bobchennan

import torch
import torch.nn as nn
import torch.nn.functional as F

# helpers

def make_unit_length(x, epsilon=1e-6):
    norm = x.norm(p=2, dim=-1, keepdim=True)
    return x.div(norm + epsilon)

bobchennan

def main(text_file, out_file):
    with open(text_file) as f:
        temp = f.readlines()
    content = [x.strip() for x in temp]

    with open(out_file, 'w') as f:
        for i in content:
            if i.split()[4] == 'bonafide':
                f.write('%s %s\n' % (i.split()[0] + '-' + i.split()[1], i.split()[4]))
            else:

bobchennan

import torch
from torch.autograd import Function

class GELS(Function):
    """ Efficient implementation of gels from
        Nanxin Chen
        [email protected]
    """
    @staticmethod
    def forward(ctx, A, b):

bobchennan

import numpy as np
#from sklearn.discriminant_analysis import _class_cov,LinearDiscriminantAnalysis
from discriminant_analysis import _class_cov,LinearDiscriminantAnalysis

X = np.random.rand(1000000, 100)
y = np.random.randint(0, 20000, (1000000, ))
print len(np.unique(y))
#model=LinearDiscriminantAnalysis(solver='eigen')
#model.fit(X,y)
_class_cov(X, y, np.array([1.]*20000)/20000)

bobchennan

import keras.backend as K
from keras import initializers
from keras.engine import InputSpec
from keras.layers import Dense, Lambda, Wrapper


class ConcreteDropout(Wrapper):
    def __init__(self, layer, weight_regularizer=1e-6, dropout_regularizer=1e-5, **kwargs):
        assert 'kernel_regularizer' not in kwargs
        super(ConcreteDropout, self).__init__(layer, **kwargs)

bobchennan

# The data can be found in https://github.com/tdhopper/topic-modeling-datasets/blob/master/README.md
# The stop word lists can be found in https://github.com/stanfordnlp/CoreNLP/blob/master/data/edu/stanford/nlp/patterns/surface/stopwords.txt
# Data clean in http://www.mjdenny.com/Text_Processing_In_R.html
library(stringr)
library(gtools)

createDataFile = function(){
processFile = function(filepath) {
  con = file(filepath, "r")
  inAbstract = FALSE

bobchennan

from multiprocessing import Pipe, Process, Manager
from time import sleep
import sharedmem

batches = Manager().dict()

def train_process(model, num_batches, save_path, q):
    cnt   = 0
    
    while True:

bobchennan

var ldap = require('ldapjs'),
    mysql = require("mysql"),
    server = ldap.createServer(),
    addrbooks = {}, userinfo = {},
    ldap_port = 389,
    basedn = "dc=example, dc=com",
    company = "Example",
    db = mysql.createClient({
      user: "abook",
      password: "abook",