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/ops.py", line 2338, in create_op
    set_shapes_for_outputs(ret)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/ops.py", line 1719, in set_shapes_for_outputs
    shapes = shape_func(op)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/ops.py", line 1669, in call_with_requiring
    return call_cpp_shape_fn(op, require_shape_fn=True)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/common_shapes.py", line 610, in call_cpp_shape_fn
    debug_python_shape_fn, require_shape_fn)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/common_shapes.py", line 676, in _call_cpp_shape_fn_impl
    raise ValueError(err.message)
ValueError: Negative dimension size caused by subtracting 2 from 1 for 'max_pooling2d_1/MaxPool' (op: 'MaxPool') with input shapes: [?,1,28,64].

Any ideas what is going wrong here?

ywy · Answer 1 · Fri Jun 02 2017 20:13:27 GMT+0800 (China Standard Time)

I also got these problem.......

Shiki · Answer 2 · Sun Jun 04 2017 19:31:55 GMT+0800 (China Standard Time)

This caused by the different version of Keras. You could do the following changes.

def discriminator_model():
    model = Sequential()
    model.add(Convolution2D(
                        64, 5, 5,
                        border_mode='same',
                        input_shape=(28, 28, 1))) #modify this variable

Goofy · Answer 3 · Sun Jun 11 2017 19:38:32 GMT+0800 (China Standard Time)

It doesn't work. After change input_shape to (28, 28, 1), got the following
Traceback (most recent call last): File "dcgan.py", line 167, in <module> train(BATCH_SIZE=args.batch_size) File "dcgan.py", line 105, in train X = np.concatenate((image_batch, generated_images)) ValueError: all the input array dimensions except for the concatenation axis must match exactly

Shiki · Answer 4 · Sun Jun 11 2017 23:30:16 GMT+0800 (China Standard Time)

The default parameter of 'data_format' for Conv2d layer and pooling layer is 'channel_last', see [Keras doc] While the input data format is 'channel_first', it causes the conflict. You can either change the parameters, or change the input shape.

This is my version and it works fine in my environment( Keras==2.0.4 && Python==3.6).

ywy · Answer 5 · Fri Jun 16 2017 16:39:45 GMT+0800 (China Standard Time)

I just modified the code to fit tensorflow backend, you can try to use this one.
This is my code(Keras==2.0.4 && Python==3.5)

ItsArghya · Answer 6 · Thu Jun 29 2017 11:08:16 GMT+0800 (China Standard Time)

Hi,
An easy fix is:

Open ~/.keras/keras.json file. In Ubuntu do: vi ~/.keras/keras.json
change your input data format from channels_last to channels_first
save it.
Run: python dcgan.py --mode train --batch_size 100

panpan · Answer 7 · Wed Oct 11 2017 19:42:13 GMT+0800 (China Standard Time)

I also got these problem.......，have you solved?

agcala · Answer 8 · Sat Dec 23 2017 13:15:34 GMT+0800 (China Standard Time)

You need to change the order of the data. Tensorflow uses [samples][height][width][channels] order, while Theano is in reverse order.

agcala · Answer 9 · Sat Dec 23 2017 13:37:14 GMT+0800 (China Standard Time)

The shape of the input should have the channel last as you are using Tensorflow.

agcala · Answer 10 · Sat Dec 23 2017 13:43:00 GMT+0800 (China Standard Time)

@MacwinWin You forgot to add the batch size which needs to be the first element.

Krishna Prasad · Answer 11 · Wed May 13 2020 17:00:36 GMT+0800 (China Standard Time)

When I try to run: python dcgan.py --mode train --batch_size 100 I get the following:


Using TensorFlow backend.
dcgan.py:41: UserWarning: Update your `Conv2D` call to the Keras 2 API: `Conv2D(64, (5, 5), input_shape=(1, 28, 28..., padding="same")`
  input_shape=(1, 28, 28)))
Traceback (most recent call last):
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/common_shapes.py", line 671, in _call_cpp_shape_fn_impl
    input_tensors_as_shapes, status)
  File "/home/marija/anaconda3/envs/tensorflow-gpu-3.5/lib/python3.5/contextlib.py", line 66, in __exit__
    next(self.gen)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/errors_impl.py", line 466, in raise_exception_on_not_ok_status
    pywrap_tensorflow.TF_GetCode(status))
tensorflow.python.framework.errors_impl.InvalidArgumentError: Negative dimension size caused by subtracting 2 from 1 for 'max_pooling2d_1/MaxPool' (op: 'MaxPool') with input shapes: [?,1,28,64].

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "dcgan.py", line 169, in <module>
    train(BATCH_SIZE=args.batch_size)
  File "dcgan.py", line 82, in train
    discriminator = discriminator_model()
  File "dcgan.py", line 43, in discriminator_model
    model.add(MaxPooling2D(pool_size=(2, 2)))
  File "/home/marija/.local/lib/python3.5/site-packages/keras/models.py", line 466, in add
    output_tensor = layer(self.outputs[0])
  File "/home/marija/.local/lib/python3.5/site-packages/keras/engine/topology.py", line 585, in __call__
    output = self.call(inputs, **kwargs)
  File "/home/marija/.local/lib/python3.5/site-packages/keras/layers/pooling.py", line 154, in call
    data_format=self.data_format)
  File "/home/marija/.local/lib/python3.5/site-packages/keras/layers/pooling.py", line 217, in _pooling_function
    pool_mode='max')
  File "/home/marija/.local/lib/python3.5/site-packages/keras/backend/tensorflow_backend.py", line 3245, in pool2d
    x = tf.nn.max_pool(x, pool_size, strides, padding=padding)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/ops/nn_ops.py", line 1821, in max_pool
    name=name)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/ops/gen_nn_ops.py", line 1638, in _max_pool
    data_format=data_format, name=name)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/op_def_library.py", line 768, in apply_op
    op_def=op_def)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/ops.py", line 2338, in create_op
    set_shapes_for_outputs(ret)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/ops.py", line 1719, in set_shapes_for_outputs
    shapes = shape_func(op)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/ops.py", line 1669, in call_with_requiring
    return call_cpp_shape_fn(op, require_shape_fn=True)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/common_shapes.py", line 610, in call_cpp_shape_fn
    debug_python_shape_fn, require_shape_fn)
  File "/home/marija/.local/lib/python3.5/site-packages/tensorflow/python/framework/common_shapes.py", line 676, in _call_cpp_shape_fn_impl
    raise ValueError(err.message)
ValueError: Negative dimension size caused by subtracting 2 from 1 for 'max_pooling2d_1/MaxPool' (op: 'MaxPool') with input shapes: [?,1,28,64].

Any ideas what is going wrong here?

The problem you have is very simple. The MaxPooling2D layer in Tf uses padding='valid' as default. When the input shape is defined, model is built alongside checking the architecture, in some cases the architecture raises a computing problem (when few functions are mathematically incooperative).

To solve the problem, you need not change the input shape(HxWxchannel - 224, 224, 1), but just go over to the code line and change the parameter padding='valid' to padding='same'. If there is no padding='valid in your code, then add it as below.

model.add(MaxPooling2D((2,2), strides=(2,2), padding='same'))

more info on the topic can be found here - https://www.tensorflow.org/api_docs/python/tf/keras/layers/MaxPool2D