From 318ceee2140c316801bf295326ea4c929f89a355 Mon Sep 17 00:00:00 2001
From: allenyang <skyyang_chuang@163.com>
Date: Thu, 12 Nov 2015 11:19:53 +0800
Subject: [PATCH 01/55] Convolutional Neural Networks

try to translate the Convolutional Neural Networks
---
 README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 5aa6472..f639e5d 100644
--- a/README.md
+++ b/README.md
@@ -72,7 +72,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md) ([@Tony Jin](https://github.com/linbojin))
   - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md)
   - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md)
-  - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)
+  - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)([@oskycar](https://github.com/oskycar))
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)
   - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)(@[Warln](https://github.com/Warln))

From 080f4062f312fec2eb3149b037b5c233f96d1da1 Mon Sep 17 00:00:00 2001
From: sstruct <sstruct@gmail.com>
Date: Thu, 12 Nov 2015 11:21:46 +0800
Subject: [PATCH 02/55] Chapter 1-3 check

---
 README.md | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/README.md b/README.md
index 5aa6472..7783dfc 100644
--- a/README.md
+++ b/README.md
@@ -101,9 +101,9 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
 ### 校对
 
 - 起步
-  - [介绍](SOURCE/get_started/introduction.md) （[此处添加审校者姓名及链接](https://github.com/xxx)）
-  - [下载及安装](SOURCE/get_started/os_setup.md)
-  - [基本用法](SOURCE/get_started/basic_usage.md)
+  - [介绍](SOURCE/get_started/introduction.md) （[此处添加审校者姓名及链接](https://github.com/xxx)）（[@yangtze](https://github.com/sstruct)）
+  - [下载及安装](SOURCE/get_started/os_setup.md)（[@yangtze](https://github.com/sstruct)）
+  - [基本用法](SOURCE/get_started/basic_usage.md)（[@yangtze](https://github.com/sstruct)）
 - 教程
   - [Overview](SOURCE/tutorials/overview.md)
   - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md)

From 612a6bed211e04f05419794e3687f8a24afe3d61 Mon Sep 17 00:00:00 2001
From: oujiao <18792435256@163.com>
Date: Thu, 12 Nov 2015 11:33:06 +0800
Subject: [PATCH 03/55] Frist commit

get the work
---
 README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 5aa6472..e1e24d1 100644
--- a/README.md
+++ b/README.md
@@ -78,7 +78,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)(@[Warln](https://github.com/Warln))
   - [Mandelbrot Set](SOURCE/tutorials/mandelbrot.md)
   - [Partial Differential Equations](SOURCE/tutorials/pdes.md) 
-  - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)
+  - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)(@[JoyLiu]https://github.com/fengsehng)
 - 运作方式
   - [总览](SOURCE/how_tos/overview.md) 
   - [变量:创建、初始化、保存和加载](SOURCE/how_tos/variables.md) 

From fbdcb35e4527ebafb8d12be20feb39aa91db2b6c Mon Sep 17 00:00:00 2001
From: ChenWeican <chenweican@gmail.com>
Date: Thu, 12 Nov 2015 12:36:10 +0800
Subject: [PATCH 04/55] regist deep_mnist_for_expert

---
 README.md | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/README.md b/README.md
index 6e8e950..1ab93d1 100644
--- a/README.md
+++ b/README.md
@@ -69,8 +69,8 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [基本用法](SOURCE/get_started/basic_usage.md) （[@PFZheng](https://github.com/PFZheng)）
 - 教程
   - [Overview](SOURCE/tutorials/overview.md) （[@PFZheng](https://github.com/PFZheng)）
-  - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md)
-  - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md)
+  - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md) ([@Tony Jin](https://github.com/linbojin))
+  - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md) ([@chenweican](https://github.com/chenweican))
   - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md)
   - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)
@@ -101,9 +101,9 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
 ### 校对
 
 - 起步
-  - [介绍](SOURCE/get_started/introduction.md) （[此处添加审校者姓名及链接](https://github.com/xxx)）
-  - [下载及安装](SOURCE/get_started/os_setup.md)
-  - [基本用法](SOURCE/get_started/basic_usage.md)
+  - [介绍](SOURCE/get_started/introduction.md) （[此处添加审校者姓名及链接](https://github.com/xxx)）（[@yangtze](https://github.com/sstruct)）
+  - [下载及安装](SOURCE/get_started/os_setup.md)（[@yangtze](https://github.com/sstruct)）
+  - [基本用法](SOURCE/get_started/basic_usage.md)（[@yangtze](https://github.com/sstruct)）
 - 教程
   - [Overview](SOURCE/tutorials/overview.md)
   - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md)

From 69f2f2d8e71d433604ed675bbfc414c2fb7a4a6b Mon Sep 17 00:00:00 2001
From: ChenWeican <chenweican@gmail.com>
Date: Thu, 12 Nov 2015 12:38:40 +0800
Subject: [PATCH 05/55] fix

---
 README.md | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/README.md b/README.md
index 1ab93d1..9be077f 100644
--- a/README.md
+++ b/README.md
@@ -69,8 +69,8 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [基本用法](SOURCE/get_started/basic_usage.md) （[@PFZheng](https://github.com/PFZheng)）
 - 教程
   - [Overview](SOURCE/tutorials/overview.md) （[@PFZheng](https://github.com/PFZheng)）
-  - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md) ([@Tony Jin](https://github.com/linbojin))
-  - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md) ([@chenweican](https://github.com/chenweican))
+  - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md)
+  - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md)([@chenweican](https://github.com/chenweican))
   - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md)
   - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)
@@ -101,9 +101,9 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
 ### 校对
 
 - 起步
-  - [介绍](SOURCE/get_started/introduction.md) （[此处添加审校者姓名及链接](https://github.com/xxx)）（[@yangtze](https://github.com/sstruct)）
-  - [下载及安装](SOURCE/get_started/os_setup.md)（[@yangtze](https://github.com/sstruct)）
-  - [基本用法](SOURCE/get_started/basic_usage.md)（[@yangtze](https://github.com/sstruct)）
+  - [介绍](SOURCE/get_started/introduction.md) （[此处添加审校者姓名及链接](https://github.com/xxx)）
+  - [下载及安装](SOURCE/get_started/os_setup.md)
+  - [基本用法](SOURCE/get_started/basic_usage.md)
 - 教程
   - [Overview](SOURCE/tutorials/overview.md)
   - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md)

From 6ee12903cbfd8f20079fcbafefc79acdc20106b2 Mon Sep 17 00:00:00 2001
From: KK4SBB <zhyh8341@gmail.com>
Date: Thu, 12 Nov 2015 13:02:02 +0800
Subject: [PATCH 06/55] request to translate the section of "CNNs"

I am now working on click through prediction in an IT company in Beijing. The techniques I use include Linear Regression, Decision Tree, Neural Network and some NLP methods.
I also have been a translator and reviewer for CSDN for nearly one year. Here are some URLs of my recent translations.
http://www.csdn.net/article/2015-11-09/2826166
http://www.csdn.net/article/2015-11-02/2826097
http://www.csdn.net/article/2015-10-11/2825883

For any other information, please feel free to contact me, zhyh8341@gmail.com
---
 README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 5aa6472..96bd3a5 100644
--- a/README.md
+++ b/README.md
@@ -72,7 +72,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md) ([@Tony Jin](https://github.com/linbojin))
   - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md)
   - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md)
-  - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)
+  - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)([@zhyhooo](https://github.com/zhyhooo))
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)
   - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)(@[Warln](https://github.com/Warln))

From 82375904d2e636758a9f9af7f6d429bb0f28f18f Mon Sep 17 00:00:00 2001
From: "Jason (Xi) Yang" <xyang40@users.noreply.github.com>
Date: Thu, 12 Nov 2015 00:30:01 -0500
Subject: [PATCH 07/55] Update README.md

---
 README.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/README.md b/README.md
index 7783dfc..7cbaea3 100644
--- a/README.md
+++ b/README.md
@@ -72,8 +72,8 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md) ([@Tony Jin](https://github.com/linbojin))
   - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md)
   - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md)
-  - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)
-  - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)
+  - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)（[@xyang40](https://github.com/xyang40)）
+  - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)（[@xyang40](https://github.com/xyang40)）
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)
   - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)(@[Warln](https://github.com/Warln))
   - [Mandelbrot Set](SOURCE/tutorials/mandelbrot.md)

From 11fdedf34ee626c9d5cf4aedb212b2b7635c184e Mon Sep 17 00:00:00 2001
From: "Jason (Xi) Yang" <xyang40@users.noreply.github.com>
Date: Thu, 12 Nov 2015 02:27:51 -0500
Subject: [PATCH 08/55] Update README.md

take down note for cnn
---
 README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 7cbaea3..7b0f35d 100644
--- a/README.md
+++ b/README.md
@@ -72,7 +72,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md) ([@Tony Jin](https://github.com/linbojin))
   - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md)
   - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md)
-  - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)（[@xyang40](https://github.com/xyang40)）
+  - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)（[@xyang40](https://github.com/xyang40)）
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)
   - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)(@[Warln](https://github.com/Warln))

From bc0570c0825b1a44d0bd5abbdcf54304dcdbb23b Mon Sep 17 00:00:00 2001
From: KK4SBB <zhyh8341@gmail.com>
Date: Thu, 12 Nov 2015 15:28:33 +0800
Subject: [PATCH 09/55] Update README.md
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Hi, since some one has already picked up the CNNs section,  how about let me translate the section of "how tos/variables" and review the section of CNNs？Thanks.
---
 README.md | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/README.md b/README.md
index 96bd3a5..8e7176a 100644
--- a/README.md
+++ b/README.md
@@ -72,7 +72,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md) ([@Tony Jin](https://github.com/linbojin))
   - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md)
   - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md)
-  - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)([@zhyhooo](https://github.com/zhyhooo))
+  - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)
   - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)(@[Warln](https://github.com/Warln))
@@ -81,7 +81,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)
 - 运作方式
   - [总览](SOURCE/how_tos/overview.md) 
-  - [变量:创建、初始化、保存和加载](SOURCE/how_tos/variables.md) 
+  - [变量:创建、初始化、保存和加载](SOURCE/how_tos/variables.md) ([@zhyhooo](https://github.com/zhyhooo))
   - [TensorBoard:可视化学习](SOURCE/how_tos/summaries_and_tensorboard.md) 
   - [TensorBoard:图表可视化](SOURCE/how_tos/graph_viz.md) 
   - [读取数据](SOURCE/how_tos/reading_data.md) 
@@ -109,7 +109,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md)
   - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md)
   - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md)
-  - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)
+  - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md) ([@zhyhooo](https://github.com/zhyhooo))
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)
   - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)

From 33ab4d7afe14ca04419e7394cdfb135516540151 Mon Sep 17 00:00:00 2001
From: floydzhang1984 <tech_universe@163.com>
Date: Thu, 12 Nov 2015 16:23:46 +0800
Subject: [PATCH 10/55] typo

---
 README.md | 6 +++---
 TOC.md    | 2 +-
 2 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/README.md b/README.md
index 5b4a64b..4ddfc68 100644
--- a/README.md
+++ b/README.md
@@ -74,7 +74,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md)
   - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)([@oskycar](https://github.com/oskycar))
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)（[@xyang40](https://github.com/xyang40)）
-  - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)(@[Warln](https://github.com/Warln))
+  - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)([@Warln](https://github.com/Warln))
   - [Mandelbrot Set](SOURCE/tutorials/mandelbrot.md)
   - [Partial Differential Equations](SOURCE/tutorials/pdes.md) 
   - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)
@@ -95,7 +95,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [示例使用](SOURCE/resources/uses.md)
   - [FAQ](SOURCE/resources/faq.md)
   - [术语表](SOURCE/resources/glossary.md)
-  - [Tensor排名、形状和类型](SOURCE/resources/dim_types.md)
+  - [Tensor排名、形状和类型](SOURCE/resources/dims_types.md)
 
 ### 校对
 
@@ -132,7 +132,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [示例使用](SOURCE/resources/uses.md)
   - [FAQ](SOURCE/resources/faq.md)
   - [术语表](SOURCE/resources/glossary.md)
-  - [Tensor排名、形状和类型](SOURCE/resources/dim_types.md)
+  - [Tensor排名、形状和类型](SOURCE/resources/dims_types.md)
 
 ## 进度记录
 
diff --git a/TOC.md b/TOC.md
index 0cf1ace..0833484 100644
--- a/TOC.md
+++ b/TOC.md
@@ -31,7 +31,7 @@
   - [示例使用](SOURCE/resources/uses.md)
   - [FAQ](SOURCE/resources/faq.md)
   - [术语表](SOURCE/resources/glossary.md)
-  - [Tensor排名、形状和类型](SOURCE/resources/dim_types.md)
+  - [Tensor排名、形状和类型](SOURCE/resources/dims_types.md)
 
 
   

From 103b9969d9da9d50989fed3b5aed3e23812223c9 Mon Sep 17 00:00:00 2001
From: ericxu <feitongxiaoke@gmail.com>
Date: Thu, 12 Nov 2015 16:36:18 +0800
Subject: [PATCH 11/55] first commit

---
 README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 4ddfc68..b4c5423 100644
--- a/README.md
+++ b/README.md
@@ -77,7 +77,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)([@Warln](https://github.com/Warln))
   - [Mandelbrot Set](SOURCE/tutorials/mandelbrot.md)
   - [Partial Differential Equations](SOURCE/tutorials/pdes.md) 
-  - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)
+  - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)([@ericxk](https://githu.com/ericxk))
 - 运作方式
   - [总览](SOURCE/how_tos/overview.md) 
   - [变量:创建、初始化、保存和加载](SOURCE/how_tos/variables.md) ([@zhyhooo](https://github.com/zhyhooo))

From c6a9da8c7a412fe233fb92c6ad7d72e308c3a4d8 Mon Sep 17 00:00:00 2001
From: ericxu <feitongxiaoke@gmail.com>
Date: Thu, 12 Nov 2015 18:13:38 +0800
Subject: [PATCH 12/55] first commit

---
 README.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/README.md b/README.md
index b4c5423..9ccfbee 100644
--- a/README.md
+++ b/README.md
@@ -75,9 +75,9 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)([@oskycar](https://github.com/oskycar))
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)（[@xyang40](https://github.com/xyang40)）
   - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)([@Warln](https://github.com/Warln))
-  - [Mandelbrot Set](SOURCE/tutorials/mandelbrot.md)
+  - [Mandelbrot Set](SOURCE/tutorials/mandelbrot.md)([@ericxk](https://githu.com/ericxk))
   - [Partial Differential Equations](SOURCE/tutorials/pdes.md) 
-  - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)([@ericxk](https://githu.com/ericxk))
+  - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)
 - 运作方式
   - [总览](SOURCE/how_tos/overview.md) 
   - [变量:创建、初始化、保存和加载](SOURCE/how_tos/variables.md) ([@zhyhooo](https://github.com/zhyhooo))

From 98eff57315f43ef6ab2057009cd3fd1f4ea16594 Mon Sep 17 00:00:00 2001
From: bingjin <545915891@qq.com>
Date: Thu, 12 Nov 2015 20:19:56 +0800
Subject: [PATCH 13/55] claims to translate two sections of the TensorFlow
 offical tutorial

---
 README.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/README.md b/README.md
index 4ddfc68..57bf320 100644
--- a/README.md
+++ b/README.md
@@ -71,13 +71,13 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [Overview](SOURCE/tutorials/overview.md) （[@PFZheng](https://github.com/PFZheng)）
   - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md) ([@Tony Jin](https://github.com/linbojin))
   - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md)([@chenweican](https://github.com/chenweican))
-  - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md)
+  - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md)([@bingjin](https://github.com/bingjin))
   - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)([@oskycar](https://github.com/oskycar))
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)（[@xyang40](https://github.com/xyang40)）
   - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)([@Warln](https://github.com/Warln))
   - [Mandelbrot Set](SOURCE/tutorials/mandelbrot.md)
   - [Partial Differential Equations](SOURCE/tutorials/pdes.md) 
-  - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)
+  - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)([@bingjin](https://github.com/bingjin))
 - 运作方式
   - [总览](SOURCE/how_tos/overview.md) 
   - [变量:创建、初始化、保存和加载](SOURCE/how_tos/variables.md) ([@zhyhooo](https://github.com/zhyhooo))

From 3537affaad38716f8824d49bdcbd9891f68d66b2 Mon Sep 17 00:00:00 2001
From: frank-tancf <frank.tancf@gmail.com>
Date: Thu, 12 Nov 2015 21:32:02 +0800
Subject: [PATCH 14/55] first commit

---
 README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 4ddfc68..6d9743e 100644
--- a/README.md
+++ b/README.md
@@ -81,7 +81,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
 - 运作方式
   - [总览](SOURCE/how_tos/overview.md) 
   - [变量:创建、初始化、保存和加载](SOURCE/how_tos/variables.md) ([@zhyhooo](https://github.com/zhyhooo))
-  - [TensorBoard:可视化学习](SOURCE/how_tos/summaries_and_tensorboard.md) 
+  - [TensorBoard:可视化学习](SOURCE/how_tos/summaries_and_tensorboard.md) ([@thylaco1eo](https://github.com/thylaco1eo))
   - [TensorBoard:图表可视化](SOURCE/how_tos/graph_viz.md) 
   - [读取数据](SOURCE/how_tos/reading_data.md) 
   - [线程和队列](SOURCE/how_tos/threading_and_queues.md) 

From 2ce498ba5317de98063922b06eaf4cb80c01eb52 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Thu, 12 Nov 2015 21:45:06 +0800
Subject: [PATCH 15/55] Update README.md

---
 README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 5b4a64b..007618c 100644
--- a/README.md
+++ b/README.md
@@ -87,7 +87,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [线程和队列](SOURCE/how_tos/threading_and_queues.md) 
   - [添加新的Op](SOURCE/how_tos/adding_an_op.md) 
   - [自定义数据读取](SOURCE/how_tos/new_data_formats.md) 
-  - [使用gpu](SOURCE/how_tos/using_gpu.md) 
+  - [使用gpu](SOURCE/how_tos/using_gpu.md) ([@lianghyv](https://github.com/lianghyv))
   - [共享变量](SOURCE/how_tos/variable_scope.md) 
 - 资源
   - [总览](SOURCE/resources/overview.md) 

From 0c3c01a302957e995c3f91ad13d3a81bf30e1844 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Thu, 12 Nov 2015 22:32:36 +0800
Subject: [PATCH 16/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 67 ++++++++++++++++---------------------
 1 file changed, 29 insertions(+), 38 deletions(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 8a46c30..ab7a5ce 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -1,25 +1,21 @@
-# Using GPUs <a class="md-anchor" id="AUTOGENERATED-using-gpus"></a>
+# 使用GPUs <a class="md-anchor" id="AUTOGENERATED-using-gpus"></a>
 
-## Supported devices <a class="md-anchor" id="AUTOGENERATED-supported-devices"></a>
+## 支持的设备 <a class="md-anchor" id="AUTOGENERATED-supported-devices"></a>
 
-On a typical system, there are multiple computing devices. In TensorFlow, the
-supported device types are `CPU` and `GPU`.  They are represented as
-`strings`. For example:
+在一套标准的系统上通常有多个计算设备.TensorFlow中支持CPU和GPU这两种.我们用指定字符串
+`strings`来标示这些设备.比如:
 
-*  `"/cpu:0"`: The CPU of your machine.
-*  `"/gpu:0"`: The GPU of your machine, if you have one.
-*  `"/gpu:1"`: The second GPU of your machine, etc.
+*  `"/cpu:0"`:机器的CPU
+*  `"/gpu:0"`:机器的GPU,如果你只有一个的话.
+*  `"/gpu:1"`:机器的第二个GPU,以此类推...
 
