I am just starting to learn cntk. However, I have a basic question that is holding me back from progressing. I have the following test that passes:
import numpy as np
from cntk import input_variable, plus
def test_simple(self):
x_input = np.asarray([[1, 2, 2]], dtype=np.int64)
assert (1, 3) == x_input.shape
y_input = np.asarray([[5, 3, 3]], dtype=np.int64)
assert (1, 3) == y_input.shape
x = input_variable(x_input.shape[1])
assert (3, ) == x.shape
y = input_variable(y_input.shape[1])
assert (3, ) == y.shape
x_plus_y = plus(x, y)
assert (3, ) == x_plus_y.shape
res = x_plus_y.eval({x: x_input, y: y_input})
assert 6 == res[0, 0, 0]
assert 5 == res[0, 0, 1]
assert 5 == res[0, 0, 2]
I understand that the shape of the output is (1, 1, 3) as the first and second axis are the batch and default dynamic axis respectively.
However, why do I need to set the shape of the input variables as (3,) instead of (1, 3). Using (1, 3) fails.
Why is there an inconsistency between the shape of the input node in the graph and the numpy data used as input to that node?
Thank you,
Paddy
This is explained a little bit in the description of "arguments" for Function.forward. Another description is here. The reason for your confusion probably is that CNTK does some "helpful" conversions.
If you specify your input as (1,3) then you need to provide a list of (1,3) arrays in case of a minibatch without a sequence axis or a list of (x,1,3) arrays in case of a minibatch with a sequence axis (where x is potentially different for each sequence in the minibatch). Similarly, if you specify an input as (3,) then you need to either provide a list of (3,) vectors or a list of (x,3) vectors.
The confusion probably arises from the case when a list is not provided. In that case CNTK iterates over the leading axis of the provided tensor and creates a list out of those elements e.g. a (5,1,3) tensor becomes a batch of 5 elements each having a shape of (1,3).
Related
I would like to whiten each image in a batch. The code I have to do so is this:
def whiten(self, x):
shape = x.shape
x = K.batch_flatten(x)
mn = K.mean(x, 0)
std = K.std(x, 0) + K.epsilon()
r = (x - mn) / std
r = K.reshape(x, (-1,shape[1],shape[2],shape[3]))
return r
#
where x is (?, 320,320,1). I am not keen on the reshape function with a -1 arg. Is there a cleaner way to do this?
Let's see what the -1 does. From the Tensorflow documentation (Because the documentation from Keras is scarce compared to the one from Tensorflow):
If one component of shape is the special value -1, the size of that dimension is computed so that the total size remains constant.
So what this means:
from keras import backend as K
X = tf.constant([1,2,3,4,5])
K.reshape(X, [-1, 5])
# Add one more dimension, the number of columns should be 5, and keep the number of elements to be constant
# [[1 2 3 4 5]]
X = tf.constant([1,2,3,4,5,6])
K.reshape(X, [-1, 3])
# Add one more dimension, the number of columns should be 3
# For the number of elements to be constant the number of rows should be 2
# [[1 2 3]
# [4 5 6]]
I think it is simple enough. So what happens in your code:
# Let's assume we have 5 images, 320x320 with 3 channels
X = tf.ones((5, 320, 320, 3))
shape = X.shape
# Let's flat the tensor so we can perform the rest of the computation
flatten = K.batch_flatten(X)
# What this did is: Turn a nD tensor into a 2D tensor with same 0th dimension. (Taken from the documentation directly, let's see that below)
flatten.shape
# (5, 307200)
# So all the other elements were squeezed in 1 dimension while keeping the batch_size the same
# ...The rest of the stuff in your code is executed here...
# So we did all we wanted and now we want to revert the tensor in the shape it had previously
r = K.reshape(flatten, (-1, shape[1],shape[2],shape[3]))
r.shape
# (5, 320, 320, 3)
Besides, I can't think of a cleaner way to do what you want to do. If you ask me, your code is already clear enough.
I'm having some trouble porting some code over from tensorflow to pytorch.
