TF construct eight-part text

• Import: import related packages
• train and test: Process the training and test data sets, including out-of-order operations
• models.Sequential: Building Network Models
• model.compile: Set the optimization algorithm, configure the training method
• model.fit: Set the learning rate during training, the number of iterations, batch_size, and train the model;
• model.summary: print out network structure and parameter statistics

Simple implementation

# 1. importimport tensorflow as tf# 2. train and testw = tf.Variable(tf.constant(5, dtype=tf.float32))# 3. model.sequential# 4. model.compile and 5. model.fitlr = 0.2epochs = 40for epoch in range(epochs):with tf.GradientTape() as tape:loss = tf.square(w + 1) # define lossfunctiongrads = tape.gradient(loss, w) # Find the gradient of loss to ww.assign_sub(lr * grads)print("After %s epoch, w is %f, loss is%f" % (epoch,w.numpy(), loss)) # 6. model.summary()