-If a TensorFlow operation has both CPU and GPU implementations, the
-GPU devices will be given priority when the operation is assigned to
-a device. For example, `matmul` has both CPU and GPU kernels.  On a
-system with devices `cpu:0` and `gpu:0`, `gpu:0` will be selected to run
-`matmul`.
+如果一个TensorFlow中CPU和GPU都有实现,指派一个设备时GPU有优先权.比如`matmul`中CPU
+和GPU核心都有.系统里同时有`cpu:0`和`gpu:0`,那么`gpu:0`会被选择运行`matmul`.
 
-## Logging Device placement <a class="md-anchor" id="AUTOGENERATED-logging-device-placement"></a>
+## 记录使用设备的位置 <a class="md-anchor" id="AUTOGENERATED-logging-device-placement"></a>
 
-To find out which devices your operations and tensors are assigned to, create
-the session with `log_device_placement` configuration option set to `True`.
+为了知道你的操作和Tensor被指派到哪个设备运行，用`log_device_placement`创建`session`,并将配置
+选项设置为`True`.
 
 ```python
 # Creates a graph.
@@ -32,7 +28,7 @@ sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
 print sess.run(c)
 ```
 
-You should see the following output:
+你应该能看见以下输出:
 
 ```
 Device mapping:
@@ -46,12 +42,10 @@ MatMul: /job:localhost/replica:0/task:0/gpu:0
 
 ```
 
-## Manual device placement <a class="md-anchor" id="AUTOGENERATED-manual-device-placement"></a>
+## 手动选择指定设备 <a class="md-anchor" id="AUTOGENERATED-manual-device-placement"></a>
 
-If you would like a particular operation to run on a device of your
-choice instead of what's automatically selected for you, you can use
-`with tf.device` to create a device context such that all the operations
-within that context will have the same device assignment.
+如果你想让指定设备运行操作而不用系统自动为你分配的，你可以用`with tf.device`
+创建一个设备环境，在这个环境下的所有操作都在统一的设备上运行.
 
 ```python
 # Creates a graph.
@@ -65,7 +59,7 @@ sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
 print sess.run(c)
 ```
 
-You will see that now `a` and `b` are assigned to `cpu:0`.
+你会发现现在`a`和`b`操作都被分配给了`cpu:0`.
 
 ```
 Device mapping:
@@ -78,11 +72,10 @@ MatMul: /job:localhost/replica:0/task:0/gpu:0
  [ 49.  64.]]
 ```
 
-## Using a single GPU on a multi-GPU system <a class="md-anchor" id="AUTOGENERATED-using-a-single-gpu-on-a-multi-gpu-system"></a>
+## 在多GPU系统里使用单一GPU<a class="md-anchor" id="AUTOGENERATED-using-a-single-gpu-on-a-multi-gpu-system"></a>
 
-If you have more than one GPU in your system, the GPU with the lowest ID will be
-selected by default. If you would like to run on a different GPU, you will need
-to specify the preference explicitly:
+如果你的系统里有多个GPU，ID最小的GPU会被默认选中.如果你想用别的GPU，
+可以明确指定你的偏好：
 
 ```python
 # Creates a graph.
@@ -96,8 +89,7 @@ sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
 print sess.run(c)
 ```
 
-If the device you have specified does not exist, you will get
-`InvalidArgumentError`:
+如果你指定的设备不存在，你会收到`InvalidArgumentError`提示：
 
 ```
 InvalidArgumentError: Invalid argument: Cannot assign a device to node 'b':
@@ -106,10 +98,8 @@ Could not satisfy explicit device specification '/gpu:2'
    values: 1 2 3...>, _device="/gpu:2"]()]]
 ```
 
-If you would like TensorFlow to automatically choose an existing and
-supported device to run the operations in case the specified one doesn't
-exist, you can set `allow_soft_placement` to `True` in the configuration
-option when creating the session.
+如果你希望TensorFlow自动选择一个存在的且被支持的设备以防你指定的不存在，
+你可以在你创建的会话的配置选项里把`allow_soft_placement`选项设为`True`
 
 ```python
 # Creates a graph.
@@ -125,11 +115,10 @@ sess = tf.Session(config=tf.ConfigProto(
 print sess.run(c)
 ```
 
-## Using multiple GPUs <a class="md-anchor" id="AUTOGENERATED-using-multiple-gpus"></a>
+## 使用多个GPU <a class="md-anchor" id="AUTOGENERATED-using-multiple-gpus"></a>
 
-If you would like to run TensorFlow on multiple GPUs, you can construct your
-model in a multi-tower fashion where each tower is assigned to a different GPU.
-For example:
+如果你想让TensorFlow在多个GPU上运行，你可以建立multi-tower结构，在这个结构
+里每个tower分别被指配给不同的GPU运行.比如：
 
 ```
 # Creates a graph.
@@ -147,7 +136,7 @@ sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
 print sess.run(sum)
 ```
 
-You will see the following output.
+你会看到如下输出：
 
 ```
 Device mapping:
@@ -172,3 +161,5 @@ AddN: /job:localhost/replica:0/task:0/cpu:0
 
 The [cifar10 tutorial](../../tutorials/deep_cnn/index.md) is a good example
 demonstrating how to do training with multiple GPUs.
+
+原文 https://github.com/jikexueyuanwiki/tensorflow-zh/blob/master/SOURCE/how_tos/using_gpu.md

From accfe47f850ab91d5fe34f8d7b3cddf752ee0352 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Thu, 12 Nov 2015 22:37:08 +0800
Subject: [PATCH 17/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index ab7a5ce..2fe7dab 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -159,7 +159,6 @@ AddN: /job:localhost/replica:0/task:0/cpu:0
  [  98.  128.]]
 ```
 
-The [cifar10 tutorial](../../tutorials/deep_cnn/index.md) is a good example
-demonstrating how to do training with multiple GPUs.
+The [cifar10 tutorial](../../tutorials/deep_cnn/index.md) 这个例子很好的演示了怎样用GPU集群训练
 
 原文 https://github.com/jikexueyuanwiki/tensorflow-zh/blob/master/SOURCE/how_tos/using_gpu.md

From 661c40995415619b80b65926d96bdfe2271f7676 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Thu, 12 Nov 2015 22:38:40 +0800
Subject: [PATCH 18/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 2fe7dab..bddf6e9 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -14,8 +14,7 @@
 
 ## 记录使用设备的位置 <a class="md-anchor" id="AUTOGENERATED-logging-device-placement"></a>
 
-为了知道你的操作和Tensor被指派到哪个设备运行，用`log_device_placement`创建`session`,并将配置
-选项设置为`True`.
+为了知道你的操作和Tensor被指派到哪个设备运行，用`log_device_placement`创建`session`,并将设置为`True`.
 
 ```python
 # Creates a graph.

From 29408fbc99c8ee13a055ef2086ad2fa6c8eef6e1 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Thu, 12 Nov 2015 22:39:26 +0800
Subject: [PATCH 19/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index bddf6e9..610436e 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -41,7 +41,7 @@ MatMul: /job:localhost/replica:0/task:0/gpu:0
 
 ```
 
-## 手动选择指定设备 <a class="md-anchor" id="AUTOGENERATED-manual-device-placement"></a>
+## 手动指定运行设备 <a class="md-anchor" id="AUTOGENERATED-manual-device-placement"></a>
 
 如果你想让指定设备运行操作而不用系统自动为你分配的，你可以用`with tf.device`
 创建一个设备环境，在这个环境下的所有操作都在统一的设备上运行.

From 9e1dbd7bdd2ef9d08de82bb20b74655de51851c6 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Thu, 12 Nov 2015 22:40:28 +0800
Subject: [PATCH 20/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 610436e..179bbec 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -158,6 +158,6 @@ AddN: /job:localhost/replica:0/task:0/cpu:0
  [  98.  128.]]
 ```
 
-The [cifar10 tutorial](../../tutorials/deep_cnn/index.md) 这个例子很好的演示了怎样用GPU集群训练
+[cifar10 tutorial](../../tutorials/deep_cnn/index.md) 这个例子很好的演示了怎样用GPU集群训练.
 
 原文 https://github.com/jikexueyuanwiki/tensorflow-zh/blob/master/SOURCE/how_tos/using_gpu.md

From a55222d7ab63f2522a8c5cc9c9fd004a9fcb4854 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Thu, 12 Nov 2015 22:44:02 +0800
Subject: [PATCH 21/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 179bbec..8c8ba34 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -158,6 +158,6 @@ AddN: /job:localhost/replica:0/task:0/cpu:0
  [  98.  128.]]
 ```
 
-[cifar10 tutorial](../../tutorials/deep_cnn/index.md) 这个例子很好的演示了怎样用GPU集群训练.
+[cifar10 tutorial](.../tensorflow-zh/blob/master/SOURCE/tutorials/deep_cnn/index.md) 这个例子很好的演示了怎样用GPU集群训练.
 
 原文 https://github.com/jikexueyuanwiki/tensorflow-zh/blob/master/SOURCE/how_tos/using_gpu.md

From b1bb68adba1604b73a4f00abd86b88c13309f0e0 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Thu, 12 Nov 2015 22:45:15 +0800
Subject: [PATCH 22/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 8c8ba34..179bbec 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -158,6 +158,6 @@ AddN: /job:localhost/replica:0/task:0/cpu:0
  [  98.  128.]]
 ```
 
-[cifar10 tutorial](.../tensorflow-zh/blob/master/SOURCE/tutorials/deep_cnn/index.md) 这个例子很好的演示了怎样用GPU集群训练.
+[cifar10 tutorial](../../tutorials/deep_cnn/index.md) 这个例子很好的演示了怎样用GPU集群训练.
 
 原文 https://github.com/jikexueyuanwiki/tensorflow-zh/blob/master/SOURCE/how_tos/using_gpu.md

From 72f3d16f6103910225d9bd7ddc6c7249d487edd1 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Thu, 12 Nov 2015 22:46:04 +0800
Subject: [PATCH 23/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 179bbec..be4871a 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -158,6 +158,6 @@ AddN: /job:localhost/replica:0/task:0/cpu:0
  [  98.  128.]]
 ```
 
-[cifar10 tutorial](../../tutorials/deep_cnn/index.md) 这个例子很好的演示了怎样用GPU集群训练.
+[cifar10 tutorial](../../SOURCE/tutorials/deep_cnn/index.md) 这个例子很好的演示了怎样用GPU集群训练.
 
 原文 https://github.com/jikexueyuanwiki/tensorflow-zh/blob/master/SOURCE/how_tos/using_gpu.md

From 21d8cfa684939a1d7bdf0590a53c8cabe243ddb2 Mon Sep 17 00:00:00 2001
From: ericxu <feitongxiaoke@gmail.com>
Date: Thu, 12 Nov 2015 22:48:45 +0800
Subject: [PATCH 24/55] add mandelbrot-zh.md

---
 SOURCE/tutorials/mandelbrot/mandelbrot-zh.md | 137 +++++++++++++++++++
 1 file changed, 137 insertions(+)
 create mode 100755 SOURCE/tutorials/mandelbrot/mandelbrot-zh.md

diff --git a/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md b/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md
new file mode 100755
index 0000000..57257bd
--- /dev/null
+++ b/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md
@@ -0,0 +1,137 @@
+# 曼德布洛特集合 <a class="md-anchor" id="AUTOGENERATED-mandelbrot-set"></a>
+Visualizing the Mandelbrot set doesn't have anything to do with machine
+learning, but it makes for a fun example of how one can use TensorFlow for
+general mathematics.  This is actually a pretty naive implementation of the
+visualization, but it makes the point.  (We may end up providing a more
+elaborate implementation down the line to produce more truly beautiful images.)
+可视化曼德布洛特集合与机器学习没有任何关系，但这是一个将TensorFlow应用在基础数学上的一个有趣的例子。这是一个有趣朴素的可视化的补充。（我们最终将提供更多的精心实施的线条，以产生更真实美丽的图像）。
+
+Note: This tutorial was originally prepared as an IPython notebook.
+说明：这个教程作为一个IPython的notebook。
+
+## 基本步骤 <a class="md-anchor" id="AUTOGENERATED-basic-setup"></a>
+
+We'll need a few imports to get started.
+准备开始时，我们将需要导入一些库。
+
+```python
+# Import libraries for simulation
+# 导入仿真库
+import tensorflow as tf
+import numpy as np
+
+# Imports for visualization
+# 导入可视化库
+import PIL.Image
+from cStringIO import StringIO
+from IPython.display import clear_output, Image, display
+import scipy.ndimage as nd
+```
+
+Now we'll define a function to actually display the image once we have
+iteration counts.
+现在我们将定义一个函数来显示迭代计算出的图像。
+
+```python
+def DisplayFractal(a, fmt='jpeg'):
+  """Display an array of iteration counts as a
+     colorful picture of a fractal."""
+  """显示迭代计算出的彩色分形图像。"""
+  a_cyclic = (6.28*a/20.0).reshape(list(a.shape)+[1])
+  img = np.concatenate([10+20*np.cos(a_cyclic),
+                        30+50*np.sin(a_cyclic),
+                        155-80*np.cos(a_cyclic)], 2)
+  img[a==a.max()] = 0
+  a = img
+  a = np.uint8(np.clip(a, 0, 255))
+  f = StringIO()
+  PIL.Image.fromarray(a).save(f, fmt)
+  display(Image(data=f.getvalue()))
+```
+
+## 会话和变量初始化 <a class="md-anchor" id="AUTOGENERATED-session-and-variable-initialization"></a>
+
+For playing around like this, we often use an interactive session, but a regular
+session would work as well.
+为了，我们常常使用交互式会话，但是普通会话也能正常使用。
+
+```python
+   sess = tf.InteractiveSession()
+```
+
+It's handy that we can freely mix NumPy and TensorFlow.
+非常好用的是我们可以自由的混合使用NumPy和TensorFlow
+
+```python
+# Use NumPy to create a 2D array of complex numbers on [-2,2]x[-2,2]
+# 使用NumPy创建一个在[-2,2]x[-2,2]范围内的2维复数数组
+
+Y, X = np.mgrid[-1.3:1.3:0.005, -2:1:0.005]
+Z = X+1j*Y
+```
+
+Now we define and initialize TensorFlow tensors.
+现在我们定义并初始化一个TensorFlow的tensors。
+
+```python
+xs = tf.constant(Z.astype("complex64"))
+zs = tf.Variable(xs)
+ns = tf.Variable(tf.zeros_like(xs, "float32"))
+```
+
+TensorFlow requires that you explicitly initialize variables before using them.
+TensorFlow在使用之前需要明确的初始化初值。
+
+```python
+tf.initialize_all_variables().run()
+```
+
+## 定义并运行计算<a class="md-anchor" id="AUTOGENERATED-defining-and-running-the-computation"></a>
+
+Now we specify more of the computation...
+现在我们说明一些更多的计算
+
+```python
+# Compute the new values of z: z^2 + x
+# 计算一个新的值z: z^2 + x
+zs_ = zs*zs + xs
+
+# Have we diverged with this new value?
+# 我们用这个新值分形么？
+not_diverged = tf.complex_abs(zs_) < 4
+
+# Operation to update the zs and the iteration count.
+# 更新zs并且迭代计算
+#
+# Note: We keep computing zs after they diverge! This
+#       is very wasteful! There are better, if a little
+#       less simple, ways to do this.
+# 说明：在分形之后我们继续计算zs，这个计算消耗特别大！
+#      如果稍微简单点，这里有更好的方法做这件事。
+#
+step = tf.group(
+  zs.assign(zs_),
+  ns.assign_add(tf.cast(not_diverged, "float32"))
+  )
+```
+
+... and run it for a couple hundred steps
+。。。继续执行几百步步骤
+
+```python
+for i in range(200): step.run()
+```
+
+Let's see what we've got.
+让我们看看我们得到什么。
+
+```python
+DisplayFractal(ns.eval())
+```
+
+![jpeg](mandelbrot_output.jpg)
+
+Not bad!
+不错！
+
+

From dc0f04e85ea717cea9792307bbbc21908c8c097b Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Thu, 12 Nov 2015 22:49:53 +0800
Subject: [PATCH 25/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index be4871a..0e90e6e 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -41,7 +41,7 @@ MatMul: /job:localhost/replica:0/task:0/gpu:0
 
 ```
 
-## 手动指定运行设备 <a class="md-anchor" id="AUTOGENERATED-manual-device-placement"></a>
+## 手工指定使用的设备 <a class="md-anchor" id="AUTOGENERATED-manual-device-placement"></a>
 
 如果你想让指定设备运行操作而不用系统自动为你分配的，你可以用`with tf.device`
 创建一个设备环境，在这个环境下的所有操作都在统一的设备上运行.

From ca94f2a0ec961c1174ab2a1b07044a52ce90fd0b Mon Sep 17 00:00:00 2001
From: ericxu <feitongxiaoke@gmail.com>
Date: Thu, 12 Nov 2015 22:54:35 +0800
Subject: [PATCH 26/55] add mandelbrot-zh.md

---
 SOURCE/tutorials/mandelbrot/mandelbrot-zh.md | 1 +
 1 file changed, 1 insertion(+)

diff --git a/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md b/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md
index 57257bd..96a7b91 100755
--- a/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md
+++ b/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md
@@ -135,3 +135,4 @@ Not bad!
 不错！
 
 
+> 原文：[Mandelbrot Set](http://tensorflow.org/tutorials/mandelbrot/index.md)  翻译：[ericxk](https://github.com/ericxk)  校对：[](https://github.com/)

From 010dc841a3603dd480b00137f37ef2e7fc090676 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Thu, 12 Nov 2015 22:55:09 +0800
Subject: [PATCH 27/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 0e90e6e..314e5c8 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -98,7 +98,7 @@ Could not satisfy explicit device specification '/gpu:2'
 ```
 
 如果你希望TensorFlow自动选择一个存在的且被支持的设备以防你指定的不存在，
-你可以在你创建的会话的配置选项里把`allow_soft_placement`选项设为`True`
+你可以在创建的`session`里把`allow_soft_placement`设为`True`
 
 ```python
 # Creates a graph.

From 13a52ea50d5b74cf7522d049678c1967c3b39e36 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Thu, 12 Nov 2015 22:59:04 +0800
Subject: [PATCH 28/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 314e5c8..14aedc0 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -44,7 +44,7 @@ MatMul: /job:localhost/replica:0/task:0/gpu:0
 ## 手工指定使用的设备 <a class="md-anchor" id="AUTOGENERATED-manual-device-placement"></a>
 
 如果你想让指定设备运行操作而不用系统自动为你分配的，你可以用`with tf.device`
-创建一个设备环境，在这个环境下的所有操作都在统一的设备上运行.
+创建一个设备环境，在这个环境下的所有操作都统一运行在你所指定的设备上.
 
