深度学习-Tensorflow2.2-tf.data输入模块{2}-tf.data输入实例-10

# -*- coding: utf-8 -*-
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2' # 修改警告级别，不显示警告
import tensorflow as tf


# 下载数据集并划分为训练集和测试集
(train_images,train_lables),(test_images,test_labels) = tf.keras.datasets.mnist.load_data()
# 归一化
train_images = train_images/255
test_images = test_images/255
print(train_images.shape)
# 建立模型创建dataset
ds_train_img = tf.data.Dataset.from_tensor_slices(train_images)
print(ds_train_img)

# -*- coding: utf-8 -*-
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2' # 修改警告级别，不显示警告
import tensorflow as tf


# 下载数据集并划分为训练集和测试集
(train_images,train_lables),(test_images,test_labels) = tf.keras.datasets.mnist.load_data()
# 归一化
train_images = train_images/255
test_images = test_images/255
print(train_images.shape)
# 创建dataset
ds_train_img = tf.data.Dataset.from_tensor_slices(train_images)
print(ds_train_img)
ds_train_lab = tf.data.Dataset.from_tensor_slices(train_lables)
print(ds_train_lab)
# 合并(元组)
ds_train = tf.data.Dataset.zip((ds_train_img,ds_train_lab))
print(ds_train)
# 取出其中10000个组件进行乱序,无限重复每次输出64张图片
ds_train = ds_train.shuffle(10000).repeat().batch(64)

# 建立模型
model = tf.keras.Sequential([
    tf.keras.layers.Flatten(input_shape=(28,28)),
    tf.keras.layers.Dense(128,activation="relu"),
    tf.keras.layers.Dense(10,activation="softmax")
])
# 编译模型
model.compile(optimizer="adam",
              loss="sparse_categorical_crossentropy",
              metrics=["accuracy"])
steps_per_epochs = train_images.shape[0]//64 # 我们上面是无限循环迭代，定义每一个epochs训练多少步
# 训练模型 一共训练5次每次训练 train_images.shape[0]//64个组件
model.fit(ds_train,epochs=5,steps_per_epoch=steps_per_epochs)

# -*- coding: utf-8 -*-
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2' # 修改警告级别，不显示警告
import tensorflow as tf


# 下载数据集并划分为训练集和测试集
(train_images,train_lables),(test_images,test_labels) = tf.keras.datasets.mnist.load_data()
# 归一化
train_images = train_images/255
test_images = test_images/255
# print(train_images.shape)
# 创建dataset
ds_train_img = tf.data.Dataset.from_tensor_slices(train_images)
# print(ds_train_img)
ds_train_lab = tf.data.Dataset.from_tensor_slices(train_lables)
# print(ds_train_lab)

# 合并(元组形式放入)
ds_train = tf.data.Dataset.zip((ds_train_img,ds_train_lab))

# print(ds_train)
# 取出其中10000个组件进行乱序,无限重复每次输出64张图片
ds_train = ds_train.shuffle(10000).repeat().batch(64)
ds_test = tf.data.Dataset.from_tensor_slices((test_images,test_labels))# 创建test数据集放入一个元组
ds_test = ds_test.batch(64)# 每次输出64个组件
# 建立模型
model = tf.keras.Sequential([
    tf.keras.layers.Flatten(input_shape=(28,28)),
    tf.keras.layers.Dense(128,activation="relu"),
    tf.keras.layers.Dense(10,activation="softmax")
])
# 编译模型
model.compile(optimizer="adam",
              loss="sparse_categorical_crossentropy",
              metrics=["accuracy"])
steps_per_epochs = train_images.shape[0]//64 # 我们上面是无限循环迭代，定义每一个epochs训练多少步
# 训练模型 一共训练5次每次训练 train_images.shape[0]//64个组件/测试数据ds_test 每次测试10000整除64个组件
model.fit(ds_train,
          epochs=5,
          steps_per_epoch=steps_per_epochs,
          validation_data=ds_test,
          validation_steps=10000//64)