Chainer is an open source deep learning framework written in Python.
Chainer
A Powerful, Flexible, and Intuitive Framework for Neural Networks
For more info see chainer.org
Questions tagged [chainer]
194 questions
0
votes
1 answer
Does chainercv project the full copy of original faster R-CNN caffe code?
I find the chainercv project : https://github.com/chainer/chainercv
I have read this code, the comment and code is excellent, but few detail still confuse me. I wonder:
Does this code the exactly same logic as…
machen
- 283
- 2
- 10
0
votes
0 answers
How to implement existing word2vector into EmbedID?
I know the way to initialize the EmbedID randomly. But now I want to load existing w2v, like the Glove, and update it constantly with the training process. How can I do that?
Thanks for your help.
Georgie Ji
- 15
- 2
0
votes
1 answer
Composing functions in Chainer
I wish to create a new Function object in Chainer by composing a number of existing functions. I haven't been able to find any method to do this in the docs. I could implement the composite function directly, this might be computationally more…
James Griffin
- 153
- 9
0
votes
1 answer
Do we need to derive differential/gradient w.r.t. input data in backward function(chainer)?
I am implementing a very complex Function in my research, It use Belief Propagation in this layer. I have derived the gradient w.r.t. W(parameter) of this layer, But because its complex, I haven't derived the gradient w.r.t. input_data(the data come…
machen
- 283
- 2
- 10
0
votes
1 answer
Does slice or index of chainer.Variable to get item in chainer has backward ability?
Does the following code chainer.Variable still have ability to hold graph and can backward (gradient flow) after slice(a[0,1] or index(a[0]):
>>> a = chainer.Variable(np.array([[1,2,3],[10,11,12]]))
>>> a
variable([[ 1, 2, 3],
[10, 11,…
machen
- 283
- 2
- 10
0
votes
1 answer
What is argument "grad_outputs" in chainer's backward function
3 questions:
what is grad_outputs in chainer?
one example in chainer's function F.transpose, how to explain this backward code?
def backward(self, inputs, grad_outputs):
gy = grad_outputs[0]
inv_axes = self.axes
if self.axes:
…
machen
- 283
- 2
- 10
0
votes
1 answer
how to use NStepLSTM or BiNStepLSTM in chainer?
how to use NStepLSTM or BiNStepLSTM?
I see reference it must supply argument as list of variables, what is this list?
How to use it to composite of other layer Link??
machen
- 283
- 2
- 10
0
votes
1 answer
Why in chainer define self define function backward function must return gradient shape the same as input?
Two questions:
https://docs.chainer.org/en/stable/tutorial/function.html written: the backward function must return the same shape as the arguments of the forward method? because in some cases, the input data and parameter need not to be same shape…
machen
- 283
- 2
- 10
0
votes
1 answer
How to plugin already written C/C++ layer in chainer
I have already written C++ layer which is using CPU, I want to plugin chainer framework, How to do that ? Can chainer mix of CPU and GPU layers together?
machen
- 283
- 2
- 10
0
votes
1 answer
In chainer, How to write BPTT updater using multiple GPUs?
I don't find example because existing example only extends training.StandardUpdater, thus only use One GPU.
machen
- 283
- 2
- 10
0
votes
0 answers
how to write chainer Iterator has the ability the same as MultiprocessIterator functionality
I want to write the RNN/LSTM iterator the same as https://docs.chainer.org/en/stable/tutorial/recurrentnet.html example, but add multi process fetch ability the same as MultiprocessIterator functionality.
The url link's example as following:
from…
machen
- 283
- 2
- 10
0
votes
1 answer
How to implement naive batch gradient descent?
everyone. I have a question about the implement of gradient descent. I have found several optimizers, like ada_grad, adam, sgd and so on, they're perfect. But I'm attempting to implement the naive gradient method batch gradient descent that is with…
Georgie Ji
- 15
- 2
0
votes
1 answer
Warning: nvcc path != CUDA_PATH while installing chainer with sudo CUDA_PATH
I was running a code in orenmel, it is cool with CPU. But when I ran -train_context2vec.py- with GPU, I recieved：
Warning: nvcc path != CUDA_PATH
...
cuda.check_cuda_available()
RuntimeError: CUDA environment is not correctly set up, cannot…
Yg Meng
- 53
- 1
- 1
- 8
0
votes
1 answer
Chainer, why does my model expect **array** of images?
I've written a very simple, toy, chainer model:
class Upscale(chainer.Chain):
def __init__(self):
super(Upscale, self).__init__(
d1=L.Deconvolution2D(3, 40, 4, stride=2),
c1=L.Convolution2D(40, 3, 4)
)
def…
Jeffrey
- 11,063
- 1
- 21
- 42
0
votes
1 answer
IOError on train_mnist.py
When train_mnist.py is executed on Ubuntu 15.10 64bit and python 2.7, the error(IOError: [Errno 2] No such file or directory: 'result/cg.dot') was occurred below. Let me know how to solve it.
survey@SURVEY-C11:~/chainer/examples/mnist$ python…
user7452691
- 23
- 5