 ```python
 # Creates a graph.
@@ -74,7 +74,7 @@ MatMul: /job:localhost/replica:0/task:0/gpu:0
 ## 在多GPU系统里使用单一GPU<a class="md-anchor" id="AUTOGENERATED-using-a-single-gpu-on-a-multi-gpu-system"></a>
 
 如果你的系统里有多个GPU，ID最小的GPU会被默认选中.如果你想用别的GPU，
-可以明确指定你的偏好：
+可以指明你的偏好：
 
 ```python
 # Creates a graph.

From 7b59c8769f72b07097566a6d18fedf7d0055e9b0 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Thu, 12 Nov 2015 23:01:57 +0800
Subject: [PATCH 29/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 14aedc0..2aa3e2c 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -9,12 +9,12 @@
 *  `"/gpu:0"`:机器的GPU,如果你只有一个的话.
 *  `"/gpu:1"`:机器的第二个GPU,以此类推...
 
-如果一个TensorFlow中CPU和GPU都有实现,指派一个设备时GPU有优先权.比如`matmul`中CPU
-和GPU核心都有.系统里同时有`cpu:0`和`gpu:0`,那么`gpu:0`会被选择运行`matmul`.
+如果一个TensorFlow系统中兼有CPU和GPU实现,当你指派设备时GPU有优先权.比如`matmul`中CPU
+和GPU核心都有.那么在`cpu:0`和`gpu:0`中,`gpu:0`会被选择运行`matmul`.
 
 ## 记录使用设备的位置 <a class="md-anchor" id="AUTOGENERATED-logging-device-placement"></a>
 
-为了知道你的操作和Tensor被指派到哪个设备运行，用`log_device_placement`创建`session`,并将设置为`True`.
+为了获取你的操作和Tensor被指派到哪个设备上运行，用`log_device_placement`创建`session`,并设置为`True`.
 
 ```python
 # Creates a graph.

From 60ec45c5213ba34c388126881fea5d79188133ff Mon Sep 17 00:00:00 2001
From: ericxu <feitongxiaoke@gmail.com>
Date: Thu, 12 Nov 2015 23:13:13 +0800
Subject: [PATCH 30/55] add mandelbrot-zh.md

---
 SOURCE/tutorials/mandelbrot/mandelbrot-zh.md | 32 ++++----------------
 1 file changed, 6 insertions(+), 26 deletions(-)

diff --git a/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md b/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md
index 96a7b91..9890d33 100755
--- a/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md
+++ b/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md
@@ -1,26 +1,17 @@
 # 曼德布洛特集合 <a class="md-anchor" id="AUTOGENERATED-mandelbrot-set"></a>
-Visualizing the Mandelbrot set doesn't have anything to do with machine
-learning, but it makes for a fun example of how one can use TensorFlow for
-general mathematics.  This is actually a pretty naive implementation of the
-visualization, but it makes the point.  (We may end up providing a more
-elaborate implementation down the line to produce more truly beautiful images.)
-可视化曼德布洛特集合与机器学习没有任何关系，但这是一个将TensorFlow应用在基础数学上的一个有趣的例子。这是一个有趣朴素的可视化的补充。（我们最终将提供更多的精心实施的线条，以产生更真实美丽的图像）。
+虽然可视化曼德布洛特集合与机器学习没有任何关系，但这是将TensorFlow应用在基础数学上的一个有趣的例子。这是一个有趣简单的可视化的实现。（我们将提供一个更复杂的实现来产生更多真正美丽的图像）。
 
-Note: This tutorial was originally prepared as an IPython notebook.
-说明：这个教程作为一个IPython的notebook。
+说明：本教程使用了IPython的notebook。
 
 ## 基本步骤 <a class="md-anchor" id="AUTOGENERATED-basic-setup"></a>
 
-We'll need a few imports to get started.
-准备开始时，我们将需要导入一些库。
+开始时，我们需要导入一些库。
 
 ```python
-# Import libraries for simulation
 # 导入仿真库
 import tensorflow as tf
 import numpy as np
 
-# Imports for visualization
 # 导入可视化库
 import PIL.Image
 from cStringIO import StringIO
@@ -28,14 +19,10 @@ from IPython.display import clear_output, Image, display
 import scipy.ndimage as nd
 ```
 
-Now we'll define a function to actually display the image once we have
-iteration counts.
 现在我们将定义一个函数来显示迭代计算出的图像。
 
 ```python
 def DisplayFractal(a, fmt='jpeg'):
-  """Display an array of iteration counts as a
-     colorful picture of a fractal."""
   """显示迭代计算出的彩色分形图像。"""
   a_cyclic = (6.28*a/20.0).reshape(list(a.shape)+[1])
   img = np.concatenate([10+20*np.cos(a_cyclic),
@@ -51,26 +38,21 @@ def DisplayFractal(a, fmt='jpeg'):
 
 ## 会话和变量初始化 <a class="md-anchor" id="AUTOGENERATED-session-and-variable-initialization"></a>
 
-For playing around like this, we often use an interactive session, but a regular
-session would work as well.
-为了，我们常常使用交互式会话，但是普通会话也能正常使用。
+为了达到这样的目的，我们常常使用交互式会话，但普通会话也能正常使用。
 
 ```python
    sess = tf.InteractiveSession()
 ```
 
-It's handy that we can freely mix NumPy and TensorFlow.
-非常好用的是我们可以自由的混合使用NumPy和TensorFlow
+非常便捷的是我们可以自由的混合使用NumPy和TensorFlow
 
 ```python
-# Use NumPy to create a 2D array of complex numbers on [-2,2]x[-2,2]
 # 使用NumPy创建一个在[-2,2]x[-2,2]范围内的2维复数数组
 
 Y, X = np.mgrid[-1.3:1.3:0.005, -2:1:0.005]
 Z = X+1j*Y
 ```
 
-Now we define and initialize TensorFlow tensors.
 现在我们定义并初始化一个TensorFlow的tensors。
 
 ```python
@@ -79,7 +61,6 @@ zs = tf.Variable(xs)
 ns = tf.Variable(tf.zeros_like(xs, "float32"))
 ```
 
-TensorFlow requires that you explicitly initialize variables before using them.
 TensorFlow在使用之前需要明确的初始化初值。
 
 ```python
@@ -88,8 +69,7 @@ tf.initialize_all_variables().run()
 
 ## 定义并运行计算<a class="md-anchor" id="AUTOGENERATED-defining-and-running-the-computation"></a>
 
-Now we specify more of the computation...
-现在我们说明一些更多的计算
+现在我们指定更多的计算...
 
 ```python
 # Compute the new values of z: z^2 + x

From ac00c4beb846e6541991445da806cfa75507025e Mon Sep 17 00:00:00 2001
From: ericxu <feitongxiaoke@gmail.com>
Date: Thu, 12 Nov 2015 23:16:02 +0800
Subject: [PATCH 31/55] update mandelbrot-zh.md

---
 SOURCE/tutorials/mandelbrot/mandelbrot-zh.md | 23 ++++++--------------
 1 file changed, 7 insertions(+), 16 deletions(-)

diff --git a/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md b/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md
index 9890d33..f787de5 100755
--- a/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md
+++ b/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md
@@ -72,22 +72,16 @@ tf.initialize_all_variables().run()
 现在我们指定更多的计算...
 
 ```python
-# Compute the new values of z: z^2 + x
-# 计算一个新的值z: z^2 + x
+# 计算一个新值z: z^2 + x
 zs_ = zs*zs + xs
 
-# Have we diverged with this new value?
-# 我们用这个新值分形么？
+# 我们使用这个新值分形么？
 not_diverged = tf.complex_abs(zs_) < 4
 
-# Operation to update the zs and the iteration count.
-# 更新zs并且迭代计算
+# 更新zs并且迭代计算。
 #
-# Note: We keep computing zs after they diverge! This
-#       is very wasteful! There are better, if a little
-#       less simple, ways to do this.
 # 说明：在分形之后我们继续计算zs，这个计算消耗特别大！
-#      如果稍微简单点，这里有更好的方法做这件事。
+#      如果稍微简单点，这里有更好的方法来处理。
 #
 step = tf.group(
   zs.assign(zs_),
@@ -95,15 +89,13 @@ step = tf.group(
   )
 ```
 
-... and run it for a couple hundred steps
-。。。继续执行几百步步骤
+...继续执行几百步步骤
 
 ```python
 for i in range(200): step.run()
 ```
 
-Let's see what we've got.
-让我们看看我们得到什么。
+让我们看看我们得到了什么。
 
 ```python
 DisplayFractal(ns.eval())
@@ -111,8 +103,7 @@ DisplayFractal(ns.eval())
 
 ![jpeg](mandelbrot_output.jpg)
 
-Not bad!
-不错！
 
+不错！
 
 > 原文：[Mandelbrot Set](http://tensorflow.org/tutorials/mandelbrot/index.md)  翻译：[ericxk](https://github.com/ericxk)  校对：[](https://github.com/)

From 064a1e0630eeb4cbc283cf747d69164b0722aaef Mon Sep 17 00:00:00 2001
From: floydzhang1984 <tech_universe@163.com>
Date: Fri, 13 Nov 2015 10:18:49 +0800
Subject: [PATCH 32/55] =?UTF-8?q?=E6=B7=BB=E5=8A=A0=E7=BF=BB=E8=AF=91?=
 =?UTF-8?q?=E8=AF=B4=E6=98=8E?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 README.md | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 85e0733..9dfe5d8 100644
--- a/README.md
+++ b/README.md
@@ -40,6 +40,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
 * fork主仓库（<https://github.com/jikexueyuanwiki/tensorflow-zh>）
 * 按照章节认领翻译(每次申请一个章节)，在下面这个`README.md`里找还没有被人申请的章节，写上（@你的github号），给主仓库的`master`分支提pull request；
 * 提的 pull request 被确认，合并到主仓库后，代表你申请的章节*认领*完成，开始翻译；
+* 翻译的文件为README.md或者TOC.md中对应的md文件，请不要翻译单独文件夹中的index.md
 * 翻译过程请参 *翻译协作规范* ，完成翻译后提交 pull request 给主仓库的`master`分支；
 * 校核完成后，从主仓库的`master`分支合并到主`publish`分支；
 * 全部翻译完成后，生成PDF/ePub文档，放在极客学院Wiki平台发布；
@@ -77,7 +78,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)([@Warln](https://github.com/Warln))
   - [Mandelbrot Set](SOURCE/tutorials/mandelbrot.md)
   - [Partial Differential Equations](SOURCE/tutorials/pdes.md) 
-  - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)(@[JoyLiu]https://github.com/fengsehng)
+  - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)([@JoyLiu](https://github.com/fengsehng))
 - 运作方式
   - [总览](SOURCE/how_tos/overview.md) 
   - [变量:创建、初始化、保存和加载](SOURCE/how_tos/variables.md) ([@zhyhooo](https://github.com/zhyhooo))

From 22add62457dd08a4e2b120ce2a8c6e68129a8cb8 Mon Sep 17 00:00:00 2001
From: ericxu <feitongxiaoke@gmail.com>
Date: Fri, 13 Nov 2015 10:23:24 +0800
Subject: [PATCH 33/55] mv mandelbrot-zh.md to mandelbrot.md

---
 SOURCE/tutorials/mandelbrot.md               |  60 +++++-----
 SOURCE/tutorials/mandelbrot/mandelbrot-zh.md | 109 -------------------
 2 files changed, 26 insertions(+), 143 deletions(-)
 delete mode 100755 SOURCE/tutorials/mandelbrot/mandelbrot-zh.md

diff --git a/SOURCE/tutorials/mandelbrot.md b/SOURCE/tutorials/mandelbrot.md
index f1ce882..f787de5 100755
--- a/SOURCE/tutorials/mandelbrot.md
+++ b/SOURCE/tutorials/mandelbrot.md
@@ -1,36 +1,29 @@
-# Mandelbrot Set <a class="md-anchor" id="AUTOGENERATED-mandelbrot-set"></a>
+# 曼德布洛特集合 <a class="md-anchor" id="AUTOGENERATED-mandelbrot-set"></a>
+虽然可视化曼德布洛特集合与机器学习没有任何关系，但这是将TensorFlow应用在基础数学上的一个有趣的例子。这是一个有趣简单的可视化的实现。（我们将提供一个更复杂的实现来产生更多真正美丽的图像）。
 
-Visualizing the Mandelbrot set doesn't have anything to do with machine
-learning, but it makes for a fun example of how one can use TensorFlow for
-general mathematics.  This is actually a pretty naive implementation of the
-visualization, but it makes the point.  (We may end up providing a more
-elaborate implementation down the line to produce more truly beautiful images.)
+说明：本教程使用了IPython的notebook。
 
-Note: This tutorial was originally prepared as an IPython notebook.
+## 基本步骤 <a class="md-anchor" id="AUTOGENERATED-basic-setup"></a>
 
-## Basic Setup <a class="md-anchor" id="AUTOGENERATED-basic-setup"></a>
-
-We'll need a few imports to get started.
+开始时，我们需要导入一些库。
 
 ```python
-# Import libraries for simulation
+# 导入仿真库
 import tensorflow as tf
 import numpy as np
 
-# Imports for visualization
+# 导入可视化库
 import PIL.Image
 from cStringIO import StringIO
 from IPython.display import clear_output, Image, display
 import scipy.ndimage as nd
 ```
 
-Now we'll define a function to actually display the image once we have
-iteration counts.
+现在我们将定义一个函数来显示迭代计算出的图像。
 
 ```python
 def DisplayFractal(a, fmt='jpeg'):
-  """Display an array of iteration counts as a
-     colorful picture of a fractal."""
+  """显示迭代计算出的彩色分形图像。"""
   a_cyclic = (6.28*a/20.0).reshape(list(a.shape)+[1])
   img = np.concatenate([10+20*np.cos(a_cyclic),
                         30+50*np.sin(a_cyclic),
@@ -43,25 +36,24 @@ def DisplayFractal(a, fmt='jpeg'):
   display(Image(data=f.getvalue()))
 ```
 
-## Session and Variable Initialization <a class="md-anchor" id="AUTOGENERATED-session-and-variable-initialization"></a>
+## 会话和变量初始化 <a class="md-anchor" id="AUTOGENERATED-session-and-variable-initialization"></a>
 
-For playing around like this, we often use an interactive session, but a regular
-session would work as well.
+为了达到这样的目的，我们常常使用交互式会话，但普通会话也能正常使用。
 
 ```python
    sess = tf.InteractiveSession()
 ```
 
-It's handy that we can freely mix NumPy and TensorFlow.
+非常便捷的是我们可以自由的混合使用NumPy和TensorFlow
 
 ```python
-# Use NumPy to create a 2D array of complex numbers on [-2,2]x[-2,2]
+# 使用NumPy创建一个在[-2,2]x[-2,2]范围内的2维复数数组
 
 Y, X = np.mgrid[-1.3:1.3:0.005, -2:1:0.005]
 Z = X+1j*Y
 ```
 
-Now we define and initialize TensorFlow tensors.
+现在我们定义并初始化一个TensorFlow的tensors。
 
 ```python
 xs = tf.constant(Z.astype("complex64"))
@@ -69,28 +61,27 @@ zs = tf.Variable(xs)
 ns = tf.Variable(tf.zeros_like(xs, "float32"))
 ```
 
-TensorFlow requires that you explicitly initialize variables before using them.
+TensorFlow在使用之前需要明确的初始化初值。
 
 ```python
 tf.initialize_all_variables().run()
 ```
 
-## Defining and Running the Computation <a class="md-anchor" id="AUTOGENERATED-defining-and-running-the-computation"></a>
+## 定义并运行计算<a class="md-anchor" id="AUTOGENERATED-defining-and-running-the-computation"></a>
 
-Now we specify more of the computation...
+现在我们指定更多的计算...
 
 ```python
-# Compute the new values of z: z^2 + x
+# 计算一个新值z: z^2 + x
 zs_ = zs*zs + xs
 
-# Have we diverged with this new value?
+# 我们使用这个新值分形么？
 not_diverged = tf.complex_abs(zs_) < 4
 
-# Operation to update the zs and the iteration count.
+# 更新zs并且迭代计算。
 #
-# Note: We keep computing zs after they diverge! This
-#       is very wasteful! There are better, if a little
-#       less simple, ways to do this.
+# 说明：在分形之后我们继续计算zs，这个计算消耗特别大！
+#      如果稍微简单点，这里有更好的方法来处理。
 #
 step = tf.group(
   zs.assign(zs_),
@@ -98,13 +89,13 @@ step = tf.group(
   )
 ```
 
-... and run it for a couple hundred steps
+...继续执行几百步步骤
 
 ```python
 for i in range(200): step.run()
 ```
 
