Python 3.5.2 (v3.5.2:4def2a2901a5, Jun 25 2016, 22:18:55) [MSC v.1900 64 bit (AMD64)] on win32
Type "copyright", "credits" or "license()" for more information.
>>> import word2vec_basic
Found and verified text8.zip
Data size 17005207
Most common words (+UNK) [['UNK', 418391], ('the', 1061396), ('of', 593677), ('and', 416629), ('one', 411764)]
Sample data [5243, 3081, 12, 6, 195, 2, 3135, 46, 59, 156] ['anarchism', 'originated', 'as', 'a', 'term', 'of', 'abuse', 'first', 'used', 'against']
3081 originated -> 5243 anarchism
3081 originated -> 12 as
12 as -> 3081 originated
12 as -> 6 a
6 a -> 12 as
6 a -> 195 term
195 term -> 2 of
195 term -> 6 a
Initialized
Average loss at step 0 : 275.96685791
Nearest to b: lim, mathbb, pron, sadd, postmodernism, yearning, interim, circumstance,
Nearest to s: astronomers, hallelujah, ona, heiress, sparkling, proverb, rulings, bartle,
Nearest to when: superpower, gaels, cutaway, novarum, ananda, geostationary, panthera, hypocrisy,
Nearest to seven: herbivorous, hyperplasia, kenyatta, ajanta, zadok, eternally, fairness, hine,
Nearest to of: mumbai, guidebook, arlington, phase, slowdown, palomar, hardcover, phonetics,
Nearest to system: getty, instructed, archers, beowulf, empowerment, arrears, grandsons, nicea,
Nearest to its: federico, bins, transducers, stanhope, range, freight, menai, vaduz,
Nearest to called: massed, desertification, doesn, morphology, monasteries, canceled, watering, lumpur,
Nearest to known: bolivia, banzer, humanism, adele, finnic, kwajalein, filtration, putting,
Nearest to will: horrors, fr, analysis, moravians, landslide, parenting, isomer, insulated,
Nearest to people: pathological, anagram, jonas, scenario, intercepts, guru, prequels, kirchhoff,
Nearest to nine: philosophy, dukes, trusting, szabo, contradicting, columba, citation, forks,
Nearest to also: jean, positive, articulated, serious, shepard, rabin, science, supplement,
Nearest to eight: response, amour, hissarlik, badminton, tuscany, heightened, ils, ashamed,
Nearest to are: jovian, provider, supervision, bosom, henslow, gimmicks, acute, burundi,
Nearest to all: robertson, mammoths, shapeshifting, mobilize, wasteful, nearing, kansas, resentment,
Average loss at step 2000 : 113.10768539
Average loss at step 4000 : 52.4376370575
Average loss at step 6000 : 33.8352457421
Average loss at step 8000 : 23.7887491972
Average loss at step 10000 : 18.0617327156
Nearest to b: gland, indians, lim, taliban, tezuka, circumstance, nine, rendering,
Nearest to s: and, condom, the, aarhus, UNK, holmes, of, gland,
Nearest to when: deposits, gland, geostationary, experimental, allowing, were, algebra, hypocrisy,
Nearest to seven: eight, analogue, zero, nine, six, reginae, gland, phi,
Nearest to of: and, in, for, from, with, the, agave, roper,
Nearest to system: psi, instructed, law, empowerment, saskatchewan, archaeology, celebrated, obligation,
Nearest to its: the, agave, their, range, a, kicking, aarhus, established,
Nearest to called: sadler, victoriae, mathbf, experimented, UNK, anthony, monasteries, doesn,
Nearest to known: bob, bolivia, phi, seer, music, helped, convention, humanism,
Nearest to will: fr, horrors, analysis, rfc, skill, situated, vogt, mya,
Nearest to people: reginae, perceived, music, jonas, september, married, pathological, scenario,
Nearest to nine: gland, zero, reginae, eight, gb, victoriae, cl, altenberg,