So I have a matrix with dimensions 10x30 representing 10 examples each with 30 features. Then I have another matrix with dimensions 10x5 containing indices of the the 5 closest examples for each examples in the first matrix. I want to 'gather' using the indices contained in the second matrix the 5 closet examples for each example in the first matrix leaving me with a 3d tensor of shape 10x5x30.
In tensorflow this is done with tf.gather(matrix1, matrix2). Does anyone know how i could do this in pytorch?
How about this?
matrix1 = torch.randn(10, 30)
matrix2 = torch.randint(high=10, size=(10, 5))
gathered = matrix1[matrix2]
It uses the trick of indexing with an array of integers.
I had a scenario where I had to apply gather() on an array of integers.
Exam-01
torch.Tensor().gather(dim, input_tensor)
# here,
# input_tensor -> tensor(1)
my_list = [0, 1, 2, 3, 4]
my_tensor = torch.IntTensor(my_list)
output = my_tensor.gather(0, input_tensor) # 0 -> is the dimension
Exam-02
torch.gather(param_tensor, dim, input_tensor)
# here,
# input_tensor -> tensor(1)
my_list = [0, 1, 2, 3, 4]
my_tensor = torch.IntTensor(my_list)
output = torch.gather(my_tensor, 0, input_tensor) # 0 -> is the dimension
I'm trying to writting a layer to merge 2 tensors with such a formula
The shapes of x[0] and x[1] are both (?, 1, 500).
M is a 500*500 Matrix.
I want the output to be (?, 500, 500) which is theoretically feasible in my opinion. The layer will output (1,500,500) for every pair of inputs, as (1, 1, 500) and (1, 1, 500). As the batch_size is variable, or dynamic, the output must be (?, 500, 500).
However, I know little about axes and I have tried all the combinations of axes but it doesn't make sense.
I try with numpy.tensordot and keras.backend.batch_dot(TensorFlow). If the batch_size is fixed, taking a =
(100,1,500) for example, batch_dot(a,M,(2,0)), the output can be (100,1,500).
Newbie for Keras, sorry for such a stupid question but I have spent 2 days to figure out and it drove me crazy :(
def call(self,x):
input1 = x[0]
input2 = x[1]
#self.M is defined in build function
output = K.batch_dot(...)
return output
Update:
Sorry for being late. I try Daniel's answer with TensorFlow as Keras's backend and it still raises a ValueError for unequal dimensions.
I try the same code with Theano as backend and now it works.
>>> import numpy as np
>>> import keras.backend as K
Using Theano backend.
>>> from keras.layers import Input
>>> x1 = Input(shape=[1,500,])
>>> M = K.variable(np.ones([1,500,500]))
>>> firstMul = K.batch_dot(x1, M, axes=[1,2])
I don't know how to print tensors' shape in theano. It's definitely harder than tensorflow for me... However it works.
For that I scan 2 versions of codes for Tensorflow and Theano. Following are differences.
In this case, x = (?, 1, 500), y = (1, 500, 500), axes = [1, 2]
In tensorflow_backend:
return tf.matmul(x, y, adjoint_a=True, adjoint_b=True)
In theano_backend:
return T.batched_tensordot(x, y, axes=axes)
(If following changes of out._keras_shape don't make influence on out's value.)
Your multiplications should select which axes it uses in the batch dot function.
Axis 0 - the batch dimension, it's your ?
Axis 1 - the dimension you say has length 1
Axis 2 - the last dimension, of size 500
You won't change the batch dimension, so you will use batch_dot always with axes=[1,2]
But for that to work, you must ajust M to be (?, 500, 500).
For that define M not as (500,500), but as (1,500,500) instead, and repeat it in the first axis for the batch size:
import keras.backend as K
#Being M with shape (1,500,500), we repeat it.
BatchM = K.repeat_elements(x=M,rep=batch_size,axis=0)
#Not sure if repeating is really necessary, leaving M as (1,500,500) gives the same output shape at the end, but I haven't checked actual numbers for correctness, I believe it's totally ok.