-Let's see what we've got.
+让我们看看我们得到了什么。
 
 ```python
 DisplayFractal(ns.eval())
@@ -112,6 +103,7 @@ DisplayFractal(ns.eval())
 
 ![jpeg](mandelbrot_output.jpg)
 
-Not bad!
 
+不错！
 
+> 原文：[Mandelbrot Set](http://tensorflow.org/tutorials/mandelbrot/index.md)  翻译：[ericxk](https://github.com/ericxk)  校对：[](https://github.com/)
diff --git a/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md b/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md
deleted file mode 100755
index f787de5..0000000
--- a/SOURCE/tutorials/mandelbrot/mandelbrot-zh.md
+++ /dev/null
@@ -1,109 +0,0 @@
-# 曼德布洛特集合 <a class="md-anchor" id="AUTOGENERATED-mandelbrot-set"></a>
-虽然可视化曼德布洛特集合与机器学习没有任何关系，但这是将TensorFlow应用在基础数学上的一个有趣的例子。这是一个有趣简单的可视化的实现。（我们将提供一个更复杂的实现来产生更多真正美丽的图像）。
-
-说明：本教程使用了IPython的notebook。
-
-## 基本步骤 <a class="md-anchor" id="AUTOGENERATED-basic-setup"></a>
-
-开始时，我们需要导入一些库。
-
-```python
-# 导入仿真库
-import tensorflow as tf
-import numpy as np
-
-# 导入可视化库
-import PIL.Image
-from cStringIO import StringIO
-from IPython.display import clear_output, Image, display
-import scipy.ndimage as nd
-```
-
-现在我们将定义一个函数来显示迭代计算出的图像。
-
-```python
-def DisplayFractal(a, fmt='jpeg'):
-  """显示迭代计算出的彩色分形图像。"""
-  a_cyclic = (6.28*a/20.0).reshape(list(a.shape)+[1])
-  img = np.concatenate([10+20*np.cos(a_cyclic),
-                        30+50*np.sin(a_cyclic),
-                        155-80*np.cos(a_cyclic)], 2)
-  img[a==a.max()] = 0
-  a = img
-  a = np.uint8(np.clip(a, 0, 255))
-  f = StringIO()
-  PIL.Image.fromarray(a).save(f, fmt)
-  display(Image(data=f.getvalue()))
-```
-
-## 会话和变量初始化 <a class="md-anchor" id="AUTOGENERATED-session-and-variable-initialization"></a>
-
-为了达到这样的目的，我们常常使用交互式会话，但普通会话也能正常使用。
-
-```python
-   sess = tf.InteractiveSession()
-```
-
-非常便捷的是我们可以自由的混合使用NumPy和TensorFlow
-
-```python
-# 使用NumPy创建一个在[-2,2]x[-2,2]范围内的2维复数数组
-
-Y, X = np.mgrid[-1.3:1.3:0.005, -2:1:0.005]
-Z = X+1j*Y
-```
-
-现在我们定义并初始化一个TensorFlow的tensors。
-
-```python
-xs = tf.constant(Z.astype("complex64"))
-zs = tf.Variable(xs)
-ns = tf.Variable(tf.zeros_like(xs, "float32"))
-```
-
-TensorFlow在使用之前需要明确的初始化初值。
-
-```python
-tf.initialize_all_variables().run()
-```
-
-## 定义并运行计算<a class="md-anchor" id="AUTOGENERATED-defining-and-running-the-computation"></a>
-
-现在我们指定更多的计算...
-
-```python
-# 计算一个新值z: z^2 + x
-zs_ = zs*zs + xs
-
-# 我们使用这个新值分形么？
-not_diverged = tf.complex_abs(zs_) < 4
-
-# 更新zs并且迭代计算。
-#
-# 说明：在分形之后我们继续计算zs，这个计算消耗特别大！
-#      如果稍微简单点，这里有更好的方法来处理。
-#
-step = tf.group(
-  zs.assign(zs_),
-  ns.assign_add(tf.cast(not_diverged, "float32"))
-  )
-```
-
-...继续执行几百步步骤
-
-```python
-for i in range(200): step.run()
-```
-
-让我们看看我们得到了什么。
-
-```python
-DisplayFractal(ns.eval())
-```
-
-![jpeg](mandelbrot_output.jpg)
-
-
-不错！
-
-> 原文：[Mandelbrot Set](http://tensorflow.org/tutorials/mandelbrot/index.md)  翻译：[ericxk](https://github.com/ericxk)  校对：[](https://github.com/)

From 276aacee8b4e9455057ab78833edf2e2f28b35ec Mon Sep 17 00:00:00 2001
From: ericxu <feitongxiaoke@gmail.com>
Date: Fri, 13 Nov 2015 10:27:53 +0800
Subject: [PATCH 34/55] fix images url

---
 SOURCE/tutorials/mandelbrot.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/SOURCE/tutorials/mandelbrot.md b/SOURCE/tutorials/mandelbrot.md
index f787de5..38ee478 100755
--- a/SOURCE/tutorials/mandelbrot.md
+++ b/SOURCE/tutorials/mandelbrot.md
@@ -101,7 +101,7 @@ for i in range(200): step.run()
 DisplayFractal(ns.eval())
 ```
 
-![jpeg](mandelbrot_output.jpg)
+![jpeg](../images/mandelbrot_output.jpg)
 
 
 不错！

From 515be6ddca50bbbf634f966d92b770dce90d5713 Mon Sep 17 00:00:00 2001
From: ericxu <feitongxiaoke@gmail.com>
Date: Fri, 13 Nov 2015 10:41:48 +0800
Subject: [PATCH 35/55] add images

---
 SOURCE/images/mandelbrot_output.jpg | Bin 0 -> 20185 bytes
 1 file changed, 0 insertions(+), 0 deletions(-)
 create mode 100755 SOURCE/images/mandelbrot_output.jpg

