hzy46 / deep-learning-21-examples Goto Github PK

《21个项目玩转深度学习———基于TensorFlow的实践详解》配套代码

Python 91.35% HCL 0.15% HTML 1.59% Jupyter Notebook 2.33% Makefile 0.03% Shell 1.11% MATLAB 0.67% CSS 0.44% JavaScript 2.10% Dockerfile 0.23%

deep-learning-21-examples's Introduction

Deep-Learning-21-Examples

本工程是《21个项目玩转深度学习———基于TensorFlow的实践详解》的配套代码，代码推荐的运行环境为：Ubuntu 14.04，Python 2.7、TensorFlow >= 1.4.0。请尽量使用类UNIX系统和Python 2运行本书的代码。

欢迎有兴趣的同学猛戳下面的链接购买：

京东：https://item.jd.com/12327358.html
天猫：https://detail.tmall.com/item.htm?id=566399074299
当当：http://product.dangdang.com/25245282.html

写作这本书的初衷是希望提供一个深度学习实践案例的合集，让每个人都可以在“动手实验”的过程中，比较轻松地掌握深度学习和TensorFlow的使用技巧，并且能完整地做出一些有意思的项目。希望大家能在实践的过程中找到乐趣，如果有任何问题，欢迎通过issues提出纠错或改进建议。

快速指引

以下是各章代码详细的运行方法：

本书包含的项目主要有三部分：第1～11章主要介绍CNN相关的项目，其中8~11章为GAN模型和它的几个重要变体；第12～17章主要介绍RNN、LSTM 相关的项目；第18～21章主要介绍强化学习相关的项目。

各章data文件夹下载地址

本工程主要包含的是项目的源代码文件，以chapter_1/ ~ chapter_21/ 标识。在有的章节中，还会额外需要一些数据集、模型等文件，我们用chapter_<章节名>_data/来表示。

例如：第四章有chapter_4/和chapter_4_data/两个文件夹，前者是源代码，后者是在运行程序时会用的数据文件。

由于数据文件通常比较大，建议读者在用到时候再进行下载，下载地址为：https://pan.baidu.com/s/1i7pKvFf，提取码：1kmf。

其他链接

意见反馈邮箱：hzydl21 [at] 163.com
我的专栏博客：AI Insight

致谢

感谢电子工业出版社的孙学瑛女士和宋亚东先生为本书付出的努力，同时也感谢刘婧源同学给本书的宝贵建议。本书对于来自于网络的代码，在书中和配套代码的Readme文件中，均作了引用说明，在此也对原作者表示深深的感谢。

deep-learning-21-examples's People

Contributors

Stargazers

Watchers

Forkers

beabetterman mldl sunshinelu sumeng123 simplejian wudahht liferlisiqi apiguo dongchaoanderuni bestbyte yyuhanggg ys2518 yzk-yzk qinchangping duboya ziqiangchen trestea verylove aiyoubug huosan0123 wuqixiaobai kur-collect chisyliu e271141 james97 canyuchen ltttiantian wojiaoziyou whyou5945 sunshinezhihuo turpure wwxfromtju zgd716 zpeng1989 jluxcs angzz binwangh leofengxin luoyexuge fantasticliar jodejoester yjjiang11 chaoongithub fxlou xbigot allensmile dylan891111 lanjianchun ccccchar charlieyqin a-li-peng binyi10 wangle1218 bluce raivien muyuanyang sunsj2014 corefan zhongkailv hezhipengzipp hellogiantman1989 codewanghust xanderchang carlpc yangvict yuconan rongbohui dgq2011 hhaandroid zhangtaozhir shenqidewo sps1204 yemika wanjinchang ckqsars hcna-ai shaunstanislauslau viella4585 chengyao0929 n-ever-ever cclauss coderpreacher wanfengkai lwdgit brucew91 zer0zh1 wintersmilesb101 lillianfly simon137 coldplayplay zhangbo-tj cringyman haohaohaohaohaohaozhang ivonnechen yifengyiye vivienzou1 cynthiaulee killvxk pkuswg sidney1994