Nearest to also: jean, zionist, reginae, serious, crispin, probe, supplement, confusing,
Nearest to eight: six, nine, gland, zero, five, seven, reginae, phi,
Nearest to are: is, ba, kramnik, hoax, were, african, analogue, supervision,
Nearest to all: kansas, expanded, asterism, profession, complexity, references, robertson, represents,
Average loss at step 12000 : 13.7994806267
Average loss at step 14000 : 11.7659612741
Average loss at step 16000 : 9.8469510901
Average loss at step 18000 : 8.50730247939
Average loss at step 20000 : 7.85234803987
Nearest to b: lim, and, gland, circumstance, indians, tezuka, nine, pron,
Nearest to s: and, zero, holmes, the, or, birkenau, his, of,
Nearest to when: deposits, were, geostationary, and, gland, experimental, analogue, ananda,
Nearest to seven: eight, nine, zero, five, six, three, two, four,
Nearest to of: in, and, for, with, from, nine, eight, agave,
Nearest to system: psi, instructed, law, archers, UNK, cartier, nicea, empowerment,
Nearest to its: the, their, his, agave, a, absalom, aarhus, range,
Nearest to called: sadler, massed, UNK, monasteries, victoriae, experimented, pair, mathbf,
Nearest to known: dasyprocta, bob, bolivia, injuring, arg, phi, bug, hmong,
Nearest to will: fr, would, rfc, horrors, bosniaks, analysis, emerson, situated,
Nearest to people: reginae, perceived, scenario, odes, music, intercepts, anagram, pathological,
Nearest to nine: eight, six, seven, five, zero, four, dasyprocta, three,
Nearest to also: jean, zionist, which, crispin, amber, reginae, cth, confusing,
Nearest to eight: nine, five, six, zero, seven, three, four, two,
Nearest to are: is, were, was, kramnik, analogue, hoax, in, mathbf,
Nearest to all: expanded, kansas, rhenish, asterism, robertson, complexity, profession, represents,
Average loss at step 22000 : 7.24495614147
Average loss at step 24000 : 7.01978718054
Average loss at step 26000 : 6.66928812242
Average loss at step 28000 : 6.14945300984
Average loss at step 30000 : 6.17055390692
Nearest to b: and, gland, circumstance, lim, d, grants, landscapes, indians,
Nearest to s: and, zero, of, his, the, or, inches, six,
Nearest to when: deposits, speedup, and, geostationary, analogue, gland, experimental, were,
Nearest to seven: nine, eight, five, six, four, three, zero, two,
Nearest to of: in, and, for, from, s, nine, eight, iota,
Nearest to system: psi, empowerment, instructed, archers, cartier, law, nicea, obligation,
Nearest to its: their, the, his, a, agave, absalom, surroundings, amdahl,
Nearest to called: UNK, massed, sadler, primigenius, abitibi, victoriae, experimented, bagapsh,
Nearest to known: dasyprocta, adele, well, bob, used, seer, bolivia, injuring,
Nearest to will: would, fr, could, rfc, emerson, cpa, bosniaks, foam,
Nearest to people: reginae, odes, pathological, music, intercepts, scenario, perceived, guru,
Nearest to nine: eight, six, seven, five, four, three, zero, dasyprocta,
Nearest to also: which, zionist, crispin, sometimes, jean, cth, trinomial, reginae,
Nearest to eight: nine, six, five, seven, four, three, zero, abitibi,
Nearest to are: were, is, analogue, was, have, hoax, kramnik, anoa,
Nearest to all: rhenish, asterism, reuptake, kansas, expanded, dasyprocta, represents, profession,
Average loss at step 32000 : 5.86945372009
Average loss at step 34000 : 5.86404296362