diff --git a/SOURCE/images/mandelbrot_output.jpg b/SOURCE/images/mandelbrot_output.jpg
new file mode 100755
index 0000000000000000000000000000000000000000..8e261d44a8478d39cd7b076a3f870b0420e86e45
GIT binary patch
literal 20185
zcmbq)1z42rw)W5=l7rHq^dL$%BH#=qH83=&ba$tK7$Y^5#L(T{ARyfxQi9}=(vtqe
z_np1Zj`N?r_y3RA<IM0r*L%%+*IM_wS3JLGelGz?6(I@`00ssCfPwx8_&o=Z1>oc2
z;^E@p<Kf{E5a1IM0q+vsxkE%vPC){sr(tBEr=g=`VtLHY^zacg9UX@-=OZ3o0RaI<
zc2P+YK8eTt0(`eO!5|<YAi6_Db@wh6-vhb_eE;$5cPoGl9}^!_2MgmN0Fw*@iwxs;
z2Y?a%J8>}ncme+Fg@K8Mjf0DaPe6DFy`h>EfQf;Hg^7)YgM*EY-tC9J4!|bEA%DOt
zjZ2~a8t<VKkS`FCj?XMp-bM)>{mH^_>>NZuc<(+HH4Q5pI|nC7Ku}0nL{v=nIT#`*
zub`-*3DeTn(bY3CH8Z!cw6b<_b#wRd^zsgV8}jb`hmY{6=$P2J&+!R~8JStxIk|cH
z1r?Q5)it$s^$lO!J370%dwToE#wRAHrf0s*F8^3rU0dJSL~R`$9vz>Yp8Yz%xRnb7
zfc2+X=>Pu|>>uPJL(7GUjg5tkcPkeLrU!aqkzwOJ;Ke1ER>ynoMDdU>5FaRmNH1?A
zVCIMZq%?LOCA`NXu*`aJE7~8D{d<B1{Xa?eSHb>Uu2}#P76w{*SY!Yxz+nSNMl(eW
zM+O}5pRX896+jwxCN&EB=;fkRtUSOdDEXtya~m>A&)m1LU6qU=x+Fr5jD!bZpc=-%
zSeT^vZIra&>+gNp8Ij|bCDM4<P*@O_8zluqjv1lY1y=~ofOByC=Rfk_5Uk5vV~WrF
zCkUpN34f#%g%;rEG3Xh93NsFFZdU_T2lL_p)t~+SN+WmVTn20p3Y4(3yaT98tB_qJ
zM!~az=IdQMdWM-J8bQ%YWtXtLryZhRp|g@W55T#Ra(ar&cNGxv!}eBh$)<AC-gr;c
z#pGN>s;iprNrB8{wFH!T&_YtFgMonC)fFnt9IQ^Nxikrv!6Ks3;#BPP<{Sv|kC}zk
z*8OGE|NpyuxrakvpipEdbOB`c?8|y`sB1=Y==}dpbr{is&u~OFBvNA5UO0P?Z^if|
z^!dF3h^d1q&0nj+_N;)W<J<C(3k0VtbK*-@m$I@wXP&3kOrUnAP-qzuJ~Pu4tiyQG
zRQ_9(DDfDc;@Qt%dV=}0Q#J>@@OAeuIf=1u^KW@hfFaVS*%253dK?k>(SeQ|K{b64
z|3eU%RY_ixX36!Emh+4`r{dkkc+bPAN}EjXr+IIKi0`t#c*cyIX~uP2G;eE1ZD0pf
zmrp}KQo!wEA`4b)uv>M_&&v|n(aK0c)gs0c<iClB{F8G3QQ80N_t4`sOk|!uYFkz3
zpx+<HDo(mF16FJ1w9H6||DTm3sUeX_UY8_qoo@my2mv;86qA?7WX+&A_YwknGg30?
z&I;g-6z)JGLiH)Qx*<>4x6t_N&w1VTqlZ^+-O_tLTG_C=nsAf)7>ZCAqp3UQr8XzK
z>tNA5m7Q!8&qbT_lD;&@ipOJ5GpYl7O8hO|4T>!J7u1kMQ_0OEo-wt3@|!j#IrI|-
z6&_LFhX-bx1+$hXn$>$Bc4?COo6quqkPT6gHGKBa4G~Jhe!<Dv7`1myDpWUg;gf@O
zM52jF)3q8RW634QkaS9Fste7JE&VU~X@3gap;E2!dpbScP}yt<H3TKL>8lY4F-M84
zG207gF%4_)u;{w00y-ZWh#yrSM@H7jU)&jVU%4qfRuCJL$Lg*)WnQk>?$e1`NZY@_
zjQ<%^C?9q%w1BW@Wnp*xKxVg!$T462^Wk!2yg}p9-!C-+l#1qF04S1(h%lR+Zntdm
zndG1t9JFJ;@D<LHCJhEyWXs=Y)XVEE7k*9YljsK2>eYX4y3k7K2GlkLIhdMp;y@Pe
zid1W-xP>~mYL|y(?57h59do{f&{0H@FmDg4lRMxb%b<>JK4)72jAPmLbyh$??gq7y
z3j!ISXxj79J{%Y8_{D0W$MACa*V<04pctT0^Ek?fp_r<!YmBwF`e@FjAmnC~s@TTH
zb6Jhd37&4|n1wSWza2PMO?q6+tA6UZ_O6@wp^N?S>*HCs2c=1+^zy+}?GF>6<5VUo
zlkFlW!k;#U;2cH$G`)=}CuCG2f|AS}uc0f>D_r)nU`}XmqGGVxe>n;U6)u$^1~wf{
z`3r;xQFul)=y8XDI+*z{*Z**)zc|(Z?L8NC4>dl8yoHwRTw}V+yP9{#ON+(eoC*K4
zyDh!FM&s$lxVr6UmUkeE9niCMZ|JC2?0@4>x<XEDtD@seJx4W88!+*zs5^jn43734
zCeE*Kc)_k3x%tQ}>>mXS241}P@O}e|)weLs)z`jEF#2F6_F5?z(^L18@-y4fw_IFb
zPS+~JbBTR~b5*x5Hf<p9C9}Sz(;vL9d4Y$OVx7HunFN|YzLFZ^teEQ=d!YL0@kn~i
zD!JIvwlde<rUxa5bxQ@|F38Yp$bPFv+*2e+iRIQA%0W`4PW`Rfz}qO-svyoq9@8O5
zfWZ;+Cd-io1$#9yP$?g6&-w5zY{&@}70_OqEp&J@`{NrE7y%<-4*#_sdh~sh#Jl)4
zi`qPanX%9Y;IT{FRYY#HZ?=Jn=ZCyrXH-DS+1$~UN;OX3oL=H@z$0xCVy{n#SVgYo
z^?2g9a5YUAT}%=CHU8`#{N)cTk#e>qL-CJ4`3Cmjg>^?9olj3KZ9@yeN0QYZH`OAA
zqo-jy3&LcDpi1X%%$y>Tvgt;%Avxt%qDb$+a6HD1#5uf&^-h`@sQgZQC(YcQSsC>y
zBv`0Pw`(A44+41_6Z=H(wSjAim&_bZz6UKBs0qZzn#wSn^()Zh6m-tyB2`y0>t@at
zs<iIPpGSK-E;^C|WFIcAx(f7_K_GXZLB_yrHuwFoGvJ)S+iU;(f<7{wJ4{o~5>%^I
zNvQ^<872E!O=$2t+oEr6`p-u7cLs&cj#9<PtIt)}%0MJVzXk&}SpI+>EJ;q8=;M4q
zw_@Gb)p{LE1GijtCHO4wj`?>dj_l6ITwvuB_5=dB9!Igs9a~usLXIr929!nsA1$*#
z*u<I^B1*jtpNU-3l1eOfhl0uRTGg-TQ@q3Lj!X4teL#1seKy@*KRhp;DGu1^CDQ2F
zrZWv1;N$V87#G2#=btMDn9k4ldoV?5#9>i-sMUq37cbM2++k{c(0ILHT5DgO@bYeg
zrqSfrCzGxXsKf0Im+5T!CucZ7bt9{C<Dt!2)peX92uX?5=MDXPwT@dPRqbnUH^24J
zOFagj47!}u=h&yJnv#dSS|!&{uA{!<5wj>iE>-0&xX#FB&$$d}4H!e!hnPmKZ|T&)
z6;}D~VW;qUb@BE*t*$5E3FX+t<znj)aE`fK{|rxLJ5_N?W9-XV(ZjrKoZNGUry%8_
z0lx>pD(z3uR9UL>Dkwt1pyCrGe_keSoDcz3*R5lNMnBio43W;Ed|2UVIjX=5Vl!|S
z)2wpKvtH|*34&`G=0sHrsVv{Z{0oFLKT0(tFxR*mxM5ffw^JzlpM&S0)JP-tjw@<g
zMdz}14LV*<?Dwt!2Q@vF3<eI|f;Q&gEk_x@f@f(OXk-!i@Red0Z3yU%dJ5F_LTH2I
zllfyW=QFl9<SF;3Vjgqa=zN_^sCOsi1&cLswkIU?X%VUGe)fo#=I`osglLvx2c#aM
zWVWcmYW$p7gr2>UTvNHhpBmOD?~dk;MQvGD77$eeBo(r;X?o(;_R6WWu*n+jRYa-V
zDA`g4lJF6MSb^V9_Vv2&;60Y#&fXtkl~I4s?|mxB$p@1Bns|sqT@X&W*e#jF%B{?;
zn`)0hAfNO#k$X}P@up6_{nAhAEKIz@#Mn9w1_{j9YfN8--4qLyjJiWs7?Vwn?B=P$
z`NxVBoE1$UYg6|~)^DFP7ON#X@Jw5u+LK4{?beUVAT##B1temN)DN+XY^r-CB}qyH
zvQ4K}RW~?lKCeeu5sJNvPH~zrB^+p~j2i7)dljlO){5yAdUjSie56M*+wciCeRx?T
zDdzBb!OWT?qbc}RaS}F}p4VIa^%a|jm9-&H`b%fI!vZaD*J)Pz*r<h%19eL#S3gQ6
z+7J&l@6b5n87)@tuKqHz?$glPy)ma`<f{;{<txu)i4^Vf)=Dhvd@b!vN_3$8qh1^<
z_5sO(`*ye=*IuE_U5?x@-O+k6Nw%YRl_ISX4KRbFUocNMBH#VGt|7=?f2Y@pV|hf!
zt_#&DGTMq}Ler_Q;_;HRRIGxL+Y9ke4^Kzmaq=pGO?GH~Na9}*@~Ro`@Fcgsav9Vb
zx1Uuji_A3k4^Z5UGipxd`~I?Z45aed@Kb1Wr%?98%-6w2-154KUD`Ad2#B1=Nn@Pe
z2i_$=OFT<93Z}$OM;kE~i$RGWh^A6aY}XcQ9_fcW3I=}qC&Iyzo8`57oJLzV3GId_
zs6CmULISWb;qH9Sb$>9xza=PWbQ;S-16NLLue1Cs8UL6QwH7@$N-LIsVl;r1n4f8G
z1?slSSpu9=ijhRF8Sp7y`N75zAT&@bcHyIaE1HouT6P-}W%UW~(|q+EdJj*xFHxk%
zjMt7a_D5~u$}Qh`7M?jK@X0z=m596O)ZVFZKt3_p#lk`vHG8|U%PlDOmm0Q|YBUn`
zt@Jyu%`1nHGChm8kaEr6n7;5Pu`b0@A?}%8r&>CgzPhPeFw&1dm9YO}#K*YS?>^)m
z9bQ*ZTKL6crhYLUh9)S((-&u@45Ard71ivgE9VY>1A^Qj#A<pkFScyud<mG>0nxRR
zoqSsxt4w66s(OWwdR;Jw*u7q+F9b%k`ZI0-y9ww^Q~5;}(8$(>zx!RZ)Sfzw(PWtW
zs?p=J`s#%iJzJK}(Mok0KPSRj6)R{ZZKGCXf4ftz-oY->68jCraW0|{S8M}o{3lhx
z^r5a77kTAVzMeiFdFHTh3@y3Sp)(tCtY%?-@0gI@!I6GpzQwjfn%0*Q%2)d(zX9+2
z_ry$8ZVu<CqE@9@lqtLdzwT2LCWtb#m_(S&N%C+;Z;?Vi(l_1<sdr^Y1Rv9-=HI{&
zU?+*Oik2t({+};*i*K|9PeEq^(^tG%ZXYEo%G+sbJdH8FuH^9ky7tjvncU04R$A*J
zcBJ;CCsDZ;U(ZuFXwfe&ZT^z{ekZ;8J%1qwrXW;2^yoa%S)MeYK&CxYNEL<jkEN%C
zC#n1fWIXPytLDlxK18&>VEk;k$BV`-4WYui+6#e~{1%9wm(K!<E!6dk3|;pJ_Rnri
z59P{h@1@(g#hVW(MKT-n$xR$|OAm@I?%FRt0}A#0ELh+X-4ewL+&$iO%s=R7k>l;@
z@zea#u$-jb()1fpx_z~nCC0mHxAXb6g%fqyglQSQ&&Jb9c4Td+d!?6JK7;xEJ9g>|
z;s9bv8ybDoMlESSv&da!^z`<&)^Xm-p*?*zgZXH=#9-6v(XUql(ME486@F}Y>Rm`p
z-{9sAj<0H)FQ&3}9G-E3R){i#J1rfGs9w~ptYh<Z_Xei~l-Aac3w@a-3}t8z@iS#s
zta6|>noF^>zT3AsBFu}+wN4bCDarY~uN~AHVKBu<zt?5i`gUDLCsGB^-g|`|mB=1D
zW*!IVx^GNb8QCq$es!VEUY$-;f;_G@O71>U+Dyq#mN6*#Dla<`@iSkb0%jnB5$ufT
z;P4w@s1|GQ+e&C!IvB|(vJLII+tl9Yf$<aOUHe60)8(WvS&y%{AJ3=IYVkK<>Vl@>
zWy7klh`<!JUh!+9-Y*KNE>ES1V3|`*FWjiclM%i2^wIT{K(b)4mz?w+B08vM6fwX0
zl$~x;=K?zP%fJ3JR;C^D%3FPl&C6!ShAaL<uEc>8j`QLJ&}W353m>r(J(UWkgec!~
zFzo;4Sc(ODZ|uLi<rew3#7b}&#Z+5M4j?b_6ge{T-ekmZO6I%!pxL;#tYz*FLzNw+
zL8w8G8712&N=fzGCuAt6z*2Ct7(5&M?dn%JM<JR{#SmuSJEd?3pkb!r!5z!cxjHCC
zC-#p_FU%8>QaKXo+H~5m==2LM4gAG*7-CY)BtWHQ<D_t(4>48Yz@lZ}G+hj$o1kNO
zZ$vHlFs_tkKn;lC?(HK=&JAy$s^s$LAWpX*q3VK=AR(&sxIW%mk^_#A4q0AbWMDYv
z>BjWW-vGj2bp0e8gA=IQ;r@opt6@E4)R|bEz~w|0TvhPJhk5LUk%M8>RRO_1OTrqn
z^`rNX3PdEOS_7hCn<c^@Q!6#6KKe^do<4~ZMnf^>kk0?aK%j*{x7VS;2t=YH2j)rh
zPjdM)ZsvQawGg^9(cQURqZ2z|-_jgiiE#cSKETN_N@vo#50w^eB1+nOSSRsD(rJrp
zs&!qEQ_6hJ^vu6`AR9Y&?B3p~t{H;=L%;KzApTH_-+<|?nwGk1bKV~R*Wq3TRoZw{
z)<-T!;-7nI^}5z#Dp@NfhO%mhsWEg`u_kUL#UeH&gwAj!a%hw1^rz|lt0}99rtT2C
zds(dCy;ANS#$GaM@zrXElZN#&J7J0R?9NPoTSLNwPd_^=m43hWt(DjTtAIV@(xK&H
z9UF^C@A3THVkhH`L9P^EVXBoxJIQWf-w|8UP<%5sGc)B~RdtbenjuKpZGfZ14Mayi
zY-qT2-j)Zo?Czj6bkO)19rS{NZi!V4W;f=WJ9LD}njEcZ35u-+z*o>wuzKX*J=tg2
z$N4bKqs2sRiYQr_cxqo(2E3}scn;0e(=1S$u^j*kRYDwYAuEaa&=4hz-fjbJ7!9%W
zTKGCNev1<p%)xY$-DxKwu$;-iA#0i4NmX>6#K=buHy|`n--&m5P^FQVyVKM;z{b(Z
zO=(RgvyQ&87My$YBicd(m6S~kU9D`4^*Ic-FgahmCSWo>n42xA>~%bM2g?W{kci-v
z1cu<F+%3RI#7~~)VF&LZT?Q`E0i3rVh-LCI0Fsj47VI<HMxe~h`}CicrAXR0$~V>D
zdTC}<Gijyvv%j7T{mI*1A}O*hZx+YJmLi}vGviy3Lb84J8}POCwA`kR3nEcuH`m5w
zLhrP%($Wljf~}4}G(yZM?h+)Y(%i68N;e(xVaP-;;)!tkTJdh~@ujW%rBKRTq>&Lp
zEV&>|?Tru~_D@wK?~P_?2K*deZ2>z{o=1gh!qCr|G9K-p|G<d<{yQFow2qt{ct+pM
z^>U|%^zOm=XYA;z@~`>BwHg;%f`;YwI&=MQvWO6m-aA9y0KZF*lAXJHYKgFisi~;B
zi}@iZ2Mu$a`@|(%^;is9*EYsXKmZ<Q`D^$Oy!KZd!-I37H-_RDQ`JpLE7(CKj^4p9
zz1JYF#d0QHot#G3cLd(OaZxg&_jb@od1OURNvWh>yC}FmVW<+%3lhbXg=)Oqn|flR
z(=4Bt0UwyTP1`_?M7)&788N=C7h`5^TXJ9^RN$6ov4dwNQ4{K#ul^h4&`D)Mb{ME+
zWf%GePtjWVA+!vh4(0}Gy^}_pEoNXe6eAE0reZj*u3|P36^DzB$-_5cZ<tPSV>H;J
z`#;o-&FqLxX9s<9(VG9N*G%*&?6h(Ru0_~RL@+d((o$^G;V(UG-JU$VzQo8~6C711
z%A67~Tx{J2itobj;+jWFW|>irmi8<*y=2j!uLVOG7Xv}%OgtjnMy_QM<^Apo(jqx~
z1-06^z<FFRz)G><BX)io6_uV8@%UB5s49n}wO(sjWgb5sUW#IL#Bj@+zWh>nbyiuQ
z+6H`Nton)h<VVY#y-bg47YQjuShu?J!mc#SWWkHaoPx9ryoxHI<kdl%yLCelXD7oL
zPJL+leRfk>FFt_g33IlE{&Egy_0T~C;uJAt?Uo|NnrQEWcrb?~e%AIIAStJpXV2|F
z_#|@to<@{ovlwyaMBq8YOo?qfPtf}cr`iP{H=9-O39CacV5!~iZjLG3<gv&~`HJXu
zz_4Bnip9#H5m}xy{dFpaRcNWGAwVyARGAB1;O2Vx2Q>T(tHeZKzpKQ=QX2(fT?ikc
zTcKKAN%8EqUpbhpYz!Cq8i&PD9hfW6-26<Mi_JuY?rI(uo!3smQ>(xmN<h#i$&VPP
z4^X^7mI~&|L~yG%-#oQowSGE!n9R{2lR(6~t**CNIqzjpJT28I?v%1#F)jSOijNDj
z8%M~$wlU#jUNWm#z9_;IYutu=pNeYbn7Y(%D%+6b>zv`muGwz@U^$Ln=1HMo7~~o(
zZxu22#l4}@TwtaTa+!-#0y`&>qCF2W@`oYz;%cB&vzVFSxK(D}=&I)*He3k?!l=;E
z{Li++ALaCd<l!|ajN^<y=7K2$4*D~bMHbGF6s9OYt$OQS!TR{aZJ1r;T{pLz(SuL}
zs~BQ8Umtrw`1k~A%{_<;2*c|{*a`?(KI{?$>6TL3B`^HEwN^r>3|FnyPDKWNCYH9?
zIeK%ZPwJ;-s>j~Oy@hP){J?2k!Nv;pULD2hm*(UPYkVWnuMn@||8k)q3Z}5+<ASCv
zSNShawe5Jku%TRsgH`s&U0%NdqE8oHD+!=^nc6wY4IXnhI77KAVuhB2z((zrbb4n{
z_D<4xa1gQdzUC?k)`mBa!w=QQl$EOsxGdd3U^q`Z^JBgMlGbTgr+9zxaN`0Qo$#aj
zXtwufgZy#?#JD@btx_W%SD*o5J_u>==zH~^Er^t2!692Jrh%8O?uF?FLx(q=+r@>3
z9sB*br_k;%u8fOlcG{gbg!jdB+X(ikk0*a;Zrn$B?E7T1pg_K|f?ayX*>|4jb9F(^
zDtK6eti_|TDl0uAJ>61qHXw{gF7$nkA|#DC^20B4RjfUJ1Heu!CTGgyRy~XnE^VbU
zR0VD}B7P|m!vkHOmU-XoZSk(Gl(bAeTKEbvv+Jp;R;G#xDxRLWf_US~C+VP>9L)*|
zWqj5lD<xlP5Ast~WurA&W3jsj+NhS$Dg90Jw)SPb=<KOby2>F{bY+K0>Ql&9Vam%)
zs+E)9fbA(@`aOuIc#Ko(OYiCC`o>xqPG;0SBRLT5&mq-PhNk0{do5Cfg$1#XCR0&k
z3!~ekVX<hOFJ<||xc;bZ{TnSkP4}?Lk;p!0{2lAo@;bNgK$X$GRu+IC1md|AF;u8k
z{1NV>en0*Ls<#i(Ch77$k8UM*#n2PPQwoFT?SyVAbQ<1&@D^gMFW&`#?mQ#vrb}^_
z+v3t1`nKXAq($4)1qATUrRD9wu%ErbeI`<jG4^G)gn>%Ox%HrM+S#}82A{^pZ`LWN
zcA*rrq!>Z`)>R$tI+3-DQhJJ7%VDghgHvs2n7(a_VWKe~7;_F`+<uWSq<`x;{Q~M>
z^*8EZQMJ_Udv!@=BLpaDfl2gK3;+*UZ7swfvyZMIQ9j230#p?}RrRd7KUTfNKxfHy
z+bUQ<t@0rK$Ti0dxEhyEXV1GG0GgG!$wY6Ot^z}S)KoxsKE7~@!(!3Se4jR_KvC#U
zw}WH4$irP#t{!Spb|ocsG`a6DObWq28{PsT#{}MpLYdiV8-4?#jS5a(QHFiJCvtX#
ze9%df&}j2sdrkRyN=bdDvn#QHFqhRhmXzFRL;9>jQw9b00Y%5YG{i8qgp<zaO0d9?
zNhnc`((2JPx_&*K!oT>1S*WV$y!n2B)VSfWVsuralB~?^1$ITK!PPF>yVkcLymNUH
z2W=+>QCqcqIJ0xqPJK(RT~fp2V|k+n6x~xXni|zndJY+e9-Ef=VDXB9vom3QsICX6
zXx9jC!0^dQ4%qDx!H1-N(O$FV2o-)*r!3gz(!0c9Pvv?)-=JJ^YNRPKs~7k&in`ah
z&ne0!`edU<4EE{HCyyGT@p|niEnW=Wp*JnzTx{@Js@Za<#s)p#kFG1@vN=du$D-3A
zqDbEo-r$LFACuA%|Ij0&`-f`)9=h<lA{+7ZqWJi5j4eEWV*YKkz>|9Bu%>GLxO3C4
z@mjhS#?+UGVtx=<QceuxpnuC$ab=t)jk}LO%dk}cgddKa!{xUNex*rD&)$pEg1G{n
zES=>z8D0F90WA0VOQ&Bx)7253vNK<_TTZTDs$38XoNI7wxhyESLf0yKM~LGkf-fdP
zu2BLN0^!aBN)%hlLM=kY>Jv#xTkc;a%H`26H1OBXA=>`_Wz(bZl;T72h87Z%l?R9(
z^8SgZB6>mG1S71rvaEfOuucXl!;WRodzz{`)xZ;#Pc!NoyEgK1?z4vV;_{8oa=$%o
zWZ50Wuc_*{5!@LPsXoo-viD4MUWG%XyW?NRF}#1{w7aW!I_gFG$mbK;{PY~GP|G#I
zw`iw9fBGtdk)n6pS#~P^nzOrD@KQTOeb0h87%joN(8GW)R#m8{CqflIN!_o|a!{d}
zHwG{>J#Wj(e}6@7qEcrE9i(vXS;@qq`(VTD*$QJ->Y`@iwt{LTE-D%DM(SF}3?wj7
zEzM_=Lr5t~_8A)zNL;nW1;+kxTY3+LW&3?3<W+*}O$l5^uFdd*fp3@4eK8MnyV%@j
z9o9QcUlHv-7n=DTZbT&5=6PG{_th3)-g*15rFx#Q@Ne{*8+oPnRG#8o_ASeUv6LWd
z_4z4o6m0#KX824V!)6>|ytzGKh##iB<TmI!mXXQjhOZvCO3oXu{N76Dn|KUEvsahZ
z+2>T_2%gUPED`L8&$$hkHKHdg)sRngLz$0ES$yBV5|-po5y=ze+MPORjqE4wmSfM&
zlWben9@-FmeV_fVK<&_V&Gg00@iOf<LFeS_(Z&XIsu|+@p{A}kv08AhRnSDNZhgPI
zI|KykxwQ;*0`ccWgT|bZvZYSPvw0rRn`SwOzO;-2iWt_x`{r!XTIM@HE8FSEI0=~O
z7`hqz3Ku?qQo%QU#bO)()6q{9HFumZpp?~HgDlrL!!QZ_WidVCJ^F)=wS)J`g66Q1
z*mSi7SYbICTwT1m7<U!0Jme<jrD1bJ)rDqI?P4<j27SW(+dZW-0!B*tkwqaDK&&A=
z`s+cp;ssvcO2Pu!b-Uq&(C8>{vIqmlY;P}R&WVmv3!7rM6TafdItEkYS&Q_v7$Zw>
zTmnz7wq*)u^|>2J#meXXt6v?LW_L|G`yVw*aMtzNb-%9a-d4>KUM#)mL(}b4r4BZ>
zlw)7K;*2?edba$cXmS1EHaf}PQgvGPc|jZ<hX(&xqiBJ5Xpo0K8p%N37x-DX$yLB(
zdxVUlqL@kn&~rLLxH1G@J~v%weGJ&LvM$~Uw3|8MdWW^?VmM>VSQYgm+<XCHg9#6!
zeH%%{BVkdjjL_3;QzV{Z?Je%ih%BQdG|22Z2tInymP<spS`y-Z%AS$b<O98bR?HmY
z7ve60)KG#c^+yx2MurA~D4U5A*s<Nv;Xa&VpBox&&@oNS<D%<CX%TzJ&R)pRWy0q-
z07=81kp0<OnEQ?BzG~2Sb5yQs6o2I9o4dm;4l1!C7b9W9vb&bvK&wlUh0ySc%tsOI
zJz0BPvCIq!3{|5JcdCL!13@BM=kgMc0iUNTHkv)^tR7gAl1-Ek8*I#_9!(W`MKpam
zT~txBKweA_Y7Kc=n9e7Tot@07TqGm1xrc(NSZm4q`mi8;xzwu%-6m6I_AC?z62a)X
zft#$~0I`wx>fdvFZI?YmSL=sX=BPL2=r5f%owtK|OA5M(AD>O$X!$2!_2py_%m<-4
z!=r!V46gR0b!IGNoANg>2PnyU$6m)d;yv}TX!l_OOrZ6{5{uvFA+h(>fK;~Ycp)pq
zwKTKxvSqo~X7R;4I7f*w`lkPD!ufZ02?xn%SY;84P_H!-=5_WSc3NYfl2SjW8~`Jn
zT_Qe<ERlP5*<_9FK6iw4DlU=UXGhnTWeQNIcdGfaAfkkS6zbe5>w+=AjN`n~0ZEJ-