deep-learning-21-examples's Issues

第3章 3.3.4的训练一定要GPU参与吗？

chapter_3/slim# python train_image_classifier.py --train_dir=satellite/train_dir --dataset_name=satellite --dataset_split_name=train --dataset_dir=satellite/data --model_name=inception_v3 --checkpoint_path=satellite/pretrained/inception_v3.ckpt --checkpoint_exclude_scopes=InceptionV3/Logits,InceptionV3/AuxLogits --trainable_scopes=InceptionV3/Logits,InceptionV3/AuxLogits --max_number_of_steps=100000 --batch_size=32 --learning_rate=0.001 --learning_rate_decay_type=fixed --save_interval_secs=300 --save_summaries_secs=2 --log_every_n_steps=10 --optimizer=rmsprop --weight_decay=0.00004
WARNING:tensorflow:From train_image_classifier.py:398: create_global_step (from tensorflow.contrib.framework.python.ops.variables) is deprecated and will be removed in a future version.
Instructions for updating:
Please switch to tf.train.create_global_step
INFO:tensorflow:Summary name /clone_loss is illegal; using clone_loss instead.
INFO:tensorflow:Fine-tuning from satellite/pretrained/inception_v3.ckpt
2018-05-31 22:46:18.586097: I tensorflow/core/platform/cpu_feature_guard.cc:137] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 AVX512F FMA
INFO:tensorflow:Error reported to Coordinator: <class 'tensorflow.python.framework.errors_impl.InvalidArgumentError'>, Cannot assign a device for operation 'InceptionV3/AuxLogits/Conv2d_2b_1x1/biases/RMSProp_1': Operation was explicitly assigned to /device:GPU:0 but available devices are [ /job:localhost/replica:0/task:0/device:CPU:0 ]. Make sure the device specification refers to a valid device.
[[Node: InceptionV3/AuxLogits/Conv2d_2b_1x1/biases/RMSProp_1 = VariableV2_class=["loc:@InceptionV3/AuxLogits/Conv2d_2b_1x1/biases"], container="", dtype=DT_FLOAT, shape=[6], shared_name="", _device="/device:GPU:0"]]

Caused by op u'InceptionV3/AuxLogits/Conv2d_2b_1x1/biases/RMSProp_1', defined at:
File "train_image_classifier.py", line 573, in
tf.app.run()
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/platform/app.py", line 48, in run
_sys.exit(main(_sys.argv[:1] + flags_passthrough))
File "train_image_classifier.py", line 540, in main
global_step=global_step)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/optimizer.py", line 474, in apply_gradients
self._create_slots([_get_variable_for(v) for v in var_list])
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/rmsprop.py", line 111, in _create_slots
self._zeros_slot(v, "momentum", self._name)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/optimizer.py", line 796, in _zeros_slot
named_slots[_var_key(var)] = slot_creator.create_zeros_slot(var, op_name)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/slot_creator.py", line 174, in create_zeros_slot
colocate_with_primary=colocate_with_primary)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/slot_creator.py", line 148, in create_slot_with_initializer
dtype)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/slot_creator.py", line 67, in _create_slot_var
validate_shape=validate_shape)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 1203, in get_variable
constraint=constraint)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 1092, in get_variable
constraint=constraint)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 425, in get_variable
constraint=constraint)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 394, in _true_getter
use_resource=use_resource, constraint=constraint)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variable_scope.py", line 805, in _get_single_variable
constraint=constraint)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variables.py", line 213, in init
constraint=constraint)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/variables.py", line 309, in _init_from_args
name=name)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/state_ops.py", line 133, in variable_op_v2
shared_name=shared_name)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/gen_state_ops.py", line 927, in _variable_v2
shared_name=shared_name, name=name)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/op_def_library.py", line 787, in _apply_op_helper
op_def=op_def)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/ops.py", line 2956, in create_op
op_def=op_def)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/ops.py", line 1470, in init
self._traceback = self._graph._extract_stack() # pylint: disable=protected-access