#Now we can use batch dot properly:
firstMul = K.batch_dot(x[0], BatchM, axes=[1,2]) #will result in (?,500,500)
#we also need to transpose x[1]:
x1T = K.permute_dimensions(x[1],(0,2,1))
#and the second multiplication:
result = K.batch_dot(firstMul, x1T, axes=[1,2])
I prefer using TensorFlow so I tried to figure it out with TensorFlow in past few days.
The first one is much similar to Daniel's solution.
x = tf.placeholder('float32',shape=(None,1,3))
M = tf.placeholder('float32',shape=(None,3,3))
tf.matmul(x, M)
# return: <tf.Tensor 'MatMul_22:0' shape=(?, 1, 3) dtype=float32>
It needs to feed values to M with fit shapes.
sess = tf.Session()
sess.run(tf.matmul(x,M), feed_dict = {x: [[[1,2,3]]], M: [[[1,2,3],[0,1,0],[0,0,1]]]})
# return : array([[[ 1., 4., 6.]]], dtype=float32)
Another way is simple with tf.einsum.
x = tf.placeholder('float32',shape=(None,1,3))
M = tf.placeholder('float32',shape=(3,3))
tf.einsum('ijk,lm->ikl', x, M)
# return: <tf.Tensor 'MatMul_22:0' shape=(?, 1, 3) dtype=float32>
Let's feed some values.
sess.run(tf.einsum('ijk,kl->ijl', x, M), feed_dict = {x: [[[1,2,3]]], M: [[1,2,3],[0,1,0],[0,0,1]]})
# return: array([[[ 1., 4., 6.]]], dtype=float32)
Now M is a 2D tensor and no need to feed batch_size to M.
What's more, now it seems such a question can be solved in TensorFlow with tf.einsum. Does it mean it's a duty for Keras to invoke tf.einsum in some situations? At least I find no where Keras calls tf.einsum. And in my opinion, when batch_dot 3D tensor and 2D tensor Keras behaves weirdly. In Daniel's answer, he pads M to (1,500,500) but in K.batch_dot() M will be adjusted to (500,500,1) automatically. I find tf will adjust it with Broadcasting rules and I'm not sure Keras does the same.
I am a pytorch user. I have got a pretrained model in tensorflow and I would like to transfer it into pytorch. In one part of model architecture, I mean in tensorflow-defined model, there is a function tf.space_to_depth which transfers an input size of (None, 38,38,64) to (None, 19,19, 256). (https://www.tensorflow.org/api_docs/python/tf/space_to_depth) is the doc of this function. But I could not understand what this function actually do. Could you please provide some numpy codes to illustrate it for me?
Actually I would like to make an exact similar layer in pytorch.
Some codes in tensorflow reveals another secret:
Here is some codes:
import numpy as np
import tensorflow as tf
norm = tf.random_normal([1, 2, 2, 1], mean=0, stddev=1)
trans = tf.space_to_depth(norm,2)
with tf.Session() as s:
norm = s.run(norm)
trans = s.run(trans)
print("Norm")
print(norm.shape)
for index,value in np.ndenumerate(norm):
print(value)
print("Trans")
print(trans.shape)
for index,value in np.ndenumerate(trans):
print(value)
And here is the output:
Norm
(1, 2, 2, 1)
0.695261
0.455764
1.04699
-0.237587
Trans
(1, 1, 1, 4)
1.01139
0.898777
0.210135
2.36742
As you can see above, In Addition to data reshaping, the tensor values has changed!
This tf.space_to_depth divides your input into blocs and concatenates them.
In your example the input is 38x38x64 (and I guess the block_size is 2). So the function divides your input into 4 (block_size x block_size) and concatenates them which gives your 19x19x256 output.
You just need to divide each of your channel (input) into block_size*block_size patches (each patch has a size of width/block_size x height/block_size) and concatenate all of these patches. Should be pretty straightforward with numpy.
Hope it helps.