zU0#teQh!`OGZu>+8)K;MHPWu{#%6_l_ne5O-m-a@)Ae&6UF50|V2<AR2ln}<`uxn%
za&Ymy#lZ(5Xkvl?-SryOtXnXVkcIxGS-L)c8#X=xmp1{w;HrF=QV09EQrqRf0Y9wO
zZmNFTOlBLh{s!E=k&ERx_YYXt=6e}1+{AeH>o?#)pq=~Xr^E-Ri1V$0Uy1dJ2GeE|
ziZ9201DGWno^)UJu_^rqoG+7HUrSvnn=bac8maP$wA(Iuy&t)BF=W;F5@>|?0J74`
zK3(lTmRj>ECu`h4b+v_GQBae;=moc!{@C6emMFWEOVw`xXoA98VjaKX;IjRO()rHw
z4{=!QlB&fzvmjKpOk(!Xntg+Ret6A8{@4liP__wp%_j4fca2oGR@ff)Tqc+gxlcRH
zklKE&?8=H!aPY0jMS<Nz^g?U2_}=F)lg&~+%toBzozzdBrq>HY(l*}E9R`4O*(ZFX
z^jH~|BN(&oy-Y@+K>ALRDeYh1Z<KCiE;?M#)CZeBtMsv{k*-MbqRutUuvM$FtP@1%
z9!L)R1}J{}QuBlWzh0CQbA}^gT|IRTCLvL1r=_6Ze-&Qf`GI4x%dc!+!7Q}yGBAso
z{pNf<y*cVUH`%u-&7kg1y*5{LFCChTT)+}CNMGOscA4JXdZy-qVy*ShA?Dd`eEfd5
z-c^^YN+XCc9+bN~dG<Hp-U35}a`o}|k<E40?@AYkq6HwEPcB47@UPD@SwGiKrI&{1
zt7v8|3z~;_jB87SE%i|YQx493)V?*$)aUZUMky&ROD<bmBgw4y1?|n;HaP|rsv6K-
zR3`>)JAdE4HiyS)#4jyFNRRiAg!QsY;y>hNz!A6Rb<17<okj7iW;az#PGbmQB<B4a
za8jr@NNNWF<RZeurc?b--fQU33lNKn6X3)!)ZH{Co>U;erV!T(EG6zwdQ-!4HfY)0
zBUdatb?CVmRlkr_-+4a_F?x$ehiMJO%a<RW@+qOjs1zI>Me<5|Wl5`v_r;ztce>|!
z!x~Uci~JJBB-++nFL@NM+Wbp{^U!bqYl3d`;0l^k1K-_BOQ}~hY5igh;}7Ugqtmp$
zj@W=|^&+|TotCeyCMf|3z5Fj(BB!i_M3h}rrWO+8SQq2LSo-57KlbmLFFl3ZQDzn`
zt~wKEBq=3YmmzU7q6Mm(tSbzjGts<$Ih2*f4ByQ(jGTlElLe7bekNy%(xo#^Bg#DZ
z4Uk&0kyHQNua+6~y>kX8@_^s>sv!T|>)O9%W2{-R+zka%D*2N7^E&)W-0404jIlx2
zo^VjeM>dE=Da$F~<i<SyZkBWZ!+Z}Zq4gCPACm*AYy%rr?#UI{d@d71LP5?z!O}^I
zM;t_@PtRlfwlgI;xS~g*Davk>$M&|1r_~llz3SfWkK=MAxAp>s+!Yy;J7&?E*MAVp
z=;M_Zipi6<Q9t|wCaE|iKgw>E`Pl;}24|(Vk>SfV|4aU)WmZbdg~Qr2H^1M2r=@6~
zoROXL@x~lItM*dN^&c&S5cdb5OO*}@D?F6wiDWWFV2e;KgdS;CDVfH($H@WMZ4do?
zZWtq~r8+<vd+aho6&6_A)9BtsI-y!PF*N!txfm62bQMakG^x(JnE^i<{Q4iEGe_E;
z2H(MrN7(rgmJqWceLA1eNUrShl8*9+m4%|zOW|Or>Jh6i<~^p*A4)cH*yD*vsycI_
z_#`w}7ipB8_X`8w%YLF(<O>-JkDSmt%0k63EiGNr#2l9lz@BOiKNMCOPz?KWkPi~N
zoLrxnS)~1luFbN&b-uMX)jt4waL<7r9Z|Awvp@{m5cb)^$#>`?3k=0Ck9H&Zia}9>
z(y=74dayPd5p6cNw5s+_fLQh8b$j}c9BFi+a8PI`5m<spke*V}j-ZoI55&`<5sbB~
z$C)GhNjBd_0c@G1Pdc*2S(P@8?)W1(V)ksq1~Pq&1?ey*5)(z9TLLWtLnn^4^XW{i
zJL(nZoAKe3UXeIYV;$k8PP7s)?&p`x3;Og8!bvoU<!5#IK6LW;V{|>Oy1cYWNGH=5
z<z-hNP8${vSls`mIi|yF{>U4-<a%vg`O#RKVK+a;{X{aL<5G+2iqb&SV5eHE`gF^?
zYU93XN|*9lp`L++snGmYR(R%>b+aq4G;zi-Md5p6o^RMo?!9zQsEgwwn?{Gqo;E{W
zm)3a#hSZm(DqvqDVQiGJh_6)ij=N0u_eZr{=ra9A;$JOG0Z2aB*e}xt(K%(*PzjpS
ze|Lw5T*awvBPbx-rt|)c9ZV;F51%`BI|w3rB{gh+#efT2t&dhZIzFc!av!x9Q<$Zd
zFB%ChyW2tJg6gK<%05ac(hxsmb{Tg#-&SEN?`28#j>Su`5a4Fsk>Y*6igZ?$2)HQR
zEZ4QD%q&1{%=9X9T3vhaa%7vC8W$0ki7<tdT5+v}N1`?SFY~Ow-BWns-Bn<5L~pJ&
z+bB=fb9HSUVo=`j4#zdY6tt7O+U1DleGPVdm8~N9W6SpWet-J7Ri5^M`%<HYCF;g|
zD$KauakQ(VQ^Em@j?;Ey)7vDy*pf|mS(o&u&5qO4zAUeN+ZK)XVggQ=r<wT?Qz_o#
zFRMM>?u9yfqlb~Xu%M#1yf#<iEWbKweo_3^8gCN=EYr((Y5CcgsbR7TiS6j5ixa4e
zuIGhfq8$$hJ3RJbnL{xcEU=>lH$B#%JZjrkEmT=o)<IKaV<0)_a$w$O0n}23Gip>>
zYsYG`!fMBSG6Xj+Nri6uIlCV<GEzfFBcDWZKV8^`?SmozH5c2#Gv9J??<F^>@s{)b
zvh|96Y8kqtPM;VGB)1Rk@*I?_YuLebOh)&kXxDxW9lEv6sJIbdDf(E!*jZsa@q3lr
z(E=8Z7PP1>knZL^>LlxCs4_6eC(595SsR=3Vkx}oS}xZCF;Es>H5R;m=<nZ5qQ|dh
ze^H}ZJ#Nxap+NoC0#Wb9vQc+I96SLNwJl+SJ)tE!ZQCEexPer&yKsf*e8Y6ALPq*Y
z4WDG+hZf|`3UjpDBGEv86sV`fb{hF9)Z4x(YO=U7OhrlmD%7Gj@zHr?PyBYheD`_J
z6$%?l)e)#5<G$*BaPzRXgsl3GW(X^EX*F;%LS$u}J^F@{A@z=~9&M~=M-e`tR#J`%
zwSO$Q_8VYcs~so2wx4rdjAm%RD0lVr_RQ{3U*#Y^TN<pAx^%f5qZW&2bXre;1FVyi
z`WGtg#g}gUS{NqXXBD!L8SKP!@sA;rDgmW>zU4)s4?;AVL|MJC4_#84w@1^XOYNxi
z7xllxw&zra54On*rgpsKt5zfR<GzbxhpteWbm1}A{RXg;@kw^%Xsc13%#BQJTkdR*
z2`PqJ<kA<+yv67dJpM&T$8%uN|8o0=?nkm2ee$=&pRZgyg=Wq*;^ry47ql3;`~qb4
zo{)C;pTq9G7pCKd@@V5>BTvuJ-OkM}j-j_nn=lY^mY-3L{JMJ9(l^K6eT6Y%1!iNS
zeyKbfF2xc8AT@PDq<C-8MHnqzXi56pgk|q+HSb*BJpGcJsjzrEQos5~`{Xt+`?prD
z1-6DPFG6RRE2ua~>|mVkH$b&mv)E4ppIf4vpNt&?3y@`zdW~(7?Y?#lvwdznIhw3(
z8MnKFkhiSZqA;g3aqJRpL+=-<ROJ_M-l8S_0u3i<EnLa=`|>zejAu1SLhl5b%7t}X
z(O6IX?*7HOR9!Zm9;j~6)uDPeD1<g0&v6^}`i9%g<&UvKxhKy+Am5gNpE;V9mbXkE
z>k%6CAZT{-XP%A@3RW>{$!{lnH}hf-+ECD`kMo^{Xpem%t*SMQn*bJR<P~(*!(kEy
zR*$4=nvQJAwx(&YX<L4P%yy@DIT^)~k0rk7Qc6z^a2@4Qe?uGAU9Dy}hTJt*N3>ne
z-%zd0{mfNi3159wL{;8ujIpu_w$^Vgq-xzYT1#4o+JUVy4<~;E;&L4XEZYQ123kMk
zl2SP1PyxusOVz($o#u37Exhn;3j*}+`AqG{eM!xJsGns9FS+Av!_feowF+pvY0(h=
z<n9eLDJO`6xa{C)H+_(!7Aa2tG6_9g$daza2*yG>b7jP_N-`hv=jztyGCu}P4GRxb
zJ2twX4Le;hq)ZApHlb*EW(!2s#DknJXu<_yUqJUQ7e;CVp6JcWw^9x&0)NEnzC@3f
zMx>8_C|J(eGH?E<Q4b^wB4ohw_HUY~anBKC{Y))E<jR;(eD}*lqkqgI8%RaT-C69&
zY0dOC?}%|wPS&gyFFeF(r)wu9MtCaE=y|q+U>?N38gk0!O8ZPyA{b>DV_itA{<$gi
z*VS3bm_?-i4r|1vq|4P!(-cuUcGwy+ebvXupK|2Gi>XeQsAp;PD|yc_Q0Q9awkFNj
z1-op^91COjr4E(R>G8SJ9RJvZ^BVdg{~rqWXAOQ|0#f5jXR^hOeggty!CYL-l&_@e
zKV+3ATjiR&-6HfY3jaF`fIi&PD`@AC!g9DG--p{{9%>qjF`O$s7Og6CHcHwlL%&u&
zLL88-c;PVcg1<1=HX{Dw;QR*m?R}<hwiwsT?Rs+SQ}Zg7(Q`Yy%bs~meY`amSCJ(K
zzd>ug1!{i<vZxmpRvP0|Wy?A{skV;DLSu*&ugG?|NbwzcQ(=4G!E7B<9IQ}BkyF7_
zpk%dtp4M)_FJZJ)XM0|Ca7QO+^L%vOJeS7gxRj<Q3>mq}hfcj^o$p<CZ!7$@^`tH>
zf<1yw?7W@?res@2I|EM0t9F5EL!JDd{5B#!&ICoFdvJ~%`M<Q|HDYU0t?htV+HBZl
z2qm>3ImK8VK|K%N7+Y?jx~6>E(nPW~$hij1ePXLjdkrGp(!=j3Dw=hed+PB#TWDjk
zz-*?Y%gBjIcpE$5)2+KF<$5Qt&DWXt#^pS5CPXfdcO1C-k?a|vn8t;6?#ufsgR&me
zKMA8Z6ge$};QSDqVYeh7L?7_2c^urSn5DB%3qJTTQ9+LN+6NZN3#Ke8=ZjnMh;Bqj
zSuLYUcpJKh)>Yv1F+EEI-Eh(~A>2z>&tV1~&eL~`uoqjjRhzM<2!g4_kAn4tcNUGi
zaAn=7+n%;B77XNYDJ;@`q;hHPUR=Q&&%fgG^?QyI=@Pc_;cm#BXs^jAIpcF!oYMWa
z+mA85zUl2K$$Kp&=H_Xo&1gd^#5>W0!Av#G(0o&MEz_3qgpUgT<_`YR8&5p?^62r>
zr?EfX@}`nvbs-DC0mB+!_~BI={COGh6~TWCT{#NfwcaiFzv3%pm3`BVYCE~IMzswY
z<SBo<CvY7^;8Z)Ly8qgd1g%vcr-+TA?Bq1JMe29U;lv=0@59l!3z*2CK%bS={@gXa
zsd6tTfK$pZs?z5EveG-!U0EJA+X2JtUhxHkwkOP&<X%2bh<){Mv+}k2LonuLPb^1N
z@b_BpG0eD72)CelwLTqhp}63@ta^*OYTnFRnD2?M!5F@x?3wCV92zFv>%UdZ!_mhI
z{!A(o9(?t)IsxWJa#B*}qtRM^0u9dqP8b{sN6*T%K6nS_*46?!gWO)BPbS1<zP9|^
z!EJhU^kLB+#!Xezi0#?#DS+--aJaK(M3O?l+`94U<;S@}B#Pniy1@(s4c3fIBCEOU
zLRHHIx*|7VQV8w)V)3-fdvN|l&JP@eK#j&{k1YpTDHSuHAEr-eSV<Q%3-mf!>dEkN
zT9zcjM2)fp%nfIHA1xZU*a8CH6{h=m%=BntMD=)MI<@vNq=wm%B?jq`OTRO?)Xu(a
z=*_qM{&G9LSwcwxqEGc8muZiYht|2jhjI1Vib$jRV>s0(aisl>r_o!i8}A#vU0Zma
z8sN03F}}lAUW7e3SbwS?HIbF63Tb&V-hU|0`9pLew6XDi&5U>cvu?VT94|Vzo-0tc
zL#x&u^AyjODB>g?T+tiklw?oMy#LV4;83zQhWhPaCRbd`<#qDv9-&_h{j?Da1T{N1
zhiH>(T2cF3QvA=KQ^w}!Rqhk1`HoR;-LEgqCD4p|3!9^`;;P{IdY|F-18=0;g;w_k
z5Kca|44GeMrF9uL_Hk(y{Pavjl-?%u&L>YVL)MFEj_>-(lDT^cu<|HA9-jc!t{Uc5
zCxa`#LcNb`>JOsR>Fl<IQgUnYifv|k>+!K8N${wK$5Ud@1j?bEqw6&b0mUSYlq!n{
z?thf!R*Y$tkt3u(K_H(L&e%5fZIhfk&vN9w3H_dLaO?D&$a6!`5gZro<7ybNnYEAk
z$Yo_|g+ysLw8RKB0Q?!(6$$W|#?+rt6rvA;6;#VdxjPIc53DhCRG}xk{}$zOu!9k*
z+IXdu^uX~lqy~FNh8EXxJ{Fc%@&HoruMh}b4L^=i!wODxd}S>LikW5YBp+3O_b`Qu
zW>T5Y;C5dDi<*^HRoYZpD&+ZRbu0Yrj7D_DgzW<h9b>i`U)LGmdEZbbm71zc#D?B|
z4gAA8mR;^&gq`)~Uhp!IOK_@Ds^Tfz^yF%N5qe;9Waq1Nt#eM1nXZ<QDX0lEGy07f
z(&d^<s8%qT_nk=8J)vJXw0IE#`3%+4RlVJb8tXc(ubhIUQNNDRE#lJmUCUb&Me`Ik
z$~AhnElUj9oF!nVM@3&|D?k;Q?1S}{<74TGvzGLTk*j9;Rrmd05ku8Yb-TsbWK~JM
z{dmyaRNG6EF+>M+R=~ECE=nh)*B5a-Tq^kOI-2wKL}I0q+vMrdw9-eIN5f1<Mzupx
zvB^iXaL98}*rWO<fyeZ!jHSh9Z_Tw{WreJLpnqEZV=5g4trQBKi0wdUN8~~O7-(Z*
z$w7InA6o&=$`W6<@nb*Zt6)f?5%Q#P)_Ar`<GdinnS-<W1737$ceX3Ez0XB0?(^P@
zY%}LOKFnP56_qj?*n6Rf;nHlQu+BW*4y^84Jkj=()eE=$l_&fTcRuo;y`dU%Nj*4O
z8Bf9-w#0F!R*^0(Z8e3H*=JOuzPBu|)%Qxss0?iE*=4=+FhmmjtI1;(%iV3Nj3aws
zm&uNm?ND$Yl8#LOuu85&B=q#GzU`(N%<{#Qze;HEFr9_+^1F=pX(_t!xxXONIG4L}
zc%_?Bnr_r-n%I*c7;AFb2>amPZPe`fbaAm}ci#*>>27c9Pe&N8U_riS0(<1CmbQ9v
z?FUlOWq9`BJXT&j#*7YuM7m0)E*q_rQ<i<&I-FKDys!VH(V-=WLeU21V$Z(#CUp(2
zF;y_LKs|}b<Umc(RC!O|drhnuCoy@STA)jG12_wHR%Z1)^stbbA;A*F(~FV~jL-|O
zj1~$STSgzsiMuV<fjKv&Xu@N|brKb$O^a?d^H&f0FHgdX?YoRnOwi%^Q8+??_wZBH
zit`{PO%DlPFjRxz8tsJ|=!up;&tBZV|6FxjPr;W2(THh4S=;!$LA>n65~SWVRpWeU
zoaM-f|BLqVa8Gh4G}91z5cm1d3>`60%)8vy<|69Kz^$!JN%a^6<c2khxgNe!CsMTS
zSV0;A^ndxR8_UI=8l4`F6&tlnjMdNf(}eo%?47lu|I~-DH9cJMT^q-N)nVm2d~Gg>
ziM4e8W;OjK>B$jly#5=+8cDg<vIO>FmV1w3`u!{(ahQTm2*)Rm3l5Z=>a(!CUSTYZ
zL>%2H>v$SXlM9VB){+7qc+vAit$GF$cz5X`VkWbhB=2&m>$LRtl5gjl_q`k70`-~Z
zHPolth{{JHr>l?mUrcyt&AbR4o`%m51S&@{ZOIjT^Auy(GiQiQJvcIqkD3JwrD!|O
zQ|(5Z1G*(rd))Sw!v)88Q&>$A3&CVnKgCHM@um7}z3bwpNee&xS{RbU6Fs3@0e-R^
zP>6P<Ue>LQK;3}7c>7|u7X7`?y?t-?n|9q+iWf!`1d--C#vwr7xq1V#D3LA|Vi(l@
zICYL{4%6Yy(9ZUiy*Ck(m`5t%nhTcZIa1^adj%`MGIT=E8+PUQM9=`|=LmWY4ibki
zG5MhDO8sm03%jMIF7n!qqcQpljbxP{&US5IJFC73KJH#oPZrxDCnC<Wc(;_ATe_^L
z_72K|fOi!r?upzDZ<lbpDhN-{S8{jvdfQds)-@k4!YnCLitK!~`FbyF;lifzvPKM;
zvk(inqXh*iE?w|Q^&umhf>#02qaH-5X<OD}T>UD-lUCd4iK=O+Syi3ed|AB#1=6W?
z__z>Hp@BAh@`1jU=*sr-_PJ)9Sm9iOeqX@op@-eeq|^ZG6nR~FXIZXS?=%oSJ*6-4
z0@LA%p|s~;TAZShvHNK=uJJ1i+!Jp*vg|lS0@Q~{(pPG%mAK@pqqMYoxekv4=bcq2
zTu*jK`swb1!a5>F=#C5XZHM3yK|s49KKUW1ly>+9Yvg#=cv!RjG1~Z4f3dS1+-v?O
zqcAYiFrV+{p#|6g5N))KB?Ade_tH{9(y^_yF`=duH%{kMh$^a{<@ZeC8HL>zkLtqZ
z?DD!x4NS_9i=9M$DZye2?w9*VHLJ2@xYeU1Wubg=gfNj=f?m{(Xz^X737jeLb?C!m
zN+B-m_!NDXQ;Cb{^=Qi++t<b&rHi7>66Ps2o55?05#E&$E{+w$9d^#<Z(#xVMhN<H
z^QcAhU0Y4b#RKbVMNal_!1|3~9v)j0`XNwB0`|v``r?MpPwivJ#gN~dTOci^zXA{=
zxr5RrIUm4K9gOISG2Yf(3ia~dPT#4Al}U{ck?1(gJBOaO>sSbXTe8C&t}hUyu1j0|
zM|e$Oe(kiannzzN$fgOKnMH@@yuXEszx}K+CDX(J=7VDt#$n-71FH?liKX+ko_R{p
zc<?*bv~QxxAUZ$$@7BmIA!J}V$$f{HEejT8lMOOG)_mQwJt)B=2^xR#{$2>WuoiU-
zZh#*tv;uvgOg40OZzY}A?mfP3Ib6W!v?pFn#Sk-VFs;dZ*NxwimPO&=?rTJ(6^s%B
z$R6Z&eWaOK-RtCD_R1TI9!iGMtR9APDRZHqV`PhsoZ7~qo?!Z?85N}f?8nb~yKJnb
zBW<XjnKp_tlP$*!Gld=<UugRt(8bU-qEG?9*v;>2dyM4fH|o!V32U@qy+bOv(klnl
zD4mns{J9#56L(h;aqU>?@e<Dsksp;@OZ=jWI`0Fxy#2+01KPK|kBj@zqIj*R>$_VF
zm8r9tnlZ7wvJ*kYo}i`~>NmN_YpPm6ci56wh`N*;D_p{*hF`83g=>q-H2*NsUlw>I
z??$z)XPrrzVBDF7A&cLNA-%h;*C(d^U8`swm{=)c0EsS3w~`4|0(((%Ri_6@gRy@k
zB`zl$t>uMh^<-sKoECjibCaWTBjQJrJ^Pvm0|0mcyV6W<IrrGJIe)PB_sy(e=r$6X
z0kxSugcQG01V*b<L?>$7b%)2y_NlBNgrNaHOYK(Q({_b>)`TuK>!-lkI2%e+v|KoU
zr`W$$@_&6M(Ix!pMahBA3kt`Aq<aJmKGVi=<UD%w(^KFgQ>X}i3(SAaofzoCo5tI*
z)JByRqLvaQDXGvk2vs5y#j0`sKO#`V14RO0j2d=eV0K=9+lOtvF25e9ou}|;(%HrM
z_XVnEx#TBooDF=4(Ojh@=A=COxkGD_p|o^2)W14CPvvJG2Cdard$nmrQ>cen4Sr9+
z=kY5)o4GLg@r?OK_m=eOket0NaGvnD;wTDS^)i-MUMI{VCv<L`-iKTdb+STgdb0JY
z+e(6T2cX>_-i=_D(6S>vuEcyaSmjGsaMzmZNyJ;Xg9DH0kk=Tcaow*hbU{svN~&Ma
z{2bPN&g3#x>?Mi);Uf*>&?(lKc{5sZ$4vHI6+hKY?}>YAaxTJgJFSsUfjbK9DAZz;
z+`GI2ke&7Eq?@qDrwC!`xT$?No-p2&T-5QFjs4J+zVK@&5!csYZ&nc@R96uW(x~SO
z7G|8E`u+Nfh=PL@u6}hZT13s?do)_g#HczPo>F}*Y7Y};w4t9aX|N6VjulN#`H(W?
zLdI3L<lD{g9(_VV--*YRNQ?}jGYU9c%kGtQ?31RJBw@G?xR|U-!vVOK+E+t1+0p!N
z$!%|#ANX15gPx3)vQU#^G#{LhBj^5ciTl|6Bv&{Kp_Q?>hGh|QjvLH`E|U`yhJBYH
zq(5rS(Nu&WBDB!DS4g|HD-9OA!fp*!38G1Ny8-1}xQY2s4yET7g?LK9&Cs=tN4V(9
zhjkqbT`;uh*24D8SM1)8f`L8%XkxeD8A{P+)`o4WDS#=bnEjBBVkQM{++uP*z7<`c
zr2J<Ak!1TtMxvc9TZs}FLIC_JU7X3%<5X6)Wk*@W-e*49X-32Y)@Z6Gf9_uTv^}vA
z>|N)pnVCH>?&JZH(BeK@cxEg{JgK#ZLH#vcRfKYcvY__fHoL^egmy8c%8Xb()6@8+
z<f_TMzT_lQZNtRk%>tgdSGVs*`O4E+$XJ)O^T2RzA8k9)a^uPNe85Sej3vX}l15o}
znaom2D|?pjn7IjY0Q*<-Qj(IMU0!FUSQN2&Q=uD~=NC^J16q&Mp<AQh!d{iSmV`BW
z4e@w;Y+%+fhq#A^PSSF{Zl}9gnyMf?%UXM%+v_saF`=AIj%7ws!E0bEN>!){r8y|q
zNl;aw;_7$03)G-OvtwvY(KoJ5qHh-(mKkOC>-p%uMr0(eGSj{PvqkCWnV0e$-8bb{
zyK=;_TN{9J?F|X>WxI3Zs~EG4+|Hi@wo<#ad?#%JN;I}&DJbw2B+x=cp{Hv-uv-Xj
z!X6blI>DU1*?$$d2^#=RU#JsI%e6&;?pz?X(`Z9hRMV31H(6J9&-9W7rUu}`QE&yc
zcK%wcMJq*~RA1XL^-Ee&Z#%QM?^4mre!y*%ySf9ndrev}E!3yUt?JR<9@QsTd)BUN
ztNqfopsCXG(F&GBZzfEYydy2iy#42|knh6myQDqx47cu*t3UZi>x<aT?(h4J<@dgi
zo0{_MXY{82c>5ph3nf<>gue>iy3^3?q?_&MquI_%Cl*d^WZ$voXUT?lUQ@aIclUo-
zc4*VhyDVBVJ1lM=-|u%-cyGv~ufPsrBXCA-rBtctgRgJ0FK0L%Gu7K2(Kz*7L9NEi
zc+=S5*NmD3`&iw*{QNSt+k$pmDFI7WpY-F*YrV@fwfoYz1BE=FO<pTHyWF%V2bkyJ
zYxG&vnbj7g-6{4fzf<wjXJy((P)WaF0eVOh%SXzFOdKbcu6i(`h}UZq^T7v~ch2s`
znG0psgue^|s$Y@msPH5>Xroc@$ra01v;Dff+F5Ol$btvJA%ITD6AMb0X}<Me*WvYR
zrf0O{)u}5@5-&#d9^V)<r{4Z))RG5bzHd5Jk0i*fl0E018268DzKP1bGqdE6?Bt!g
zezvm8!I=&xRLhvZ0J{~ZwjFBF+}I$JG{-l!bG=kx$ihu(#iglbx~?ipZoQkgEKXnk
z(*LZYLO2t!_5G2R@yL~&o4<m@)F$oF3Mopv2n-_511DF^{9a%c=6mC++oR3jL{4d}
zxxI5$7)ru+Res)hslDcdRY=LqCCOKT3El&kEl`sv8GOc5Piv1a^;sjmG<;F_q(?K;
zx;j(;eSZ!LK$A1KP`YPKuul5C^*8>NoYL?+lnLz22^_5oX*~GrT30x*H&<#KZ6B2?
zS~#^M>(RM>t=#Yc{s4FNk8k}vlzPpyG~yjJnP-{*4qvdDbA^U_wxm^l;)&H#tT7L&
zW^pt6==iD@g|62;J@Z{qYrB{la2UEf_g%)+rU@GY&zjEP@aL6LQBjsmcaQ##HAj0@
zMM9(ynI70VS8L$vyl=xdrSJC+S<NZyG9|tOyRAVwpixC?VNFNv3&7MPv}W;-9V@lA
zYHf^bn|q?*O~j(bk&dsAdd~T>pluc~&fxu7Sb(v!EAO@Nu>Kf$(bufA&}Vkhyf0f>
zMSgaEOZltGt?@qKqPxN0nTnGy1ak{to7CmIEvU;8=ycHPTkz3Nfo_G@7G+lJKljKq
zvz?<Kw{NPjVe;FTzKbR<0x1jtx1`{TVVbBw2Q`H+1P0&4)PEoTtbEY+B4g1kr9kh%
qj742(z_5i3s^O!3G8tZ;6$wf`s^)Sg;<Kl#NG~iz@X`GLZvp@~FR=6g