InvalidArgumentError (see above for traceback): Cannot assign a device for operation 'InceptionV3/AuxLogits/Conv2d_2b_1x1/biases/RMSProp_1': Operation was explicitly assigned to /device:GPU:0 but available devices are [ /job:localhost/replica:0/task:0/device:CPU:0 ]. Make sure the device specification refers to a valid device.
[[Node: InceptionV3/AuxLogits/Conv2d_2b_1x1/biases/RMSProp_1 = VariableV2_class=["loc:@InceptionV3/AuxLogits/Conv2d_2b_1x1/biases"], container="", dtype=DT_FLOAT, shape=[6], shared_name="", _device="/device:GPU:0"]]

Traceback (most recent call last):
File "train_image_classifier.py", line 573, in
tf.app.run()
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/platform/app.py", line 48, in run
_sys.exit(main(_sys.argv[:1] + flags_passthrough))
File "train_image_classifier.py", line 569, in main
sync_optimizer=optimizer if FLAGS.sync_replicas else None)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/contrib/slim/python/slim/learning.py", line 742, in train
master, start_standard_services=False, config=session_config) as sess:
File "/usr/lib/python2.7/contextlib.py", line 17, in enter
return self.gen.next()
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/supervisor.py", line 964, in managed_session
self.stop(close_summary_writer=close_summary_writer)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/supervisor.py", line 792, in stop
stop_grace_period_secs=self._stop_grace_secs)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/coordinator.py", line 389, in join
six.reraise(*self._exc_info_to_raise)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/supervisor.py", line 953, in managed_session
start_standard_services=start_standard_services)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/supervisor.py", line 708, in prepare_or_wait_for_session
init_feed_dict=self._init_feed_dict, init_fn=self._init_fn)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/training/session_manager.py", line 279, in prepare_session
sess.run(init_op, feed_dict=init_feed_dict)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/client/session.py", line 889, in run
run_metadata_ptr)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/client/session.py", line 1120, in _run
feed_dict_tensor, options, run_metadata)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/client/session.py", line 1317, in _do_run
options, run_metadata)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/client/session.py", line 1336, in _do_call
raise type(e)(node_def, op, message)
tensorflow.python.framework.errors_impl.InvalidArgumentError: Cannot assign a device for operation 'InceptionV3/AuxLogits/Conv2d_2b_1x1/biases/RMSProp_1': Operation was explicitly assigned to /device:GPU:0 but available devices are [ /job:localhost/replica:0/task:0/device:CPU:0 ]. Make sure the device specification refers to a valid device.
[[Node: InceptionV3/AuxLogits/Conv2d_2b_1x1/biases/RMSProp_1 = VariableV2_class=["loc:@InceptionV3/AuxLogits/Conv2d_2b_1x1/biases"], container="", dtype=DT_FLOAT, shape=[6], shared_name="", _device="/device:GPU:0"]]

想问下第一章softmax回归的一个问题

softmax.py定义交叉熵的那行里面
cross_entropy = tf.reduce_mean(-tf.reduce_sum((y_ * tf.log(y))))
tf.reduce_sum函数在这里不用指明axis吗，不指明返回的是不是标量总和，怎么mean？
对python不太熟，麻烦指点一下

关于“词汇表”的问题。

你好，你十二章，如果我要训练自己的数据集，该怎样生成vocab文件呢！

Fine tune 更多的层？

想请教一下如何fine tune 更多的层还有如过换成其他的网络呢？
--checkpoint_path=satellite/pretrained/inception_v3.ckpt \ 这里我知道是加载之前训练好的模型权值
--checkpoint_exclude_scopes=InceptionV3/Logits,InceptionV3/AuxLogits \ 这里是不加载的层信息，但是这些信息是怎么来的，或者里面是说明代码，不太清楚。
--trainable_scopes=InceptionV3/Logits,InceptionV3/AuxLogits \如果换成其他网络是不是把“Inception”换成“Resnet”？？？？
如果想fine tune 之前更多的层呢？？
谢谢

第十二章的有代码部分失效

第十二章 BasicRNNCell和BasicLSTMCell已经失效，子类方法call也不能用了，应该是tensorflow更新了版本。目前还不确定有多少代码失效了，在测试中，能更改一下代码吗？谢谢