Conclusion: tf.space_to_depth() only outputs a copy of the input tensor where values from the height and width dimensions are moved to the depth dimension.
If you modify your code a little bit, like this
norm = tf.random_normal([1, 2, 2, 1], mean=0, stddev=1)
with tf.Session() as s:
norm = s.run(norm)
trans = tf.space_to_depth(norm,2)
with tf.Session() as s:
trans = s.run(trans)
Then you will have the following results:
Norm
(1, 2, 2, 1)
-0.130227
2.04587
-0.077691
-0.112031
Trans
(1, 1, 1, 4)
-0.130227
2.04587
-0.077691
-0.112031
Hope this can help you.
A good reference for PyTorch is the implementation of the PixelShuffle module here. This shows the implementation of something equivalent to Tensorflow's depth_to_space. Based on that we can implement pixel_shuffle with a scaling factor less than 1 which would be like space_to_depth. E.g., downscale_factor=0.5 is like space_to_depth with block_size=2.
def pixel_shuffle_down(input, downscale_factor):
batch_size, channels, in_height, in_width = input.size()
out_channels = channels / (downscale_factor ** 2)
block_size = 1 / downscale_factor
out_height = in_height * downscale_factor
out_width = in_width * downscale_factor
input_view = input.contiguous().view(
batch_size, channels, out_height, block_size, out_width, block_size)
shuffle_out = input_view.permute(0, 1, 3, 5, 2, 4).contiguous()
return shuffle_out.view(batch_size, out_channels, out_height, out_width)
Note: I haven't verified this implementation yet and I'm not sure if it's exactly the inverse of pixel_shuffle but this is the basic idea. I've also opened an issue on the PyTorch Github about this here. In NumPy the equivalent code would use reshapeand transpose instead of view and permute respectively.
Using split and stack functions along with permute in Pytorch gives us the same result as space_to_depth in tensorflow does. Here is the code in Pytorch.
Assume that input is in BHWC format.
Based on block_size and input shape, we can caculate the output shape.
First, it splits the input on the "width" dimension or dimension #2 by block_size. The result of this operation is an array of length d_width. It's just like you cut a cake (by block_size) into d_width pieces.
Then for each piece, you reshape it so it has correct output height and output depth (channel). Finally, we stack those pieces together and perform a permutation.
Hope it helps.
def space_to_depth(input, block_size)
block_size_sq = block_size*block_size
(batch_size, s_height, s_width, s_depth) = input.size()
d_depth = s_depth * self.block_size_sq
d_width = int(s_width / self.block_size)
d_height = int(s_height / self.block_size)
t_1 = input.split(self.block_size, 2)
stack = [t_t.contiguous().view(batch_size, d_height, d_depth) for t_t in t_1]
output = torch.stack(stack, 1)
output = output.permute(0, 2, 1, 3)
return output
maybe this one works:
sudo apt install nvidia-cuda-toolkit
it worked for me.
So, here is what I want to do:
Right now, I have padding = 'SAME' for all of my neural net layers. I would like to make my code more generic, so I can build my nets with arbitrary paddings, and I don't want to have to calculate how big the output tensors of the layers of my net are. I would like to just access the dimension at initialization/run time, the way the tf.nn functions apparently do internally, so I can initialize my weight and bias tensors in the correct dimension...
So,
How do I access the "shape" function/object of the output placeholder of a convolution?
There are two kinds of shapes -- tensor.get_shape() which gives static shape computed by Python wrappers during Graph construction (whenever possible), and tf.shape(tensor) which is an op that can be executed during runtime to get shape of the tensor (always possible). Both of these work for convolutions.
a = tf.Variable(tf.ones((1, 3, 3, 1)))
b = tf.Variable(tf.ones((3, 3, 1, 1)))
c = tf.nn_ops.conv2d(a, b, [1, 1, 1, 1], padding="VALID")
sess = create_session()
sess.run(tf.initialize_all_variables())
print c.get_shape()
print sess.run(tf.shape(c))
This gives
(1, 1, 1, 1)
[1 1 1 1]