Average loss at step 36000 : 5.67395866251
Average loss at step 38000 : 5.25235128129
Average loss at step 40000 : 5.48230646706
Nearest to b: UNK, circumstance, gland, grants, and, pron, d, landscapes,
Nearest to s: and, his, two, inches, holmes, the, or, birkenau,
Nearest to when: and, four, but, fielder, speedup, geostationary, deposits, were,
Nearest to seven: eight, six, five, four, nine, three, zero, one,
Nearest to of: in, from, for, and, abet, msg, eight, iota,
Nearest to system: psi, empowerment, instructed, cartier, archers, law, conflict, improved,
Nearest to its: their, the, his, a, agave, absalom, her, amdahl,
Nearest to called: UNK, massed, sadler, primigenius, abitibi, christiansen, victoriae, abet,
Nearest to known: used, adele, well, finnic, seer, dasyprocta, bolivia, bob,
Nearest to will: would, could, can, bosniaks, fr, may, rfc, to,
Nearest to people: reginae, odes, pathological, intercepts, music, coquitlam, scenario, perceived,
Nearest to nine: eight, seven, six, zero, five, four, three, dasyprocta,
Nearest to also: which, zionist, sometimes, crispin, generally, trinomial, reginae, cth,
Nearest to eight: nine, six, seven, five, four, zero, three, abitibi,
Nearest to are: were, is, have, was, analogue, absalon, angiotensin, kramnik,
Nearest to all: rhenish, asterism, reuptake, kansas, dasyprocta, many, expanded, any,
Average loss at step 42000 : 5.29408154821
Average loss at step 44000 : 5.32328894198
Average loss at step 46000 : 5.2740817008
Average loss at step 48000 : 5.040927809
Average loss at step 50000 : 5.12989223862
Nearest to b: gland, grants, circumstance, pron, six, d, abitibi, seven,
Nearest to s: zero, inches, his, and, nguni, pottery, recombine, vicarage,
Nearest to when: but, six, four, seven, speedup, deposits, gland, if,
Nearest to seven: eight, six, four, five, nine, three, zero, two,
Nearest to of: in, nine, for, and, from, thibetanus, reuptake, seven,
Nearest to system: psi, empowerment, instructed, cartier, archers, law, improved, conflict,
Nearest to its: their, the, his, agave, a, absalom, her, amdahl,
Nearest to called: massed, sadler, UNK, primigenius, naaman, abitibi, abet, adaptive,
Nearest to known: used, well, adele, seer, finnic, dasyprocta, epoxy, hmong,
Nearest to will: would, could, can, may, bosniaks, should, cpa, moravians,
Nearest to people: reginae, odes, coquitlam, music, pathological, intercepts, scenario, guru,
Nearest to nine: eight, seven, six, zero, four, five, three, dasyprocta,
Nearest to also: which, sometimes, zionist, thibetanus, generally, crispin, often, trinomial,
Nearest to eight: six, seven, nine, four, five, three, zero, dasyprocta,
Nearest to are: were, is, have, was, be, analogue, thibetanus, angiotensin,
Nearest to all: asterism, reuptake, two, dasyprocta, thibetanus, rhenish, many, expanded,
Average loss at step 52000 : 5.16474540925
Average loss at step 54000 : 5.10961878431
Average loss at step 56000 : 5.06780198526
Average loss at step 58000 : 5.11088050807
Average loss at step 60000 : 4.94124779272
Nearest to b: gland, microcebus, grants, d, circumstance, pron, abitibi, zero,
Nearest to s: his, zero, inches, and, michelob, recombine, vicarage, pottery,
Nearest to when: michelob, if, but, and, six, in, where, geostationary,
Nearest to seven: eight, six, five, four, nine, three, zero, two,
Nearest to of: for, in, microcebus, tamarin, thibetanus, and, abet, nine,