第三章目前遇到的几个问题

在训练的时候遇到的问题，都已经通过百度或者Google解决了，但是在验证模型测试的时候，报错了：
ResourceExhaustedError (see above for traceback): OOM when allocating tensor with shape[100,64,147,147]
这个问题应该是空间不够了，我的GPU是GTX960 2Gb，按说应该是可以的，不知道有遇到此问题的吗？
另外一个问题是在到处pb文件的时候，产生如下错误：
INFO:tensorflow:Scale of 0 disables regularizer.
百度没查到相应的解决方法。

第3章TFrecord相关代码——python3的兼容性问题

小白刚入门，写在这里也是给自己做个记录。
1.range与xrange rage()在python3返回值不再是list,这里不改会报错；
2.TypeError: 'RGB' has type class 'str', but expected one of: (class 'bytes',)
两种改法：第一种参见https://stackoverflow.com/questions/38546672/inception-build-imagenet-data-py-typeerror-rgb-has-type-class-str-but-ex
第二种参见https://lixingcong.github.io/2016/03/06/convert-data-in-python/

chapter-自动上色

首先感谢博主的书，问下博主有训练好的食物上色和动漫上色的模型嘛。
穷学生买不起gpu，cpu训练要好久好久，@。@。动漫的估计要十几天。。。
还留着模型的话给个百度链接可好。

第五章目标检测

在research目录下执行model_builder_test.py时直接报错 No module named 'object_detection'，请问大佬们有遇到吗？怎么解决呀

第6章人脸检测和人脸识别

你好，我想咨询下图6-6中R-Net的网络结构中，MP：3×3中MP代表什么含义

第17章python3上运行有几个地方需要改改

我搞了3个晚上终于才开始训练，在解压数据那里就被来来回回捣鼓了好多次，每次重新来就要等好久。
我把自己遇到的错误讲讲希望对像我这样的新手小白有用吧，大神就不用看了。
1.首先在bazel这步，我是在研究室的服务器的子用户，不能用sudo，所以我是按bazel官网上方法下载安装的。

2.bazel-bin/im2txt/download_and_preprocess_mscoco "data/mscoco" 在使用这句的时候，如果你是子用户，必须把整个路径都写上
我现在的位置是im2txt文件夹,mkdir建立一个data文件夹，然后pwd，路径复制粘贴，然后复制粘贴/download_and_preprocess_mscoco "data/mscoco"

3.第二步那句话运行之后，下载数据解压完后会调用build_mscoco_data.py来处理数据。这个文件有3处要改。
1)所有的xrange改成range
2)第213行 with tf.gfile.FastGFile(image.filename, "r") as f:
"r"改成"rb"
3)第189行return tf.train.Feature(bytes_list=tf.train.BytesList(value=[str(value)]))
改成return tf.train.Feature(bytes_list=tf.train.BytesList(value=[value.encode('utf-8') if type(value)==str else value]))

4.注意运行后面贴的那几句bazel-bin啥啥啥的前面也必须贴全路径地址。

以上都是基于我是子用户，不知道root是不是就不用写_(:з」∠)_希望对后来者有用吧。

章节一readme错误

Deep-Learning-21-Examples/chapter_1/
1.2.2 两层卷积网络分类

python softmax_regression.py 这里错误了吧，应该是 python convolutional.py

部分章节使用了很多 TensorFlow 高级 API，不利于学习

比如，楼主在 15 章中使用了很多高级 API，学习性很差。一个很简答的时历曲线学习，搞得很复杂，很不直观。

from tensorflow.contrib.timeseries.python.timeseries import estimators as ts_estimators
from tensorflow.contrib.timeseries.python.timeseries import model as ts_model
from tensorflow.contrib.timeseries.python.timeseries import  NumpyReader

