Yuta Koreeda koreyou

## sequence_mask.py
def sequence_mask(x, length, value=0.):
    xp = cuda.get_array_module(length.data)
    # create permutation on (length.ndim + 1) dimension and expand dims until it has shame rank as x
    perms = xp.arange(x.shape[length.ndim]).reshape(
        [1] * length.ndim + [-1] + [1] * (x.ndim - length.ndim -1))
    length = length.reshape([1] * (length.ndim - 1) + [-1] + [1] * (x.ndim - length.ndim))
    pad = xp.ones_like(x) * value
    mask = xp.broadcast_to(perms, x.shape) < length
    return F.where(mask, x, pad)

## reproduce_bug.py
import chainer
import chainer.functions as F
import chainer.links as L
from chainer import training
from chainer.training import extensions

# Network definition
class MLP(chainer.Chain):
    def __init__(self, n_out):
        super(MLP, self).__init__()

## chainer_reporting_demo.ipynb

      
              1 file
            
          
              0 forks
            
          
              0 comments
            
          
              0 stars
            
          
                koreyou
                / chainer_reporting_demo.ipynb
            
            
              Created
              June 26, 2017 14:29
            
              
                Demostration of how we can use chainer's reporing mechanism to monitor variables' metrics
              
          
      Sorry, something went wrong. Reload?
      Sorry, we cannot display this file.
      Sorry, this file is invalid so it cannot be displayed.
      
          Viewer requires iframe.
      
    
## file0.txt
###### Editor debris ########

*.org
*~
¥#*

# spyder
.spyderproject/

# pycharm

## extreme_learning_machine.py
# Written by Yuta Koreeda
# CC-BY
import numpy as np


class ExtremeLearningMachine(object):
    def __init__(self, n_unit, activation=None):
        self._activation = self._sig if activation is None else activation
        self._n_unit = n_unit
	def sequence_mask(x, length, value=0.):
	xp = cuda.get_array_module(length.data)
	# create permutation on (length.ndim + 1) dimension and expand dims until it has shame rank as x
	perms = xp.arange(x.shape[length.ndim]).reshape(
	[1] * length.ndim + [-1] + [1] * (x.ndim - length.ndim -1))
	length = length.reshape([1] * (length.ndim - 1) + [-1] + [1] * (x.ndim - length.ndim))
	pad = xp.ones_like(x) * value
	mask = xp.broadcast_to(perms, x.shape) < length
	return F.where(mask, x, pad)
	import chainer
	import chainer.functions as F
	import chainer.links as L
	from chainer import training
	from chainer.training import extensions

	# Network definition
	class MLP(chainer.Chain):
	def __init__(self, n_out):
	super(MLP, self).__init__()
	###### Editor debris ########

	*.org
	*~
	¥#*

	# spyder
	.spyderproject/

	# pycharm
	# Written by Yuta Koreeda
	# CC-BY
	import numpy as np


	class ExtremeLearningMachine(object):
	def __init__(self, n_unit, activation=None):
	self._activation = self._sig if activation is None else activation
	self._n_unit = n_unit