bnsh

#! /usr/bin/env python3

import re
from collections import defaultdict

def sort_word(word):
	letters = "".join(sorted([w for w in word]))
	return letters

def read_dictionary():

bnsh

#! /usr/bin/env Rscript

library(R6)

Cloaked <- R6::R6Class("Cloaked",
  public = list(
    msg = function() {
      return(paste(
        "this code _works_, but lintr (https://github.com/jimhester/lintr)",
        "complains that cloak_class.R:19:8: warning: no visible binding for",

bnsh

#! /usr/bin/python

"""Why is len(grad_input) == 3 if I use Identity, but
          len(grad_input) == 1 if I use ReLU?

   I don't know if this is the source of all my woes in my real network.

Run this like so:
	python demo.py identity
		module=Identity() len(grad_input)=3 len(grad_output)=1

bnsh

#! /usr/bin/python

"""BinaryStochastic interprets it's input as a probability
between [0,1] and
 * with probability p, outputs a 1, and
 * with probability (1-p) it outputs a 0.

Practically, this is accomplished simply by taking a random uniform
the size of the input and seeing if the random uniform is less than
the input. If true, we output a 1, if false we output a 0.

bnsh

#! /usr/bin/python

from timer_util import Timer

def fibonacci(n):
	if n == 0:
		return 0
	elif n == 1:
		return 1
	else:

bnsh

#! /usr/bin/python

"""This will preprocess the cats data.
   First, it splits the train data into training and validation.
   validation is left unmolested. (Other than scaling and cropping.)
   train tho, is further cropped and rotated randomly."""

import os
import re
import math

bnsh

{
			"client_key":"knunopkiantijnaGlok1",
			"client_secret": "iddyadvufDijSeujowfEbphihythelkekiakphyeg",
			"access_token": "yufBeokbytdiVoggUrlEk8",
			"access_token_secret": "UtgazatsectyamHadNumpashawsOltirejcickmins"
}

bnsh

require 'nn'
require 'optim'


netG = nn.Sequential()
ngf = 32
netG:add(nn.SpatialConvolution(1, ngf, 4,4,   2,2,   1,1)) 
netG:add(nn.ReLU(true))
netG:add(nn.SpatialConvolution(ngf, ngf * 2,  4,4,   2,2,   1,1))
netG:add(nn.SpatialBatchNormalization(ngf * 2))

bnsh

#! /usr/local/torch/install/bin/th

local TempDelhiCriterion, parent = torch.class('nn.TempDelhiCriterion', 'nn.MSECriterion')

function TempDelhiCriterion:__init()
	parent.__init(self)
	self.inputcpy = torch.Tensor()
end

function TempDelhiCriterion:forward(input, target)

bnsh

#! /usr/local/torch/install/bin/th

require "nn"
require "cutorch"
require "cunn"
require "cudnn"

function main()
	torch.manualSeed(12345)