TF 的 API 设计的繁复、变动性大，手册也垃圾，实际工作项目中，也是使用的高级 API 吗？

第15章运用LSTM进行单变量时序预测时遇到如下警告

如图，警告为
WARNING:tensorflow:Skipping summary for mean, must be a float, np.float32, np.int64, np.int32 or int or a serialized string of Summary.
WARNING:tensorflow:Skipping summary for observed, must be a float, np.float32, np.int64, np.int32 or int or a serialized string of Summary.
WARNING:tensorflow:Skipping summary for start_tuple, must be a float, np.float32, np.int64, np.int32 or int or a serialized string of Summary.
WARNING:tensorflow:Skipping summary for times, must be a float, np.float32, np.int64, np.int32 or int or a serialized string of Summary.
WARNING:tensorflow:Input graph does not use tf.data.Dataset or contain a QueueRunner. That means predict yields forever. This is probably a mistake.
请问如何解决？

有没有一个群或者论坛，让大家一块讨论

Chapter 9未提供数据集food_resized和anime_resized

9.4节提到“在 chapter_ 9_ data/ 中提供了两个已经整理好的数据集，分别保存为food_ resized. zip 和 anime_ resized. zip ”
但这两个文件未找到。

第三章转化为tfrecord格式错误

return tf.train.Feature(bytes_list=tf.train.BytesList(value=[value]))
TypeError: 'rock' has type str, but expected one of: bytes
您好，书是不错，可我用的python3.6 ，改了几处后还是有错，跑不了

第二章的tf.train.string_input_producer

tf.train.string_input_producer需要start_queue_runners来填充队列。
但是我在cifar10_train.py中没有找到start_queue_runners。
所以是我没有找到，还是有其他的方式来填充队列

16章的中英数据没看到啊

动漫人物头像数据集faces.zip

原文中说chapter_8_data/中已经准备好了一个动漫人物头像数据集faces.zip，没看到

chapter12:Traceback (most recent call last): File "sample.py", line 39, in <module> tf.app.run() File "C:\Program Files\Python35\lib\site-packages\tensorflow\python\platform\app.py", line 126, in run _sys.exit(main(argv)) File "sample.py", line 19, in main FLAGS.start_string = FLAGS.start_string.decode('utf-8') AttributeError: 'str' object has no attribute 'decode'

第一章在验证集上报告正确率用的数据依然是训练集上的数据.

    batch = mnist.train.next_batch(50)
    # 每100步报告一次在验证集上的准确度
    if i % 100 == 0:
        train_accuracy = accuracy.eval(feed_dict={
            x: batch[0], y_: batch[1], keep_prob: 1.0})
        print("step %d, training accuracy %g" % (i, train_accuracy))

这样修改下就是在验证集上报告正确率:
#定义验证集
validate_data_x = mnist.validation.images
validate_data_y = mnist.validation.labels
# 训练20000步
for i in range(20000):
batch = mnist.train.next_batch(50)
# 每100步报告一次在验证集上的准确度
if i % 100 == 0:
train_accuracy = accuracy.eval(feed_dict={
x: validate_data_x, y_: validate_data_y, keep_prob: 1.0})
print("step %d, training accuracy %g" % (i, train_accuracy))

关于word2vec的问题。

你好，我想问一下，有没有用word2vec生成文本的项目？

5.2.1 安装TensorFlow Object Detection API

前3步都是正确的，到了第4步执行“python object_detection/builders/model_builder_test.py”时，出现了
TypeError:new() got an unexpected keyword argument 'serialized_options'

建议在 chapter_3 增加支援Windows 命令行格式和非GPU 版 TF 的修改建议

10 undefined names in Chapter 20

flake8 testing of https://github.com/hzy46/Deep-Learning-21-Examples on Python 2.7.14

