ukakasu的博客

建议DNN配置名称配置权重初始化He initialization激活函数ELU归一化Batch Normalization正则化dropout优化器Adam学习速率调整None全连接层from tensorflow.contrib.layers import fully_connectedwith tf.name_scop...

使用TensorFlow Dataset 加速数据处理及训练https://github.com/aymericdamien/TensorFlow-Examples/blob/master/notebooks/5_DataManagement/tensorflow_dataset_api.ipynbimport tensorflow as tffrom tensorflow.examples...

tf.estimator.inputs.numpy_input_fndef numpy_input_fn(x, y=None, batch_size=128, num_epochs=1, shuffle=None, ...

应用TensorFlow高级API构建卷积神经网络两个卷积层，两个全连接层输入 [sample * 28 * 28 * 1 ] (灰度图)[ 28 * 28 1 ] --> (32个卷积核，每个大小551,sample方式卷积) --> [ 28 * 28 * 32] --> (池化 22 ，步长2)–> [14 *14 *32][ 14 * 14 32] -...

应用TensorFlow构建卷积神经网络两个卷积层，两个全连接层输入 [sample * 28 * 28 * 1 ] (灰度图)[ 28 * 28 1 ] --> (32个卷积核，每个大小551,sample方式卷积) --> [ 28 * 28 * 32] --> (池化 22 ，步长2)–> [14 *14 *32][ 14 * 14 32] ...

tf.layers.densetf.layers.dense( inputs, units, activation=None, use_bias=True, kernel_initializer=None, bias_initializer=tf.zeros_initializer(), kernel_regularizer=None, ...

应用TensorFlow构建lstmimport tensorflow as tffrom tensorflow.contrib import rnnfrom tensorflow.examples.tutorials.mnist import input_datamnist = input_data.read_data_sets("/tmp/data/", one_hot=True)...

应用TensorFlow高级api构建全连接神经网络https://github.com/aymericdamien/TensorFlow-Examples/blob/master/notebooks/3_NeuralNetworks/neural_network.ipynbimport numpy as npimport tensorflow as tffrom tensorflow....

应用TensorFlow构建全连接神经网络https://github.com/sjchoi86/tensorflow-101/blob/master/notebooks/mlp_mnist_xavier.ipynbimport numpy as npimport tensorflow as tffrom tensorflow.examples.tutorials.mnist import...

设计最优数据输入管道使用 prefetch 函数去重叠 数据读取器 和 数据消耗器的工作。推荐在输入管道的末端添加 prefetch(n) （n是batch size），以重叠 CPU 上的变换 及 GPU/TPU设备上的训练。 dataset = dataset.batch(batch_size=FLAGS.batch_size) dataset = dataset.pre...