全部代码如下：（红色部分为与笔记二不同之处）
#1.Import the neccessary libraries needed import numpy as np import tensorflow as tf import matplotlib from matplotlib import pyplot as plt ######################################################################## #2.Set default parameters for plots matplotlib.rcParams['font.size'] = 20 matplotlib.rcParams['figure.titlesize'] = 20 matplotlib.rcParams['figure.figsize'] = [9, 7] matplotlib.rcParams['font.family'] = ['STKaiTi'] matplotlib.rcParams['axes.unicode_minus']=False ######################################################################## #3.Initialize Parameters #Initialize learning rate lr = 1e-2 #----------------------changed #Initialize batch size batchsz = 512 #Initialize loss and accurate array losses = [] accs = [] #----------------------changed #Initialize the weights layers and the bias layers w1=tf.Variable(tf.random.truncated_normal([784,256],stddev=0.1)) b1=tf.Variable(tf.zeros([256])) w2=tf.Variable(tf.random.truncated_normal([256,128],stddev=0.1)) b2=tf.Variable(tf.zeros([128])) w3=tf.Variable(tf.random.truncated_normal([128,10],stddev=0.1)) b3=tf.Variable(tf.zeros([10])) ######################################################################## #4.Define preprocess function #----------------------changed def preprocess(x,y): x=tf.cast(x,dtype=tf.float32)/255. x=tf.reshape(x,[-1,28*28]) y=tf.cast(y,dtype=tf.int32) #one_hot接受的输入为int32,输出为float32 y=tf.one_hot(y,depth=10) return x,y ######################################################################## #5.Import the minist dataset offline (x_train,y_train),(x_test,y_test)=tf.keras.datasets.mnist.load_data(path=r'F:\learning\machineLearning\TensorFlow2_deeplearning\forward_progression\mnist.npz') train_db=tf.data.Dataset.from_tensor_slices((x_train,y_train)) train_db=train_db.shuffle(10000) #-----------------------changed train_db=train_db.batch(batchsz) train_db=train_db.map(preprocess) #Control the epoch times train_db=train_db.repeat(20) test_db=tf.data.Dataset.from_tensor_slices((x_test,y_test)) test_db=test_db.shuffle(1000).batch(batchsz).map(preprocess) ######################################################################## #The main function def main(): for step,(x,y) in enumerate(train_db):#Or for x,y in train_db: with tf.GradientTape() as tape: # tf.Variable # layer1 h1 = x@w1 + b1 h1 = tf.nn.relu(h1) # layer2 h2 = h1@w2 + b2 h2 = tf.nn.relu(h2) # output out = h2@w3 + b3 # compute loss loss = tf.square(y-out) # mean: scalar loss = tf.reduce_mean(loss) # compute gradients grads = tape.gradient(loss, [w1, b1, w2, b2, w3, b3]) #Update the weights and the bias #-----------------------changed for p, g in zip([w1, b1, w2, b2, w3, b3], grads): p.assign_sub(lr * g) if step % 80 == 0: print(step, 'loss:', float(loss)) losses.append(float(loss)) if step % 80 == 0: #-----------------------changed total, total_correct = 0., 0 for x,y in test_db: # layer1 h1 = x@w1 + b1 h1 = tf.nn.relu(h1) # layer2 h2 = h1@w2 + b2 h2 = tf.nn.relu(h2) # output out = h2@w3 + b3 pred=tf.argmax(out,axis=1) y=tf.argmax(y,axis=1) correct=tf.equal(pred,y) total_correct+=tf.reduce_sum(tf.cast(correct,dtype=tf.int32)).numpy() total+=x.shape[0] print(step,'Evaluate ACC:',total_correct/total) accs.append(total_correct/total) plt.figure() x = [i*80 for i in range(len(losses))] plt.plot(x, losses, color='C0', marker='s', label='训练') plt.ylabel('MSE') plt.xlabel('Step') plt.legend() plt.figure() plt.plot(x, accs, color='C1', marker='s', label='测试') plt.ylabel('准确率') plt.xlabel('Step') plt.legend() plt.show() if __name__ == '__main__': main()
其中learning rate在此处改为了1e-2，经测试若为1e-3则accurate rate会增长较慢，在20epoch下最终会达到30~40%,而1e-2则会接近80%
并且通过.map(preprocess)方法预处理了train_db，包括将图片数据标准化到（0-1），reshape到[-1,28*28]，将标签数据做one-hot处理，深度为10；通过train_db=train_db.repeat(20)代替了for epoch in range(20)；用
for p, g in zip([w1, b1, w2, b2, w3, b3], grads): p.assign_sub(lr * g)
代替了
w1.assign_sub(lr * grads[0]) b1.assign_sub(lr * grads[1]) w2.assign_sub(lr * grads[2]) b2.assign_sub(lr * grads[3]) w3.assign_sub(lr * grads[4]) b3.assign_sub(lr * grads[5])