$ flake8 . --count --select=E901,E999,F821,F822,F823 --show-source --statistics

./chapter_20/agents/_async.py:16:11: F821 undefined name 'DeepQ'
    super(DeepQ, self).__init__(sess, pred_network, target_network, env, stat, conf)
          ^
./chapter_20/agents/_async.py:83:29: F821 undefined name 'state'
      action = self.predict(state)
                            ^
./chapter_20/agents/_async.py:96:25: F821 undefined name 'ep'
                        ep, q, loss, is_update, self.learning_rate_op)
                        ^
./chapter_20/agents/_async.py:96:29: F821 undefined name 'q'
                        ep, q, loss, is_update, self.learning_rate_op)
                            ^
./chapter_20/agents/_async.py:96:32: F821 undefined name 'loss'
                        ep, q, loss, is_update, self.learning_rate_op)
                               ^
./chapter_20/agents/_async.py:96:38: F821 undefined name 'is_update'
                        ep, q, loss, is_update, self.learning_rate_op)
                                     ^
./chapter_20/agents/_n_step_q.py:15:11: F821 undefined name 'DeepQ'
    super(DeepQ, self).__init__(sess, pred_network, target_network, env, stat, conf)
          ^
./chapter_20/agents/_n_step_q.py:72:24: F821 undefined name 'global_var'
            global_v = global_var[re.sub(r'.*\/A3C_\d+\/', '', var.name)]
                       ^
./chapter_20/agents/_n_step_q.py:72:35: F821 undefined name 're'
            global_v = global_var[re.sub(r'.*\/A3C_\d+\/', '', var.name)]
                                  ^
./chapter_20/agents/_sarsa.py:16:11: F821 undefined name 'DeepQ'
    super(DeepQ, self).__init__(sess, pred_network, target_network, policy_network, env, stat, conf)
          ^
10    F821 undefined name 'DeepQ'
10

运行第七章代码报错

跑已经生成好的模型python eval.py --model_file models/wave.ckpt-done --image_file img/test.jpg
和跑训练模型 python train.py -c conf/wave.yml
都报这个错
Traceback (most recent call last):
File "eval.py", line 75, in
tf.app.run()
File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/tensorflow/python/platform/app.py", line 48, in run
_sys.exit(main(_sys.argv[:1] + flags_passthrough))
File "eval.py", line 45, in main
generated = model.net(image, training=False)
File "/Users/yushiyang/project/Deep-Learning-21-Examples-master/chapter_7/model.py", line 100, in net
conv1 = relu(instance_norm(conv2d(image, 3, 32, 9, 1)))
File "/Users/yushiyang/project/Deep-Learning-21-Examples-master/chapter_7/model.py", line 9, in conv2d
x_padded = tf.pad(x, [[0, 0], [kernel / 2, kernel / 2], [kernel / 2, kernel / 2], [0, 0]], mode=mode)
File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/tensorflow/python/ops/array_ops.py", line 1738, in pad
tensor, paddings, mode="REFLECT", name=name)
File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/tensorflow/python/ops/gen_array_ops.py", line 2580, in _mirror_pad
"MirrorPad", input=input, paddings=paddings, mode=mode, name=name)
File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/tensorflow/python/framework/op_def_library.py", line 609, in _apply_op_helper
param_name=input_name)
File "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/tensorflow/python/framework/op_def_library.py", line 60, in _SatisfiesTypeConstraint
", ".join(dtypes.as_dtype(x).name for x in allowed_list)))
TypeError: Value passed to parameter 'paddings' has DataType float32 not in list of allowed values: int32, int64

21个项目玩转深度学习第5章目标检测

根据该书的第5章的内容，对数据进行训练，我的电脑配置GTX 1080TI 12G+64 ubuntu16.04lts，在训练过程中，监测GPU使用情况，运行容量沾满，导致电脑关机。请问如何解决这个问题，我是新手，能不能说的详细点

第三章中python data_convert.py,出错，我用的是python3，windows下的，要如何改啊

INFO:root:Saving results to pic/
INFO:root:Determining list of input files and labels from pic/validation.
INFO:root:Finished finding files in 0 of 200 classes.
Traceback (most recent call last):
File "F:\algorithm\python3\21个项目玩转深度学习\Deep-Learning-21-Examples-master\chapter_3\data_prepare\data_convert.py", line 35, in
main(args)
File "F:\algorithm\python3\21个项目玩转深度学习\Deep-Learning-21-Examples-master\chapter_3\data_prepare\src\tfrecord.py", line 409, in main
command_args.validation_shards, command_args.labels_file, command_args)
File "F:\algorithm\python3\21个项目玩转深度学习\Deep-Learning-21-Examples-master\chapter_3\data_prepare\src\tfrecord.py", line 361, in _process_dataset
filenames, texts, labels = _find_image_files(directory, labels_file, command_args)
File "F:\algorithm\python3\21个项目玩转深度学习\Deep-Learning-21-Examples-master\chapter_3\data_prepare\src\tfrecord.py", line 341, in _find_image_files
random.shuffle(shuffled_index)
File "F:\Python36\lib\random.py", line 275, in shuffle
x[i], x[j] = x[j], x[i]
TypeError: 'range' object does not support item assignment