Nearest to system: empowerment, law, instructed, archers, microsite, tamarin, cartier, improved,
Nearest to its: their, the, his, tamarin, agave, her, absalom, ssbn,
Nearest to called: massed, sadler, tamarin, primigenius, michelob, naaman, abitibi, callithrix,
Nearest to known: used, well, adele, epoxy, finnic, microcebus, seer, hmong,
Nearest to will: would, could, can, may, should, to, moravians, bosniaks,
Nearest to people: reginae, odes, coquitlam, music, intercepts, pathological, cebus, saguinus,
Nearest to nine: eight, six, seven, five, four, zero, three, dasyprocta,
Nearest to also: which, sometimes, thibetanus, zionist, often, generally, tamarin, callithrix,
Nearest to eight: six, nine, seven, five, four, three, zero, two,
Nearest to are: were, is, have, angiotensin, be, kramnik, thibetanus, cebus,
Nearest to all: many, asterism, these, reuptake, two, thibetanus, dasyprocta, rhenish,
Average loss at step 62000 : 4.79670777971
Average loss at step 64000 : 4.79270891201
Average loss at step 66000 : 4.99029351902
Average loss at step 68000 : 4.88411666608
Average loss at step 70000 : 4.75195898664
Nearest to b: gland, grants, UNK, pron, d, seven, circumstance, microcebus,
Nearest to s: and, mitral, zero, inches, his, vicarage, michelob, holmes,
Nearest to when: if, michelob, but, before, where, was, during, six,
Nearest to seven: six, eight, five, four, nine, three, zero, one,
Nearest to of: for, in, microcebus, tamarin, same, iota, tabula, thibetanus,
Nearest to system: empowerment, law, improved, dinar, instructed, thaler, archers, conflict,
Nearest to its: their, his, the, tamarin, her, agave, ssbn, thaler,
Nearest to called: massed, UNK, tamarin, sadler, primigenius, michelob, naaman, mitral,
Nearest to known: used, well, epoxy, adele, such, microcebus, finnic, bug,
Nearest to will: would, could, can, may, should, must, moravians, to,
Nearest to people: reginae, odes, pathological, intercepts, coquitlam, cebus, members, saguinus,
Nearest to nine: eight, six, seven, five, four, zero, three, mitral,
Nearest to also: which, often, sometimes, zionist, thibetanus, generally, that, tamarin,
Nearest to eight: six, seven, nine, five, four, three, zero, michelob,
Nearest to are: were, is, have, be, thibetanus, while, angiotensin, was,
Nearest to all: many, these, some, asterism, reuptake, thibetanus, any, rhenish,
Average loss at step 72000 : 4.80778124154
Average loss at step 74000 : 4.75792721456
Average loss at step 76000 : 4.86112686592
Average loss at step 78000 : 4.79120120609
Average loss at step 80000 : 4.82245359421
Nearest to b: UNK, gland, d, seven, grants, microcebus, pron, david,
Nearest to s: zero, mitral, and, his, michelob, prohibition, inches, tamarin,
Nearest to when: if, michelob, but, before, pontificia, during, where, after,
Nearest to seven: six, eight, five, four, three, nine, zero, two,
Nearest to of: in, iota, tamarin, nine, microcebus, mitral, thibetanus, and,
Nearest to system: empowerment, improved, thaler, conflict, tamarin, instructed, dinar, microsite,
Nearest to its: their, his, the, tamarin, her, agave, topalov, thaler,
Nearest to called: massed, tamarin, naaman, michelob, sadler, UNK, mitral, primigenius,
Nearest to known: used, well, such, epoxy, adele, microcebus, bug, dasyprocta,
Nearest to will: would, could, can, may, should, must, moravians, to,
