Functions — github.com/Jack-Cherish/Deep-Learning

↓ 1 callersMethodbp_gradient

(self, sensitivity_array)

Tutorial/lesson-5/cnn.py:227

↓ 1 callersMethodbp_gradient

(self, sensitivity_array)

Tutorial/lesson-4/cnn.py:227

↓ 1 callersMethodbp_sensitivity_map

计算传递到上一层的sensitivity map sensitivity_array: 本层的sensitivity map activator: 上一层的激活函数

Tutorial/lesson-6/cnn.py:187

↓ 1 callersMethodbp_sensitivity_map

计算传递到上一层的sensitivity map sensitivity_array: 本层的sensitivity map activator: 上一层的激活函数

Tutorial/lesson-5/cnn.py:187

↓ 1 callersMethodbp_sensitivity_map

计算传递到上一层的sensitivity map sensitivity_array: 本层的sensitivity map activator: 上一层的激活函数

Tutorial/lesson-4/cnn.py:187

↓ 1 callersMethodcalc_delta

(self, delta_h, activator)

Tutorial/lesson-6/lstm.py:126

↓ 1 callersMethodcalc_delta

计算每个节点的delta

Tutorial/lesson-7/recursive.py:78

↓ 1 callersMethodcalc_delta

(self, sensitivity_array, activator)

Tutorial/lesson-5/rnn.py:48

↓ 1 callersMethodcalc_delta_k

根据k时刻的delta_h，计算k时刻的delta_f、 delta_i、delta_o、delta_ct，以及k-1时刻的delta_h

Tutorial/lesson-6/lstm.py:151

↓ 1 callersMethodcalc_delta_k

根据k+1时刻的delta计算k时刻的delta

Tutorial/lesson-5/rnn.py:58

↓ 1 callersMethodcalc_gradient

(self, x)

Tutorial/lesson-6/lstm.py:192

↓ 1 callersMethodcalc_gradient

计算梯度

Tutorial/lesson-3/bp.py:149

↓ 1 callersMethodcalc_gradient

内部函数，计算每个连接的梯度

Tutorial/lesson-3/bp.py:239

↓ 1 callersMethodcalc_gradient

计算每个节点权重的梯度，并将它们求和，得到最终的梯度

Tutorial/lesson-7/recursive.py:97

↓ 1 callersMethodcalc_gradient

(self)

Tutorial/lesson-5/rnn.py:69

↓ 1 callersMethodcalc_gradient_t

计算每个时刻t权重的梯度

Tutorial/lesson-6/lstm.py:242

↓ 1 callersMethodcalc_gradient_t

计算每个时刻t权重的梯度

Tutorial/lesson-5/rnn.py:81

↓ 1 callersMethodcalc_hidden_layer_delta

节点属于隐藏层时，根据式4计算delta

Tutorial/lesson-3/bp.py:51

↓ 1 callersMethodcalc_output_layer_delta

节点属于输出层时，根据式3计算delta

Tutorial/lesson-3/bp.py:60

↓ 1 callersMethodconcatenate

将各个树节点中的数据拼接成一个长向量

Tutorial/lesson-7/recursive.py:69

↓ 1 callersFunctioncorrect_ratio

(network)

Tutorial/lesson-3/fc.py:182

↓ 1 callersFunctioncrack_captcha

使用模型做预测 Parameters: captcha_image:数据 captcha_label:标签

Discuz/test.py:6

↓ 1 callersMethoddenorm

(self, vec)

Tutorial/lesson-3/fc.py:165

↓ 1 callersMethoddownload_discuz

下载验证码图片 Parameters: nums:下载的验证码图片数量

Discuz/get_discuz.py:28

↓ 1 callersMethoddump

打印层的信息

Tutorial/lesson-3/bp.py:131

↓ 1 callersFunctionevaluate

(network, test_data_set, test_labels)

Tutorial/lesson-3/mnist.py:106

↓ 1 callersMethodget_bias

(self)

Tutorial/lesson-6/cnn.py:113

↓ 1 callersMethodget_bias

(self)

Tutorial/lesson-5/cnn.py:113

↓ 1 callersMethodget_bias

(self)

Tutorial/lesson-4/cnn.py:113

↓ 1 callersMethodget_images

(self, txt_path)

Pytorch-Seg/lesson-4/dataset.py:28

↓ 1 callersFunctionget_logger

(LEVEL, log_file = None)

Pytorch-Seg/lesson-3/logger.py:3

↓ 1 callersFunctionget_max_index

(array)

Tutorial/lesson-6/cnn.py:26

↓ 1 callersFunctionget_max_index

(array)