第一章关于在验证集上报告正确率

    batch = mnist.train.next_batch(50)

     if i % 100 == 0:
        train_accuracy = accuracy.eval(feed_dict={
            x: batch[0], y_: batch[1], keep_prob: 1.0})
        print("step %d, training accuracy %g" % (i, train_accuracy))

似乎并没有使用验证集，依然是在训练集上做的啊？

第三章训练模型遇到的问题，训练过程卡住了，模型一直是model.ckpt-0

运行第三章3.3.4节开始训练的命令后，开始一直卡在INFO:tensorflow:global_step/sec=0，一直重复输出，一段时间后，然后就一直卡在INFO:tensorflow:S**ing checkpoint to path satellite/train_dir/model.ckpt，隔一段时间这个就重复输出，但是训练文件夹里保存的模型一直是model.ckpt-0，请问是什么原因？

第三章,卫星图片数据集的下载地址

  第三章的微信图片下载地址在哪?虽然书中有提供图片,但我想知道源地址,有谁能提供更多的图像训练集的下载地址,谢谢了

我有买你的书，但大哥你的百度网盘进不去阿

“在数据目录chapter_8_data/中”

由于数据文件通常比较大，建议读者在用到时候再进行下载，下载地址为：https://pan.baidu.com/s/1i7pKvFf
，提取码：1kmf。

Thanks

第七章的快速风格迁移

运行书上第一句命令时候报错如下:
TypeError: Value passed to parameter 'paddings' has DataType float32 not in list of allowed values: int32, int64，不知道怎么办。

请问项目是否能在PyTorch上复现？

请问这些项目是否都能方便地在PyTorch上复现？

Requirements:Hardware

Hi,
when training the data, the computer broke down. I guess the GPU memory is insufficient, so I want to know the Configuration requirements,

第三章，采用rennet_v1_50模型进行训练，已经训练完成，想导出模型对单张图像进行识别，最后的classfy_image_inception_v3.py出现错误，应该怎么修改

第三章，采用rennet_v1_50模型进行训练，已经训练完成，想导出模型对单张图像进行识别，最后的classfy_image_inception_v3.py出现错误，应该怎么修改呀？
KeyError: "The name 'resnet_v1/logits/SpatialSqueeze:0' refers to a Tensor which does not exist. The operation, 'resnet_v1/logits/SpatialSqueeze', does not exist in the graph."

第三章，导出模型 classify_image_inception_v3.py运行出问题

cmd: python export_inference_graph.py --alsologtostderr --model_name=inception_v3 --output_file=satellite/inception_v3_inf_graph.pb --dataset_name satellite
会报错：INFO:tensorflow:Scale of 0 disables regularizer.
请问怎么解决呀？？

第五章：深度学习中的目标检测

我的环境是win10,python3。在复现第五章的代码时，目前遇到了三个问题：
1：http://download.tensorflow.org登录不上，然后faster_rcnn_inception_resnet_v2_...那个模型我是通过其他博客找到的。

2：在运行train.py文件时
E:\Softwave\Anaconda\python.exe E:/Codes/Deep-Learning-21-Examples/Chapter_5/research/object_detection/train.py
Traceback (most recent call last):
File "E:/Codes/Deep-Learning-21-Examples/Chapter_5/research/object_detection/train.py", line 202, in
tf.app.run()
File "E:\Softwave\Anaconda\lib\site-packages\tensorflow\python\platform\app.py", line 124, in run
_sys.exit(main(argv))
File "E:/Codes/Deep-Learning-21-Examples/Chapter_5/research/object_detection/train.py", line 198, in main
worker_job_name, is_chief, FLAGS.train_dir)
File "E:/Codes/Deep-Learning-21-Examples/Chapter_5/research\object_detection\trainer.py", line 161, in train
detection_model = create_model_fn()
File "E:/Codes/Deep-Learning-21-Examples/Chapter_5/research\object_detection\builders\model_builder.py", line 77, in build
return _build_faster_rcnn_model(model_config.faster_rcnn, is_training)
File "E:/Codes/Deep-Learning-21-Examples/Chapter_5/research\object_detection\builders\model_builder.py", line 218, in _build_faster_rcnn_model
first_stage_only = frcnn_config.first_stage_only
AttributeError: 'FasterRcnn' object has no attribute 'first_stage_only'

