tensorflow2.0
二流子学程序
这个作者很懒,什么都没留下…
展开
-
Tensorflow2.0基础-笔记-利用回调函数使用Tensorboard
import tensorflow as tf#导入数据(train_image,train_labels),(test_images,test_labels)=tf.keras.datasets.mnist.load_data()#数据预处理train_image=tf.expand_dims(train_image,-1)train_image=tf.cast(train_image/255,tf.float32)train_labels=tf.cast(train_labels...原创 2021-09-12 15:47:20 · 266 阅读 · 0 评论 -
Tensorflow2.0基础-笔记-自定义训练
import tensorflow as tf#这段是因为我笔记本显存不够加的from tensorflow.compat.v1 import ConfigProtofrom tensorflow.compat.v1 import InteractiveSessionconfig = ConfigProto()config.gpu_options.allow_growth = Truesession = InteractiveSession(config=config)#导入...原创 2021-09-12 15:29:01 · 455 阅读 · 1 评论 -
Tensorflow2.0基础-笔记-图像识别-猫狗数据集
import tensorflow as tfimport matplotlib.pyplot as pltimport numpy as np%matplotlib inlineimport globimage_filenames1=glob.glob('./DataSet/猫狗数据集_2000/dc_2000/train/cat/*.jpg')image_filenames2=glob.glob('./DataSet/猫狗数据集_2000/dc_2000/train/dog/*.jpg'.原创 2021-08-18 23:23:54 · 644 阅读 · 2 评论 -
Tensorflow2.0基础-笔记-文本序列-全连接神经网络
import tensorflow as tffrom tensorflow import kerasfrom tensorflow.keras import layersimport matplotlib.pyplot as plt%matplotlib inlinedata=keras.datasets.imdb #导入电影评分数据集max_word=10000(x_train,y_train),(x_test,y_test)=data.load_data(num_words=max.原创 2021-08-18 23:13:10 · 181 阅读 · 0 评论 -
Tensorflow2.0基础-笔记- 批标准化
批标准化介绍:在训练的时候,前面层训练参数的更新将导致后面层输入数据分布的变化。以网络第二层为例:网络的第二层输入,是由第一层的参数和input计算得到的,而第一层的参数在整个训练过程中一直在变化,因此必然会引起后面每一层输入数据分布的改变。假如激活函数是sigmoid函数,如果数据在梯度很小的区域,那么学习率就会很慢甚至陷入长时间的停滞。如果上一层传递下来的数据分布为如下图:那么将会出现梯度消失的问题,同样如果全分布在右侧的或会出现梯度爆炸的问题。为此提出了一种方法批标准化(BN...原创 2021-08-18 22:47:46 · 374 阅读 · 0 评论 -
Tensorflow2.0基础-笔记- CNN卫星图像识别-手写标签
import pathlibimport numpy as npimport tensorflow as tfimport matplotlib.pyplot as plt%matplotlib inlinedata_dir='./DataSet/airplane_lake'data_root=pathlib.Path(data_dir)for item in data_root.iterdir(): print(item)#上面这段是提取数据文件,pathlib.Path().原创 2021-08-05 23:33:02 · 424 阅读 · 0 评论 -
Tensorflow2.0基础-笔记- CNN进行Mnist数据分类
import tensorflow as tffrom tensorflow import kerasimport matplotlib.pyplot as plt%matplotlib inlineimport numpy as npfashion_mnist=keras.datasets.fashion_mnist(train_images,train_labels),(test_images,test_labels)=fashion_mnist.load_data()train_i.原创 2021-08-04 20:46:10 · 196 阅读 · 0 评论 -
Tensorflow2.0基础-笔记- 利用tf.data修改mnist多分类任务
import tensorflow as tfimport pandas as pdimport numpy as npimport matplotlib.pyplot as plt%matplotlib inline(train_image,train_label),(test_image,test_label)=tf.keras.datasets.fashion_mnist.load_data()train_image=train_image/255test_image=test_i.原创 2021-08-04 00:22:02 · 106 阅读 · 0 评论 -
Tensorflow2.0基础-笔记- tf.data模块使用
import tensorflow as tfimport numpy as npdataset=tf.data.Dataset.from_tensor_slices([1,2,3,4,5,6,7,8,9]) #对数组进行切片存储for e in dataset: print(e) #这样输出的是tf.Tensor类型for e in dataset: print(e.numpy.原创 2021-08-04 00:03:28 · 131 阅读 · 0 评论 -
Tensorflow2.0基础-笔记- softmax多分类任务
import tensorflow as tfimport pandas as pdimport numpy as npimport matplotlib.pyplot as plt%matplotlib inline#输入训练数据集和测试数据集(train_image,train_label),(test_image,test_label)=tf.keras.datasets.fashion_mnist.load_data()train_image.shape #.原创 2021-07-30 21:52:00 · 312 阅读 · 0 评论 -
Tensorflow2.0基础-笔记- 多层感知器
import tensorflow as tfimport pandas as pdimport matplotlib.pyplot as plt%matplotlib inlinedata=pd.read_csv('D:\JupyterFile\DataSet\Advertising.csv') #数据集为广告与收益的关系的数据plt.scatter(data.TV,data.sales) #利用绘图matplotlib.pyplot进行绘图,参数为x,y轴plt.scat.原创 2021-07-29 14:16:32 · 112 阅读 · 0 评论 -
Tensorflow2.0基础-笔记- 逻辑回归
import tensorflow as tfimport pandas as pdimport numpy as npimport matplotlib.pyplot as plt%matplotlib inlinedata=pd.read_csv('D:\JupyterFile\DataSet\credit-a.csv',header=None) #输入数据集data.head() .原创 2021-07-27 00:09:26 · 230 阅读 · 0 评论 -
Tensorflow2.0基础-笔记- 线性回归
线性回归import tensorflow as tfimport pandas as pdimport matplotlib.pyplot as plt%matplotlib inlinedata=pd.read_csv('./DataSet/income.csv') #读取数据集plt.scatter(data.Education,data.Income) #绘图x=data.Education #将数据传给变量x,y。x是特征,只有一个,y是标签。y=data.In..原创 2021-07-26 14:47:19 · 429 阅读 · 0 评论