Tutorial/lesson-5/cnn.py:26

↓ 1 callersFunctionget_max_index

(array)

Tutorial/lesson-4/cnn.py:26

↓ 1 callersMethodget_one_sample

内部函数，将图像转化为样本的输入向量

Tutorial/lesson-3/mnist.py:38

↓ 1 callersMethodget_picture

内部函数，从文件中获取图像

Tutorial/lesson-3/mnist.py:26

↓ 1 callersFunctionget_test_data_set

获得测试数据集

Tutorial/lesson-3/mnist.py:88

↓ 1 callersFunctionget_training_data_set

获得训练数据集

Tutorial/lesson-3/mnist.py:81

↓ 1 callersFunctionget_training_dataset

基于and真值表构建训练数据

Tutorial/lesson-2/perceptron.py:68

↓ 1 callersFunctionget_training_dataset

捏造5个人的收入数据

Tutorial/lesson-2/linear_unit.py:9

↓ 1 callersFunctionget_training_dataset

基于and真值表构建训练数据

Tutorial/lesson-1/perceptron.py:68

↓ 1 callersMethodgradient_check

梯度检查 network: 神经网络对象 sample_feature: 样本的特征 sample_label: 样本的标签

Tutorial/lesson-3/fc.py:119

↓ 1 callersFunctionmask_video

(input_video, output_video, mask_path='mask.jpg')

face/video_mosaic.py:36

↓ 1 callersMethodnorm

内部函数，将一个值转换为10维标签向量

Tutorial/lesson-3/mnist.py:69

↓ 1 callersMethodpadding_black

(self, img)

Pytorch-Seg/lesson-4/dataset.py:34

↓ 1 callersMethodrandom_captcha_text

验证码一般都无视大小写；验证码长度4个字符 Parameters: captcha_size:验证码长度 Returns: captcha_text:验证码字符串

Discuz/get_discuz.py:10

↓ 1 callersMethodreset

(self)

Pytorch-Seg/lesson-4/train.py:163

↓ 1 callersFunctionsave_checkpoint

根据 is_best 存模型，一般保存 valid acc 最好的模型

Pytorch-Seg/lesson-4/train.py:39

↓ 1 callersFunctionsigmoid

(inX)

Tutorial/lesson-3/bp.py:6

↓ 1 callersFunctionsoftmax

(x)

Pytorch-Seg/lesson-4/infer.py:13

↓ 1 callersFunctiontrain

训练代码参数： train_loader - 训练集的 DataLoader model - 模型 criterion - 损失函数 optimizer - 优化器

Pytorch-Seg/lesson-4/train.py:47

↓ 1 callersMethodtrain

训练函数 labels: 样本标签 data_set: 输入样本 rate: 学习速率 epoch: 训练轮数

Tutorial/lesson-3/fc.py:84

↓ 1 callersMethodtrain

输入训练数据：一组向量、与每个向量对应的label；以及训练轮数、学习率

Tutorial/lesson-1/perceptron.py:32

↓ 1 callersFunctiontrain_and_evaluate

()

Tutorial/lesson-3/mnist.py:116

↓ 1 callersFunctiontrain_and_perceptron

使用and真值表训练感知器

Tutorial/lesson-2/perceptron.py:79

↓ 1 callersFunctiontrain_and_perceptron

使用and真值表训练感知器

Tutorial/lesson-1/perceptron.py:79

↓ 1 callersMethodtrain_crack_captcha_cnn

训练函数

Discuz/train.py:298

↓ 1 callersFunctiontrain_linear_unit

使用数据训练线性单元

Tutorial/lesson-2/linear_unit.py:19

↓ 1 callersFunctiontrain_net

(net, device, data_path, epochs=40, batch_size=1, lr=0.00001)

Pytorch-Seg/lesson-2/train.py:7

↓ 1 callersMethodtrain_one_sample

内部函数，用一个样本训练网络

Tutorial/lesson-3/bp.py:214

↓ 1 callersMethodtrain_one_sample

(self, label, sample, rate)

Tutorial/lesson-3/fc.py:96

↓ 1 callersMethodupdate

使用梯度下降算法更新权重

Tutorial/lesson-3/fc.py:46

↓ 1 callersMethodupdate_weight

内部函数，更新每个连接权重

Tutorial/lesson-3/bp.py:231

↓ 1 callersMethodupdate_weight

(self, rate)

Tutorial/lesson-3/fc.py:108

↓ 1 callersFunctionvalidate

验证代码参数： val_loader - 验证集的 DataLoader model - 模型 criterion - 损失函数 epoch - 进行第几个 epoch

