lpflpf5678的博客

import tensorflow as tffrom tensorflow.examples.tutorials.mnist import input_datamnist = input_data.read_data_sets("/root/download/inputdata",one_hot=True)batch_size = 50lr = 0.01lrd = 0.99Flage = tf.app.flags.FLAGStf.app.flags.DEFINE_integer("tra

import tensorflow as tfdef variable(): c=tf.get_variable("b",shape=[2],initializer=tf.constant_initializer(1.0)) w=tf.get_variable("w",shape=[3,2],initializer=tf.constant_initializer([[1,2],[2,3],[3,4]])) #在自定义函数内部调用会话，但是没有给变量初始化，程序报错。尽管在主程序中调

tensorflow中张量，变量，常量进行数值运算都可用符号“+ - * /”，为方便起见，tensorflow中提供了方法。如:加法add = tf.add(a, b)减法sub = tf.subtract(a, b)乘法mul = tf.multiply(a, b)除法div = tf.divide(a, b)包括矩阵、向量的运算以上是严格按照数学的计算形式，进行计算。在tensorflow中定义的一维向量（如tf.get_variable(“b”,shape=[3],in

weight = tf.get_variable("weight",shape=[5,5,1,32],initializer=tf.truncated_normal_initializer(stddev=stddev))bias = tf.get_variable("bias",initializer=tf.constant(b_constant,shape=[b_shape]))get_variable中，设置了shpae，那么initializer必须是数值，如果是张量，那么gei_variabl

import tensorflow as tfimport osbatch_size = 50Flages = tf.app.flags.FLAGStf.app.flags.DEFINE_integer("is_train",1,"train or test")class convolutin_picture(object): def __init__(self,filelist): self.filelist = filelist self.lab

import numpy as npn = np.array([[2,1,2,1], [0,-1,3,0], [2,1,-1,4], [1,2,5,7], [2,1,2,1], [0,1,3,0], [2,1,-1,4], [8,7,9,5], [6,3,2,1],

import tensorflow as tf#定义两个占位符x,y。x和y都是张量x = tf.placeholder(tf.float32)y = tf.placeholder(tf.float32)#自定义一个函数def train(x,y): with tf.Session() as sess: result = print(sess.run(x+y)) return result#调用自定义的函数，把占位符作为实参传递给函数c = train(x,y)

softmax_cross_entropy_with_logits(labels,logits)logitsandlabelsmust have the same shape[batch_size, num_classes]and the same dtype (eitherfloat16,float32, orfloat64`).sparse_softmax_cross_entropy_with_logits(labels,logits)A common use case is to have lo

import tensorflow as tfx = tf.Variable(0.0)#定义一个滑动平均值类averages_class = tf.train.ExponentialMovingAverage(0.99)#为变量列表中的每个变量创建一个影子变量。#tf.trainable_variables()返回集合图上中，没有指定trainable_variables= False的变量average_op = averages_class.apply(tf.trainable_var

import tensorflow as tfx = tf.Variable(0.0)a = tf.global_variables_initializer()with tf.Session() as sess:sess.run(a)y = tf.add(x, 10.0) **#tensorflow中定义的变量必须用tensorflow中提供的函数进行运算** **#得到的结果才能依然是变量**sess.run(tf.assign(x,y)

为什么这个时候可以尝试收集更多的数据来解决呢？因为，high variance意味着你的训练集和验证集的效果有很大差别，很可能是因为你的训练集里面的数据并不能很好地反映验证集里面的特征，或者说验证集中有一些需要学习到的东西在你的训练集中却没有，所以模型无论如何学习你的训练集，也无法很好地预测验证集的数据，因此我们应该收集更多的数据，学习更多的特征。如果你增加了数据量去训练之后，variance还是很大，那说明我们遇到了典型的 over fitting，我们就需要考虑用 正则化手段了。三、正则化（Regu

报错：FutureWarning: Passing (type, 1) or ‘1type’ as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / ‘(1,)type’.np_resource = np.dtype([(“resource”, np.ubyte, 1)])由于numpy版本过高导致的，安装低版本的numpy例如：pip inst

错误信息Cannot uninstall ‘numpy’. It is a distutils installed project and thus we cannot accurately determine which files belong to it which would lead to only a partial uninstall.Cannot uninstall ‘wrapt’. It is a distutils installed project and thus we can