literal 0
HcmV?d00001


From 40192b3560a659e9c6a8e6bea73a1447f2f80b17 Mon Sep 17 00:00:00 2001
From: ericxu <feitongxiaoke@gmail.com>
Date: Fri, 13 Nov 2015 10:46:19 +0800
Subject: [PATCH 36/55] fix title

---
 SOURCE/tutorials/mandelbrot.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/SOURCE/tutorials/mandelbrot.md b/SOURCE/tutorials/mandelbrot.md
index 38ee478..71f8ad6 100755
--- a/SOURCE/tutorials/mandelbrot.md
+++ b/SOURCE/tutorials/mandelbrot.md
@@ -1,4 +1,4 @@
-# 曼德布洛特集合 <a class="md-anchor" id="AUTOGENERATED-mandelbrot-set"></a>
+# 曼德布洛特(Mandelbrot)集合 <a class="md-anchor" id="AUTOGENERATED-mandelbrot-set"></a>
 虽然可视化曼德布洛特集合与机器学习没有任何关系，但这是将TensorFlow应用在基础数学上的一个有趣的例子。这是一个有趣简单的可视化的实现。（我们将提供一个更复杂的实现来产生更多真正美丽的图像）。
 
 说明：本教程使用了IPython的notebook。

From 48971f9eebef408f54d957aabcf82abf6f2e2236 Mon Sep 17 00:00:00 2001
From: ericxu <feitongxiaoke@gmail.com>
Date: Fri, 13 Nov 2015 10:47:17 +0800
Subject: [PATCH 37/55] fix title

---
 SOURCE/tutorials/mandelbrot.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/SOURCE/tutorials/mandelbrot.md b/SOURCE/tutorials/mandelbrot.md
index 71f8ad6..005e53f 100755
--- a/SOURCE/tutorials/mandelbrot.md
+++ b/SOURCE/tutorials/mandelbrot.md
@@ -1,5 +1,5 @@
 # 曼德布洛特(Mandelbrot)集合 <a class="md-anchor" id="AUTOGENERATED-mandelbrot-set"></a>
-虽然可视化曼德布洛特集合与机器学习没有任何关系，但这是将TensorFlow应用在基础数学上的一个有趣的例子。这是一个有趣简单的可视化的实现。（我们将提供一个更复杂的实现来产生更多真正美丽的图像）。
+虽然可视化曼德布洛特(Mandelbrot)集合与机器学习没有任何关系，但这是将TensorFlow应用在基础数学上的一个有趣的例子。这是一个有趣简单的可视化的实现。（我们将提供一个更复杂的实现来产生更多真正美丽的图像）。
 
 说明：本教程使用了IPython的notebook。
 

From 6c4c7f8c35ad16f2dbb6501950cda30ad379cbc3 Mon Sep 17 00:00:00 2001
From: floydzhang1984 <tech_universe@163.com>
Date: Fri, 13 Nov 2015 10:58:24 +0800
Subject: [PATCH 38/55] Mandelbrot Translation Finish

---
 README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README.md b/README.md
index fb051da..e67101d 100644
--- a/README.md
+++ b/README.md
@@ -76,7 +76,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)([@oskycar](https://github.com/oskycar))
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)（[@xyang40](https://github.com/xyang40)）
   - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)([@Warln](https://github.com/Warln))
-  - [Mandelbrot Set](SOURCE/tutorials/mandelbrot.md)([@ericxk](https://githu.com/ericxk))
+  - [Mandelbrot集合](SOURCE/tutorials/mandelbrot.md)([@ericxk](https://githu.com/ericxk))√
   - [Partial Differential Equations](SOURCE/tutorials/pdes.md) 
   - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)([@JoyLiu](https://github.com/fengsehng))
 - 运作方式

From 1c2aa8eb68f9adc3d3d379ebcfa2022fb0e63688 Mon Sep 17 00:00:00 2001
From: bingjin <545915891@qq.com>
Date: Fri, 13 Nov 2015 03:12:15 +0000
Subject: [PATCH 39/55] resolve conflicts for data download section

---
 README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 57bf320..d6544fd 100644
--- a/README.md
+++ b/README.md
@@ -77,7 +77,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)([@Warln](https://github.com/Warln))
   - [Mandelbrot Set](SOURCE/tutorials/mandelbrot.md)
   - [Partial Differential Equations](SOURCE/tutorials/pdes.md) 
-  - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)([@bingjin](https://github.com/bingjin))
+  - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)
 - 运作方式
   - [总览](SOURCE/how_tos/overview.md) 
   - [变量:创建、初始化、保存和加载](SOURCE/how_tos/variables.md) ([@zhyhooo](https://github.com/zhyhooo))

From 48353825280ee89419a5e6df1491e1d6f2f945bc Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Fri, 13 Nov 2015 12:05:11 +0800
Subject: [PATCH 40/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 2aa3e2c..3b0a8f1 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -158,6 +158,7 @@ AddN: /job:localhost/replica:0/task:0/cpu:0
  [  98.  128.]]
 ```
 
-[cifar10 tutorial](../../SOURCE/tutorials/deep_cnn/index.md) 这个例子很好的演示了怎样用GPU集群训练.
+[cifar10 tutorial](../../tutorials/deep_cnn/index.md) 这个例子很好的演示了怎样用GPU集群训练.
 
-原文 https://github.com/jikexueyuanwiki/tensorflow-zh/blob/master/SOURCE/how_tos/using_gpu.md
+原文:[using_gpu](https://github.com/jikexueyuanwiki/tensorflow-zh/blob/master/SOURCE/how_tos/using_gpu.md)
+翻译:@lianghyv 校对:

From e99e4948d7fa6bccfdcadaf911eaed21d37e733f Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Fri, 13 Nov 2015 12:06:25 +0800
Subject: [PATCH 41/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 3b0a8f1..cca28b4 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -161,4 +161,4 @@ AddN: /job:localhost/replica:0/task:0/cpu:0
 [cifar10 tutorial](../../tutorials/deep_cnn/index.md) 这个例子很好的演示了怎样用GPU集群训练.
 
 原文:[using_gpu](https://github.com/jikexueyuanwiki/tensorflow-zh/blob/master/SOURCE/how_tos/using_gpu.md)
-翻译:@lianghyv 校对:
+翻译:[@lianghyv] 校对:[]

From f7aff45afce446a7dd16944528c2696adea306ee Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Fri, 13 Nov 2015 12:08:09 +0800
Subject: [PATCH 42/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index cca28b4..47a0f30 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -160,5 +160,5 @@ AddN: /job:localhost/replica:0/task:0/cpu:0
 
 [cifar10 tutorial](../../tutorials/deep_cnn/index.md) 这个例子很好的演示了怎样用GPU集群训练.
 
-原文:[using_gpu](https://github.com/jikexueyuanwiki/tensorflow-zh/blob/master/SOURCE/how_tos/using_gpu.md)
-翻译:[@lianghyv] 校对:[]
+> 原文:[using_gpu](https://github.com/jikexueyuanwiki/tensorflow-zh/blob/master/SOURCE/how_tos/using_gpu.md)
+翻译:[@lianghyv](https://github.com/lianghyv) 校对:[](https://github.com/)

From ab3b834e27fb32c39e5d83e17f771cc2e1edc5ea Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Fri, 13 Nov 2015 12:12:52 +0800
Subject: [PATCH 43/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 47a0f30..10bdff1 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -160,5 +160,5 @@ AddN: /job:localhost/replica:0/task:0/cpu:0
 
 [cifar10 tutorial](../../tutorials/deep_cnn/index.md) 这个例子很好的演示了怎样用GPU集群训练.
 
-> 原文:[using_gpu](https://github.com/jikexueyuanwiki/tensorflow-zh/blob/master/SOURCE/how_tos/using_gpu.md)
+> 原文:[using_gpu](http://tensorflow.org/how_tos/using_gpu.md)
 翻译:[@lianghyv](https://github.com/lianghyv) 校对:[](https://github.com/)

From 6aea3907f502f4cd06f212aa5b5030ae6d02ac46 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Fri, 13 Nov 2015 12:21:00 +0800
Subject: [PATCH 44/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 10bdff1..28a0427 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -14,7 +14,7 @@
 
 ## 记录使用设备的位置 <a class="md-anchor" id="AUTOGENERATED-logging-device-placement"></a>
 
-为了获取你的操作和Tensor被指派到哪个设备上运行，用`log_device_placement`创建`session`,并设置为`True`.
+为了获取你的operations和Tensor被指派到哪个设备上运行，用`log_device_placement`创建`session`,并设置为`True`.
 
 ```python
 # Creates a graph.
@@ -160,5 +160,5 @@ AddN: /job:localhost/replica:0/task:0/cpu:0
 
 [cifar10 tutorial](../../tutorials/deep_cnn/index.md) 这个例子很好的演示了怎样用GPU集群训练.
 
-> 原文:[using_gpu](http://tensorflow.org/how_tos/using_gpu.md)
+> 原文:[using_gpu](http://tensorflow.org/how_tos/using_gpu/index.md)
 翻译:[@lianghyv](https://github.com/lianghyv) 校对:[](https://github.com/)

From 30676d69e4747713b97099f13e33afb321a3f14b Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Fri, 13 Nov 2015 12:26:57 +0800
Subject: [PATCH 45/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 36 ++++++++++++++++++------------------
 1 file changed, 18 insertions(+), 18 deletions(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 28a0427..8a56039 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -2,11 +2,11 @@
 
 ## 支持的设备 <a class="md-anchor" id="AUTOGENERATED-supported-devices"></a>
 
-在一套标准的系统上通常有多个计算设备.TensorFlow中支持CPU和GPU这两种.我们用指定字符串
-`strings`来标示这些设备.比如:
+在一套标准的系统上通常有多个计算设备.TensorFlow中支持CPU和GPU两种.我们用指定字符串
+`strings`来标识这些设备.比如:
 
 *  `"/cpu:0"`:机器的CPU
-*  `"/gpu:0"`:机器的GPU,如果你只有一个的话.
+*  `"/gpu:0"`:机器的GPU,如果你有一个的话.
 *  `"/gpu:1"`:机器的第二个GPU,以此类推...
 
 如果一个TensorFlow系统中兼有CPU和GPU实现,当你指派设备时GPU有优先权.比如`matmul`中CPU
@@ -17,13 +17,13 @@
 为了获取你的operations和Tensor被指派到哪个设备上运行，用`log_device_placement`创建`session`,并设置为`True`.
 
 ```python
-# Creates a graph.
+# 新建一个graph.
 a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[2, 3], name='a')
 b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2], name='b')
 c = tf.matmul(a, b)
-# Creates a session with log_device_placement set to True.
+# 新建session with log_device_placement 设置为 True.
 sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
-# Runs the op.
+# 运行这个op.
 print sess.run(c)
 ```
 
@@ -47,14 +47,14 @@ MatMul: /job:localhost/replica:0/task:0/gpu:0
 创建一个设备环境，在这个环境下的所有操作都统一运行在你所指定的设备上.
 
 ```python
-# Creates a graph.
+# 新建一个graph.
 with tf.device('/cpu:0'):
   a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[2, 3], name='a')
   b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2], name='b')
 c = tf.matmul(a, b)
-# Creates a session with log_device_placement set to True.
+# 新建session with log_device_placement 设置为 True.
 sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
-# Runs the op.
+# 运行这个op.
 print sess.run(c)
 ```
 
@@ -77,14 +77,14 @@ MatMul: /job:localhost/replica:0/task:0/gpu:0
 可以指明你的偏好：
 
 ```python
-# Creates a graph.
+# 新建一个graph.
 with tf.device('/gpu:2'):
   a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[2, 3], name='a')
   b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2], name='b')
   c = tf.matmul(a, b)
-# Creates a session with log_device_placement set to True.
+# 新建session with log_device_placement 设置为 True.
 sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
-# Runs the op.
+# 运行这个op.
 print sess.run(c)
 ```
 
@@ -101,16 +101,16 @@ Could not satisfy explicit device specification '/gpu:2'
 你可以在创建的`session`里把`allow_soft_placement`设为`True`
 
 ```python
-# Creates a graph.
+# 新建一个graph.
 with tf.device('/gpu:2'):
   a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[2, 3], name='a')
   b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2], name='b')
   c = tf.matmul(a, b)
-# Creates a session with allow_soft_placement and log_device_placement set
+# 新建session with log_device_placement 设置为 True.
 # to True.
 sess = tf.Session(config=tf.ConfigProto(
       allow_soft_placement=True, log_device_placement=True))
-# Runs the op.
+# 运行这个op.
 print sess.run(c)
 ```
 
@@ -120,7 +120,7 @@ print sess.run(c)
 里每个tower分别被指配给不同的GPU运行.比如：
 
 ```
-# Creates a graph.
+# 新建一个graph.
 c = []
 for d in ['/gpu:2', '/gpu:3']:
   with tf.device(d):
@@ -129,9 +129,9 @@ for d in ['/gpu:2', '/gpu:3']:
     c.append(tf.matmul(a, b))
 with tf.device('/cpu:0'):
   sum = tf.add_n(c)
-# Creates a session with log_device_placement set to True.
+# 新建session with log_device_placement 设置为 True.
 sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
-# Runs the op.
+# 运行这个op.
 print sess.run(sum)
 ```
 

From 4a5689ba15a6a03c8bfb1898017906c9d6b8fc08 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Fri, 13 Nov 2015 12:31:53 +0800
Subject: [PATCH 46/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 8a56039..4326c32 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -14,7 +14,7 @@
 
 ## 记录使用设备的位置 <a class="md-anchor" id="AUTOGENERATED-logging-device-placement"></a>
 
-为了获取你的operations和Tensor被指派到哪个设备上运行，用`log_device_placement`创建`session`,并设置为`True`.
+为了获取你的operations和Tensor被指派到哪个设备上运行，用`log_device_placement`新建一个`session`,并设置为`True`.
 
 ```python
 # 新建一个graph.
@@ -43,8 +43,8 @@ MatMul: /job:localhost/replica:0/task:0/gpu:0
 
 ## 手工指定使用的设备 <a class="md-anchor" id="AUTOGENERATED-manual-device-placement"></a>
 
-如果你想让指定设备运行操作而不用系统自动为你分配的，你可以用`with tf.device`
-创建一个设备环境，在这个环境下的所有操作都统一运行在你所指定的设备上.
+如果你想让指定设备运行operation而不用系统自动为你分配的，你可以用`with tf.device`
+创建一个设备环境，这个环境下的操作都统一运行在环境指定的设备上.
 