Pytorch-Seg/lesson-4/train.py:105

↓ 1 callersFunctionvideo2mp3

将视频转为音频 :param file_name: 传入视频文件的路径 :return:

face/video_mosaic.py:12

↓ 1 callersFunctionvideo_add_mp3

视频添加音频 :param file_name: 传入视频文件的路径 :param mp3_file: 传入音频文件的路径 :return:

face/video_mosaic.py:24

↓ 1 callersFunctionweight_variable

(shape)

mnist.py:38

↓ 1 callersMethodwrite

(self, message)

Pytorch-Seg/lesson-3/log.py:8

Method__getitem__

(self, index)

Pytorch-Seg/lesson-2/utils/dataset.py:19

Method__getitem__

(self, index)

Pytorch-Seg/lesson-4/dataset.py:52

Method__init__

(self)

Discuz/train.py:8

Method__init__

(self)

Discuz/get_discuz.py:6

Method__init__

(self, width, height, depth)

Tutorial/lesson-6/cnn.py:98

Method__init__

(self, input_width, input_height, channel_number, filter_width, filter_hei

Tutorial/lesson-6/cnn.py:122

Method__init__

(self, input_width, input_height, channel_number, filter_width, filter_hei

Tutorial/lesson-6/cnn.py:273

Method__init__

(self, input_width, state_width, learning_rate)

Tutorial/lesson-6/lstm.py:8

Method__init__

构造节点对象。 layer_index: 节点所属的层的编号 node_index: 节点的编号

Tutorial/lesson-3/bp.py:11

Method__init__

构造节点对象。 layer_index: 节点所属的层的编号 node_index: 节点的编号

Tutorial/lesson-3/bp.py:76

Method__init__

初始化一层 layer_index: 层编号 node_count: 层所包含的节点个数

Tutorial/lesson-3/bp.py:108

Method__init__

初始化连接，权重初始化为是一个很小的随机数 upstream_node: 连接的上游节点 downstream_node: 连接的下游节点

Tutorial/lesson-3/bp.py:139

Method__init__

(self)

Tutorial/lesson-3/bp.py:177

Method__init__

初始化一个全连接神经网络 layers: 二维数组，描述神经网络每层节点数

Tutorial/lesson-3/bp.py:186

Method__init__

构造函数 input_size: 本层输入向量的维度 output_size: 本层输出向量的维度 activator: 激活函数

Tutorial/lesson-3/fc.py:6

Method__init__

构造函数

Tutorial/lesson-3/fc.py:62

Method__init__

(self)

Tutorial/lesson-3/fc.py:156

Method__init__

初始化加载器 path: 数据文件路径 count: 文件中的样本个数

Tutorial/lesson-3/mnist.py:7

Method__init__

(self, data, children=[], children_data=[])

Tutorial/lesson-7/recursive.py:10

Method__init__

递归神经网络构造函数 node_width: 表示每个节点的向量的维度 child_count: 每个父节点有几个子节点 activator: 激活函数对象 learning_rate: 梯度下降算法学习率

Tutorial/lesson-7/recursive.py:20

Method__init__

初始化感知器，设置输入参数的个数，以及激活函数。激活函数的类型为double -> double

Tutorial/lesson-2/perceptron.py:7

Method__init__

初始化线性单元，设置输入参数的个数

Tutorial/lesson-2/linear_unit.py:5

Method__init__

(self, width, height, depth)

Tutorial/lesson-5/cnn.py:98

Method__init__

(self, input_width, input_height, channel_number, filter_width, filter_hei

Tutorial/lesson-5/cnn.py:122

Method__init__

(self, input_width, input_height, channel_number, filter_width, filter_hei

Tutorial/lesson-5/cnn.py:273

Method__init__

(self, input_width, state_width, activator, learning_rate)

Tutorial/lesson-5/rnn.py:9

Method__init__

初始化感知器，设置输入参数的个数，以及激活函数。激活函数的类型为double -> double

Tutorial/lesson-1/perceptron.py:7

Method__init__

(self, width, height, depth)

Tutorial/lesson-4/cnn.py:98

Method__init__

(self, input_width, input_height, channel_number, filter_width, filter_hei

Tutorial/lesson-4/cnn.py:122

Method__init__

(self, input_width, input_height, channel_number, filter_width, filter_hei

Tutorial/lesson-4/cnn.py:273

Method__init__

(self, filename="log.txt")

Pytorch-Seg/lesson-3/log.py:4

Functions315 in github.com/Jack-Cherish/Deep-Learning