Nearest to people: reginae, pathological, odes, members, coquitlam, cebus, intercepts, saguinus,
Nearest to nine: eight, seven, six, five, four, zero, mitral, three,
Nearest to also: which, often, sometimes, zionist, generally, thibetanus, it, trinomial,
Nearest to eight: six, seven, five, nine, four, three, zero, michelob,
Nearest to are: were, is, have, be, thibetanus, while, pathfinder, cebus,
Nearest to all: many, these, some, asterism, two, reuptake, thibetanus, any,
Average loss at step 82000 : 4.79923895121
Average loss at step 84000 : 4.79056957233
Average loss at step 86000 : 4.7452732873
Average loss at step 88000 : 4.70395690095
Average loss at step 90000 : 4.76481224179
Nearest to b: d, gland, UNK, six, pron, microcebus, grants, david,
Nearest to s: his, mitral, and, zero, inches, clemency, michelob, tamarin,
Nearest to when: if, before, michelob, but, where, after, during, while,
Nearest to seven: eight, five, six, four, nine, three, zero, one,
Nearest to of: in, for, tamarin, same, nine, microcebus, msg, and,
Nearest to system: tamarin, thaler, improved, microsite, conflict, dinar, empowerment, instructed,
Nearest to its: their, his, the, her, tamarin, agave, celera, topalov,
Nearest to called: massed, tamarin, naaman, mitral, dreamers, michelob, sadler, UNK,
Nearest to known: used, well, such, epoxy, adele, bug, microcebus, hmong,
Nearest to will: would, can, could, may, must, should, moravians, cannot,
Nearest to people: reginae, members, pathological, odes, coquitlam, cebus, intercepts, saguinus,
Nearest to nine: eight, seven, six, five, four, zero, mitral, michelob,
Nearest to also: which, often, sometimes, zionist, generally, thibetanus, trinomial, now,
Nearest to eight: seven, six, five, nine, four, three, zero, two,
Nearest to are: were, is, have, be, thibetanus, while, include, pathfinder,
Nearest to all: many, some, these, thibetanus, dasyprocta, both, any, asterism,
Average loss at step 92000 : 4.72437152827
Average loss at step 94000 : 4.62979676688
Average loss at step 96000 : 4.71152837896
Average loss at step 98000 : 4.6148717382
Average loss at step 100000 : 4.676337744
Nearest to b: d, grants, gland, david, trailed, microcebus, circumstance, thaler,
Nearest to s: his, mitral, michelob, inches, clemency, medea, zero, tamarin,
Nearest to when: if, while, where, after, during, before, michelob, but,
Nearest to seven: eight, six, five, four, nine, three, zero, two,
Nearest to of: in, tamarin, and, thibetanus, for, microcebus, nine, eight,
Nearest to system: improved, systems, law, archers, microsite, thaler, conflict, tamarin,
Nearest to its: their, his, the, her, tamarin, agave, celera, topalov,
Nearest to called: massed, UNK, tamarin, naaman, interpreted, dreamers, mitral, fright,
Nearest to known: used, such, well, epoxy, microcebus, cryo, adele, bug,
Nearest to will: would, can, could, may, must, should, to, moravians,
Nearest to people: reginae, members, odes, pathological, coquitlam, cebus, intercepts, saguinus,
Nearest to nine: eight, seven, six, five, four, zero, three, dasyprocta,
Nearest to also: which, often, sometimes, zionist, generally, now, thibetanus, still,
Nearest to eight: seven, nine, five, six, four, three, zero, dasyprocta,
Nearest to are: were, is, have, while, be, include, pathfinder, thibetanus,
Nearest to all: many, these, some, thibetanus, both, any, asterism, several,
>>>
 