Process finished with exit code 1

然后我的做法是将model_bulider.py文件第218和第222行改为了first_stage_only =False
3：在修改为first_stage_only =False后，运行train.py文件后： File "E:\Softwave\Anaconda\lib\site-packages\tensorflow\python\framework\tensor_util.py", line 328, in _AssertCompatible
(dtype.name, repr(mismatch), type(mismatch).name))
TypeError: Expected string, got ['VOC/pet_train.record'] of type 'RepeatedScalarFieldContainer' instead.

请问我该如何修改呢。。。
在复现时遇到了各种奇怪的问题，度娘谷歌了好久，本以为终于可以了，没想到。。。。。。

第二章2.1.4节相关代码似乎有误

filenames = [os.path.join(data_dir, 'data_batch_%d.bin' % i) for i in **xrange**(1, 6)] #amend to range

softmax_regression中批量梯度下降问题

写代码时，增加一个epoch的循环，最后的结果只有0.86。为什么大神里面就可以达到0.9多？

for step in range(500):
    x,y = mnist.train.next_batch(batch_size)
    for i in range(num):#num = mnist.train.num_examples/batch_size
        _,tmp = sess.run([optimizer,loss],{X:x,Y:y}) #optimizer=Adam.minimize(loss)
    #sess.run(optimizer,{X:x,Y:y})

print("training accuracy",accuracy.eval({X:mnist.train.images,Y:mnist.train.labels}))

chapter1---softmax_regression.py，损失函数定义有问题

chapter1---softmax_regression.py，line29~30损失函数定义有问题

根据y, yhat构造交叉熵损失

cross_entropy = tf.reduce_mean(-tf.reduce_sum(y * tf.log(yhat)))

cross_entropy = tf.reduce_mean(-tf.reduce_sum(y * tf.log(yhat),axis=1))

第八章，生成MNIST图像

Traceback (most recent call last):
File "main.py", line 103, in
tf.app.run()
File "D:\Anaconda3\envs\tensorflow-gpu\lib\site-packages\tensorflow\python\platform\app.py", line 48, in run
_sys.exit(main(_sys.argv[:1] + flags_passthrough))
File "main.py", line 65, in main
data_dir=FLAGS.data_dir)
File "D:\pyproject\study_TF_for21\chapter8\model.py", line 75, in init
self.data_X, self.data_y = self.load_mnist()
File "D:\pyproject\study_TF_for21\chapter8\model.py", line 468, in load_mnist
trX = loaded[16:].reshape((60000,28,28,1)).astype(np.float)
ValueError: cannot reshape array of size 9912406 into shape (60000,28,28,1)

这个怎么解决。。。windows环境

章节10 readme错误

第10章中合并图像操作好像是有问题的，应该是写的时候笔误

scipy.misc.toimage API新版已经被弃用

这里 https://github.com/hzy46/Deep-Learning-21-Examples/blob/master/chapter_1/save_pic.py#L25 使用的 API 已经被弃用

https://docs.scipy.org/doc/scipy/reference/generated/scipy.misc.toimage.html#scipy-misc-toimage
toimage is deprecated! toimage is deprecated in SciPy 1.0.0, and will be removed in 1.2.0. Use Pillow’s Image.fromarray directly instead.

这样是可以的

from PIL import Image
Image.fromarray((image_array*256).astype(np.uint8)).save(filename)

问题1:但是对于faster-rcnn来说Φ5(P)是什么
问题2:但是对于faster-rcnn来说wT是谁的参数呢