 ```python
 # 新建一个graph.
@@ -74,7 +74,7 @@ MatMul: /job:localhost/replica:0/task:0/gpu:0
 ## 在多GPU系统里使用单一GPU<a class="md-anchor" id="AUTOGENERATED-using-a-single-gpu-on-a-multi-gpu-system"></a>
 
 如果你的系统里有多个GPU，ID最小的GPU会被默认选中.如果你想用别的GPU，
-可以指明你的偏好：
+可以这样指明你的偏好：
 
 ```python
 # 新建一个graph.
@@ -97,8 +97,8 @@ Could not satisfy explicit device specification '/gpu:2'
    values: 1 2 3...>, _device="/gpu:2"]()]]
 ```
 
-如果你希望TensorFlow自动选择一个存在的且被支持的设备以防你指定的不存在，
-你可以在创建的`session`里把`allow_soft_placement`设为`True`
+如果你希望TensorFlow自动选择一个存在的且被支持的设备以防你指定的设备不存在，
+你可以在创建的`session`里把`allow_soft_placement`设置为`True`
 
 ```python
 # 新建一个graph.

From 82ac4605c80560477d862a42e6c93f9422052bdd Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Fri, 13 Nov 2015 12:33:51 +0800
Subject: [PATCH 47/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 1 -
 1 file changed, 1 deletion(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 4326c32..c99e94e 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -107,7 +107,6 @@ with tf.device('/gpu:2'):
   b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2], name='b')
   c = tf.matmul(a, b)
 # 新建session with log_device_placement 设置为 True.
-# to True.
 sess = tf.Session(config=tf.ConfigProto(
       allow_soft_placement=True, log_device_placement=True))
 # 运行这个op.

From f6934e896cf62744e031dd6edd6769e8cd42750b Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Fri, 13 Nov 2015 12:34:42 +0800
Subject: [PATCH 48/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index c99e94e..2599ff7 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -21,7 +21,7 @@
 a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[2, 3], name='a')
 b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2], name='b')
 c = tf.matmul(a, b)
-# 新建session with log_device_placement 设置为 True.
+# 新建`session` with log_device_placement 设置为 True.
 sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
 # 运行这个op.
 print sess.run(c)

From c5ae357a1753b403d54aac64c2a4a14a85c8cb48 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Fri, 13 Nov 2015 12:36:20 +0800
Subject: [PATCH 49/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 2599ff7..115ab8f 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -21,7 +21,7 @@
 a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[2, 3], name='a')
 b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2], name='b')
 c = tf.matmul(a, b)
-# 新建`session` with log_device_placement 设置为 True.
+# 新建session with log_device_placement并设置为 True.
 sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
 # 运行这个op.
 print sess.run(c)
@@ -52,7 +52,7 @@ with tf.device('/cpu:0'):
   a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[2, 3], name='a')
   b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2], name='b')
 c = tf.matmul(a, b)
-# 新建session with log_device_placement 设置为 True.
+# 新建session with log_device_placement并设置为 True.
 sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
 # 运行这个op.
 print sess.run(c)
@@ -82,7 +82,7 @@ with tf.device('/gpu:2'):
   a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[2, 3], name='a')
   b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2], name='b')
   c = tf.matmul(a, b)
-# 新建session with log_device_placement 设置为 True.
+# 新建session with log_device_placement并设置为 True.
 sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
 # 运行这个op.
 print sess.run(c)
@@ -98,7 +98,7 @@ Could not satisfy explicit device specification '/gpu:2'
 ```
 
 如果你希望TensorFlow自动选择一个存在的且被支持的设备以防你指定的设备不存在，
-你可以在创建的`session`里把`allow_soft_placement`设置为`True`
+你可以在创建的`session`里把参数`allow_soft_placement`设置为`True`
 
 ```python
 # 新建一个graph.
@@ -106,7 +106,7 @@ with tf.device('/gpu:2'):
   a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[2, 3], name='a')
   b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2], name='b')
   c = tf.matmul(a, b)
-# 新建session with log_device_placement 设置为 True.
+# 新建session with log_device_placement并设置为True.
 sess = tf.Session(config=tf.ConfigProto(
       allow_soft_placement=True, log_device_placement=True))
 # 运行这个op.
@@ -128,7 +128,7 @@ for d in ['/gpu:2', '/gpu:3']:
     c.append(tf.matmul(a, b))
 with tf.device('/cpu:0'):
   sum = tf.add_n(c)
-# 新建session with log_device_placement 设置为 True.
+# 新建session with log_device_placement并设置为 True.
 sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
 # 运行这个op.
 print sess.run(sum)

From 151929fa22d9f58d409afe112aaba66ff38b2e12 Mon Sep 17 00:00:00 2001
From: lianghyv <1436349140@qq.com>
Date: Fri, 13 Nov 2015 12:36:54 +0800
Subject: [PATCH 50/55] Update using_gpu.md

---
 SOURCE/how_tos/using_gpu.md | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/SOURCE/how_tos/using_gpu.md b/SOURCE/how_tos/using_gpu.md
index 115ab8f..b47b90e 100755
--- a/SOURCE/how_tos/using_gpu.md
+++ b/SOURCE/how_tos/using_gpu.md
@@ -21,7 +21,7 @@
 a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[2, 3], name='a')
 b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2], name='b')
 c = tf.matmul(a, b)
-# 新建session with log_device_placement并设置为 True.
+# 新建session with log_device_placement并设置为True.
 sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
 # 运行这个op.
 print sess.run(c)
@@ -52,7 +52,7 @@ with tf.device('/cpu:0'):
   a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[2, 3], name='a')
   b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2], name='b')
 c = tf.matmul(a, b)
-# 新建session with log_device_placement并设置为 True.
+# 新建session with log_device_placement并设置为True.
 sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
 # 运行这个op.
 print sess.run(c)
@@ -82,7 +82,7 @@ with tf.device('/gpu:2'):
   a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[2, 3], name='a')
   b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2], name='b')
   c = tf.matmul(a, b)
-# 新建session with log_device_placement并设置为 True.
+# 新建session with log_device_placement并设置为True.
 sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
 # 运行这个op.
 print sess.run(c)
@@ -128,7 +128,7 @@ for d in ['/gpu:2', '/gpu:3']:
     c.append(tf.matmul(a, b))
 with tf.device('/cpu:0'):
   sum = tf.add_n(c)
-# 新建session with log_device_placement并设置为 True.
+# 新建session with log_device_placement并设置为True.
 sess = tf.Session(config=tf.ConfigProto(log_device_placement=True))
 # 运行这个op.
 print sess.run(sum)

From 550890b48d35e5836ac121dc1b69b6ce3955fa6d Mon Sep 17 00:00:00 2001
From: floydzhang1984 <tech_universe@163.com>
Date: Fri, 13 Nov 2015 13:35:18 +0800
Subject: [PATCH 51/55] =?UTF-8?q?=E5=90=88=E5=B9=B6=E7=BF=BB=E8=AF=91?=
 =?UTF-8?q?=E3=80=81=E6=A0=A1=E5=AF=B9=E7=9B=AE=E5=BD=95?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 README.md | 69 ++++++++++++++-----------------------------------------
 1 file changed, 17 insertions(+), 52 deletions(-)

diff --git a/README.md b/README.md
index 1c90cdc..460f407 100644
--- a/README.md
+++ b/README.md
@@ -62,33 +62,33 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
 
 ## 参与者（按认领章节排序）
 
-### 翻译
+### 翻译 & 校对
 
 - 起步
-  - [介绍](SOURCE/get_started/introduction.md) （[此处添加翻译者姓名及链接](https://github.com/xxx)）（[@PFZheng](https://github.com/PFZheng)）
-  - [下载及安装](SOURCE/get_started/os_setup.md) （[@PFZheng](https://github.com/PFZheng)）
-  - [基本用法](SOURCE/get_started/basic_usage.md) （[@PFZheng](https://github.com/PFZheng)）
+  - [介绍](SOURCE/get_started/introduction.md) 翻译：（[@PFZheng](https://github.com/PFZheng)）校对：（[@yangtze](https://github.com/sstruct)）
+  - [下载及安装](SOURCE/get_started/os_setup.md) 翻译：（[@PFZheng](https://github.com/PFZheng)）校对： （[@yangtze](https://github.com/sstruct)）
+  - [基本用法](SOURCE/get_started/basic_usage.md) 翻译：（[@PFZheng](https://github.com/PFZheng)）校对：（[@yangtze](https://github.com/sstruct)）
 - 教程
-  - [Overview](SOURCE/tutorials/overview.md) （[@PFZheng](https://github.com/PFZheng)）
-  - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md) ([@Tony Jin](https://github.com/linbojin))
-  - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md)([@chenweican](https://github.com/chenweican))
-  - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md)([@bingjin](https://github.com/bingjin))
-  - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md)([@oskycar](https://github.com/oskycar))
-  - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)（[@xyang40](https://github.com/xyang40)）
-  - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)([@Warln](https://github.com/Warln))
-  - [Mandelbrot集合](SOURCE/tutorials/mandelbrot.md)([@ericxk](https://githu.com/ericxk))√
+  - [Overview](SOURCE/tutorials/overview.md) 翻译：（[@PFZheng](https://github.com/PFZheng)）
+  - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md) 翻译：([@Tony Jin](https://github.com/linbojin))
+  - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md) 翻译：([@chenweican](https://github.com/chenweican))
+  - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md) 翻译：([@bingjin](https://github.com/bingjin))
+  - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md) 翻译： ([@oskycar](https://github.com/oskycar)) 校对： ([@zhyhooo](https://github.com/zhyhooo))
+  - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)翻译：（[@xyang40](https://github.com/xyang40)）
+  - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md) 翻译：([@Warln](https://github.com/Warln))
+  - [Mandelbrot集合](SOURCE/tutorials/mandelbrot.md) 翻译：([@ericxk](https://githu.com/ericxk))√
   - [Partial Differential Equations](SOURCE/tutorials/pdes.md) 
-  - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)([@JoyLiu](https://github.com/fengsehng))
+  - [MNIST Data Download](SOURCE/tutorials/mnist_download.md) 翻译： ([@JoyLiu](https://github.com/fengsehng))
 - 运作方式
   - [总览](SOURCE/how_tos/overview.md) 
-  - [变量:创建、初始化、保存和加载](SOURCE/how_tos/variables.md) ([@zhyhooo](https://github.com/zhyhooo))
+  - [变量:创建、初始化、保存和加载](SOURCE/how_tos/variables.md) 翻译： ([@zhyhooo](https://github.com/zhyhooo))
   - [TensorBoard:可视化学习](SOURCE/how_tos/summaries_and_tensorboard.md) ([@thylaco1eo](https://github.com/thylaco1eo))
   - [TensorBoard:图表可视化](SOURCE/how_tos/graph_viz.md) 
   - [读取数据](SOURCE/how_tos/reading_data.md) 
   - [线程和队列](SOURCE/how_tos/threading_and_queues.md) 
   - [添加新的Op](SOURCE/how_tos/adding_an_op.md) 
   - [自定义数据读取](SOURCE/how_tos/new_data_formats.md) 
-  - [使用gpu](SOURCE/how_tos/using_gpu.md) ([@lianghyv](https://github.com/lianghyv))
+  - [使用gpu](SOURCE/how_tos/using_gpu.md)翻译：([@lianghyv](https://github.com/lianghyv))√ 
   - [共享变量](SOURCE/how_tos/variable_scope.md) 
 - 资源
   - [总览](SOURCE/resources/overview.md) 
@@ -98,46 +98,11 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [术语表](SOURCE/resources/glossary.md)
   - [Tensor排名、形状和类型](SOURCE/resources/dims_types.md)
 
-### 校对
-
-- 起步
-  - [介绍](SOURCE/get_started/introduction.md) （[此处添加审校者姓名及链接](https://github.com/xxx)）（[@yangtze](https://github.com/sstruct)）
-  - [下载及安装](SOURCE/get_started/os_setup.md)（[@yangtze](https://github.com/sstruct)）
-  - [基本用法](SOURCE/get_started/basic_usage.md)（[@yangtze](https://github.com/sstruct)）
-- 教程
-  - [Overview](SOURCE/tutorials/overview.md)
-  - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md)
-  - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md)
-  - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md)
-  - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md) ([@zhyhooo](https://github.com/zhyhooo))
-  - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)
-  - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)
-  - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md)
-  - [Mandelbrot Set](SOURCE/tutorials/mandelbrot.md)
-  - [Partial Differential Equations](SOURCE/tutorials/pdes.md) 
-  - [MNIST Data Download](SOURCE/tutorials/mnist_download.md)
-- 运作方式
-  - [总览](SOURCE/how_tos/overview.md) 
-  - [变量:创建、初始化、保存和加载](SOURCE/how_tos/variables.md) 
-  - [TensorBoard:可视化学习](SOURCE/how_tos/summaries_and_tensorboard.md) 
-  - [TensorBoard:图表可视化](SOURCE/how_tos/graph_viz.md) 
-  - [读取数据](SOURCE/how_tos/reading_data.md) 
-  - [线程和队列](SOURCE/how_tos/threading_and_queues.md) 
-  - [添加新的Op](SOURCE/how_tos/adding_an_op.md) 
-  - [自定义数据读取](SOURCE/how_tos/new_data_formats.md) 
-  - [使用gpu](SOURCE/how_tos/using_gpu.md) 
-  - [共享变量](SOURCE/how_tos/variable_scope.md) 
-- 资源
-  - [总览](SOURCE/resources/overview.md) 
-  - [BibTex 引用](SOURCE/resources/bib.md) 
-  - [示例使用](SOURCE/resources/uses.md)
-  - [FAQ](SOURCE/resources/faq.md)
-  - [术语表](SOURCE/resources/glossary.md)
-  - [Tensor排名、形状和类型](SOURCE/resources/dims_types.md)
 
 ## 进度记录
 
-- 2015-11-10, 谷歌发布全新人工智能系统TensorFlow并宣布开源, 极客学院Wiki启动协同翻译，创建 GitHub 仓库，制定协同规范  
+- 2015-11-10, 谷歌发布全新人工智能系统TensorFlow并宣布开源, 极客学院Wiki启动协同翻译，创建 GitHub 仓库，制定协同规范 
+ 
 
 
 ## 感谢支持   

From 9c1d77503c514bac3e622c1c715738edf2eff2c8 Mon Sep 17 00:00:00 2001
From: floydzhang1984 <tech_universe@163.com>
Date: Fri, 13 Nov 2015 13:36:23 +0800
Subject: [PATCH 52/55] commit

---
 README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 460f407..2959f18 100644
--- a/README.md
+++ b/README.md
@@ -82,7 +82,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
 - 运作方式
   - [总览](SOURCE/how_tos/overview.md) 
   - [变量:创建、初始化、保存和加载](SOURCE/how_tos/variables.md) 翻译： ([@zhyhooo](https://github.com/zhyhooo))
-  - [TensorBoard:可视化学习](SOURCE/how_tos/summaries_and_tensorboard.md) ([@thylaco1eo](https://github.com/thylaco1eo))
+  - [TensorBoard:可视化学习](SOURCE/how_tos/summaries_and_tensorboard.md) 翻译：([@thylaco1eo](https://github.com/thylaco1eo))
   - [TensorBoard:图表可视化](SOURCE/how_tos/graph_viz.md) 
   - [读取数据](SOURCE/how_tos/reading_data.md) 
   - [线程和队列](SOURCE/how_tos/threading_and_queues.md) 

From 19853626cf629081852acfef5ea3834b7aac87fc Mon Sep 17 00:00:00 2001
From: ericxu <feitongxiaoke@gmail.com>
Date: Fri, 13 Nov 2015 13:51:56 +0800
Subject: [PATCH 53/55] add JIAODUI

---
 README.md | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/README.md b/README.md
index 2959f18..7ed820a 100644
--- a/README.md
+++ b/README.md
@@ -69,14 +69,14 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [下载及安装](SOURCE/get_started/os_setup.md) 翻译：（[@PFZheng](https://github.com/PFZheng)）校对： （[@yangtze](https://github.com/sstruct)）
   - [基本用法](SOURCE/get_started/basic_usage.md) 翻译：（[@PFZheng](https://github.com/PFZheng)）校对：（[@yangtze](https://github.com/sstruct)）
 - 教程
-  - [Overview](SOURCE/tutorials/overview.md) 翻译：（[@PFZheng](https://github.com/PFZheng)）
-  - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md) 翻译：([@Tony Jin](https://github.com/linbojin))
+  - [Overview](SOURCE/tutorials/overview.md) 翻译：（[@PFZheng](https://github.com/PFZheng)）校对： ([@ericxk](https://github.com/ericxk))
+  - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md) 翻译：([@Tony Jin](https://github.com/linbojin)) 校对： ([@ericxk](https://githu.com/ericxk))
   - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md) 翻译：([@chenweican](https://github.com/chenweican))
   - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md) 翻译：([@bingjin](https://github.com/bingjin))
   - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md) 翻译： ([@oskycar](https://github.com/oskycar)) 校对： ([@zhyhooo](https://github.com/zhyhooo))
   - [Vector Representations of Words](SOURCE/tutorials/word2vec.md)翻译：（[@xyang40](https://github.com/xyang40)）
   - [Recurrent Neural Networks](SOURCE/tutorials/recurrent.md) 翻译：([@Warln](https://github.com/Warln))
-  - [Mandelbrot集合](SOURCE/tutorials/mandelbrot.md) 翻译：([@ericxk](https://githu.com/ericxk))√
+  - [Mandelbrot集合](SOURCE/tutorials/mandelbrot.md) 翻译：([@ericxk](https://github.com/ericxk))√
   - [Partial Differential Equations](SOURCE/tutorials/pdes.md) 
   - [MNIST Data Download](SOURCE/tutorials/mnist_download.md) 翻译： ([@JoyLiu](https://github.com/fengsehng))
 - 运作方式

From 6f7a589d81e3e1e60de103805b7a3928100dbdf2 Mon Sep 17 00:00:00 2001
From: ericxu <feitongxiaoke@gmail.com>
Date: Fri, 13 Nov 2015 13:54:28 +0800
Subject: [PATCH 54/55] add proofing

---
 README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 7ed820a..28d1302 100644
--- a/README.md
+++ b/README.md
@@ -70,7 +70,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
   - [基本用法](SOURCE/get_started/basic_usage.md) 翻译：（[@PFZheng](https://github.com/PFZheng)）校对：（[@yangtze](https://github.com/sstruct)）
 - 教程
   - [Overview](SOURCE/tutorials/overview.md) 翻译：（[@PFZheng](https://github.com/PFZheng)）校对： ([@ericxk](https://github.com/ericxk))
-  - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md) 翻译：([@Tony Jin](https://github.com/linbojin)) 校对： ([@ericxk](https://githu.com/ericxk))
+  - [MNIST For ML Beginners](SOURCE/tutorials/mnist_beginners.md) 翻译：([@Tony Jin](https://github.com/linbojin)) 校对： ([@ericxk](https://github.com/ericxk))
   - [Deep MNIST for Expert](SOURCE/tutorials/mnist_pros.md) 翻译：([@chenweican](https://github.com/chenweican))
   - [TensorFlow Mechanics 101](SOURCE/tutorials/mnist_tf.md) 翻译：([@bingjin](https://github.com/bingjin))
   - [Convolutional Neural Networks](SOURCE/tutorials/deep_cnn.md) 翻译： ([@oskycar](https://github.com/oskycar)) 校对： ([@zhyhooo](https://github.com/zhyhooo))

From 06eb78580578750ab5b41dc86c3fd39836ac7fa0 Mon Sep 17 00:00:00 2001
From: andyiac <andyiaczhang@gmail.com>
Date: Fri, 13 Nov 2015 16:11:43 +0800
Subject: [PATCH 55/55] Update README.md

---
 README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 28d1302..b80b42a 100644
--- a/README.md
+++ b/README.md
@@ -93,7 +93,7 @@ PS: 想探讨TensorFlow技术的可以加"TensorFlow技术交流群"：495115006
 - 资源
   - [总览](SOURCE/resources/overview.md) 
   - [BibTex 引用](SOURCE/resources/bib.md) 
-  - [示例使用](SOURCE/resources/uses.md)
+  - [示例使用](SOURCE/resources/uses.md) 翻译：([@andyiac](https://github.com/andyiac))
   - [FAQ](SOURCE/resources/faq.md)
   - [术语表](SOURCE/resources/glossary.md)
   - [Tensor排名、形状和类型](SOURCE/resources/dims_types.md)