 

49、word2vec - tensorflow的更多相关文章

  1. 49、[源码]-Spring容器创建-创建Bean准备

    49.[源码]-Spring容器创建-创建Bean准备

  2. NLP获取词向量的方法(Glove、n-gram、word2vec、fastText、ELMo 对比分析)

    自然语言处理的第一步就是获取词向量,获取词向量的方法总体可以分为两种两种,一个是基于统计方法的,一种是基于语言模型的. 1 Glove - 基于统计方法 Glove是一个典型的基于统计的获取词向量的方 ...

  3. 学习笔记CB009:人工神经网络模型、手写数字识别、多层卷积网络、词向量、word2vec

    人工神经网络,借鉴生物神经网络工作原理数学模型. 由n个输入特征得出与输入特征几乎相同的n个结果,训练隐藏层得到意想不到信息.信息检索领域,模型训练合理排序模型,输入特征,文档质量.文档点击历史.文档 ...

  4. EC读书笔记系列之19:条款49、50、51、52

    条款49 了解new-handler的行为 记住: ★set_new_handler允许客户指定一个函数,在内存分配无法获得满足时被调用 ★Nothrow new是一个颇为局限的工具,∵其只适用于内存 ...

  5. 一小部分机器学习算法小结: 优化算法、逻辑回归、支持向量机、决策树、集成算法、Word2Vec等

    优化算法 先导知识:泰勒公式 \[ f(x)=\sum_{n=0}^{\infty}\frac{f^{(n)}(x_0)}{n!}(x-x_0)^n \] 一阶泰勒展开: \[ f(x)\approx ...

  6. 88、展示Tensorflow计算图上每个节点的基本信息以及运行时消耗的时间和空间

    ''' Created on May 24, 2017 @author: p0079482 ''' #使用程序输出日志 import tensorflow as tf with tf.Session( ...

  7. 86、使用Tensorflow实现,LSTM的时间序列预测,预测正弦函数

    ''' Created on 2017年5月21日 @author: weizhen ''' # 以下程序为预测离散化之后的sin函数 import numpy as np import tensor ...

  8. 49、django工程(cookie+session)

    49.1.介绍: 1.cookie不属于http协议范围,由于http协议无法保持状态,但实际情况,我们却又需要"保持状态",因此cookie就是在这样一个场景下诞生. cooki ...

  9. 49、html基础认识&常用标签(1)

    从今天期我们进入前端的学习,先学习html,没有任何需要逻辑需要烧脑,只需要记忆.练习.练习.练习. 一.HTML初识 1.web服务本质 import socket def main(): sock ...

随机推荐

  1. AngularJS的工作原理1

    AngularJS的工作原理 个人觉得,要很好的理解AngularJS的运行机制,才能尽可能避免掉到坑里面去.在这篇文章中,我将根据网上的资料和自己的理解对AngularJS的在启动后,每一步都做了些 ...

  2. How To Configure Logging And Log Rotation In Apache On An Ubuntu VPS

    Introduction The Apache web server can be configured to give the server administrator important info ...

  3. 基于ASP.NET MVC的热插拔模块式开发框架(OrchardNoCMS)介绍(二)

    基于ASP.NET MVC的热插拔模块式开发框架(OrchardNoCMS)介绍(二) 之前文章中给大家说明了下我这个小小的想法,发现还是有不少人的支持和关注.你们的鼓励是对我最大的支持. 我总结了了 ...

  4. ASP.NET开发大杂烩

    ASP.NET开发大杂烩 正巧今天遇到一个获取动态生成table中的一个动态生成的TextBox的值的时候总是findcontrol不到.后来经过我们的徐总,瞬间解决,但是我觉得对于一个页面的声明周期 ...

  5. 哞哞快的 C# 高斯模糊实现

    冲动来自于 bing best 这个小工具,非常短小精干,里边的设置界面非常精美而且背景是一张模糊效果的图片,十分养眼,遂想,收集一下实现方式放到类库里以后肯定用得上.一通百度.谷歌.博客园,换了好多 ...

  6. 自动生成Code First代码

    自动生成Code First代码 在前面的文章中我们提到Entity Framework的“Code First”模式也同样可以基于现有数据库进行开发.今天就让我们一起看一下使用Entity Fram ...

  7. ARM linux解析之压缩内核zImage的启动过程

    ARM linux解析之压缩内核zImage的启动过程 semilog@163.com 首先,我们要知道在zImage的生成过程中,是把arch/arm/boot/compressed/head.s  ...

  8. sql查询语句优化需要注意的几点

    为了获得稳定的执行性能,SQL语句越简单越好.对复杂的SQL语句,要设法对之进行简化. 常见的简化规则如下:   1)不要有超过5个以上的表连接(JOIN) 2)考虑使用临时表或表变量存放中间结果. ...

  9. [置顶] java得到前一个月的年月日时分秒

    import java.util.Calendar; /** * 得到前一个月的年月日时分秒 * @author Mr.hu * 2013-6-28上午12:00:35 * Class Explain ...

  10. html5新标签布局应用指南

    html5中为了便于设计者的网站布局新添加了一些标签,本文主要讲解这些标签的实际应用方法. 大多数前端的朋友在设计网站时主要应用<div>标签构造盒子进行布局,这是种非常高效的方法,可以将 ...