[深度之眼]TensorFlow2.0项目班-模型训练之Callbacks

在上次建立的模型基础上，加入callbacks机制
分别实现了：模型参数保存、早停和动态学习率衰减

import tensorflow as tf

fashion = tf.keras.datasets.fashion_mnist
(x_train, y_train),(x_test, y_test) = fashion.load_data()
x_train, x_test = x_train / 255.0, x_test / 255.0

#函数模型
inputs = tf.keras.Input(shape=(28,28))
x = tf.keras.layers.Flatten()(inputs)
x = tf.keras.layers.Dense(128,activation='relu')(x)
predictions = tf.keras.layers.Dense(10,activation='softmax')(x)
model = tf.keras.Model(inputs,predictions)

model.compile(optimizer='adam',
              loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=False),
              metrics=['sparse_categorical_accuracy'])

callbacks = [tf.keras.callbacks.ModelCheckpoint( # 保存训练权重
                    filepath='weights.BP.hdf5', 
                    save_best_only = True ,      # 仅保存最优权重
                    verbose=1,
                    mode='val_acc',
                    save_weights_only = True),
    
             tf.keras.callbacks.EarlyStopping(   # 满足条件提前停止训练
                    monitor='val_loss',          # 监控的指标
                    min_delta=1e-2,              # 低于则结束
                    patience=3,                  # 超过设置轮数loss不减则结束
                    verbose=1),
             
             tf.keras.callbacks.ReduceLROnPlateau(   # 动态学习率衰减
                    monitor='val_acc',   
                    verbose=1,
                    model='max',
                    factor=0.8,                  # 衰减因子
                    patience=3)]

model.fit(x_train, y_train, batch_size=32, epochs=20, validation_data=(x_test, y_test), callbacks=callbacks,shuffle=True ,validation_freq=1)
model.summary()

运行结果如下：

并没有完成预先设置的epoch，在满足条件的情况下提前结束
在今后预测过程中，不再需要重新训练模型，直接调用保存好的最优模型参数：

model_best = model()
model_best.load_weights("weights.BP.hdf5")
model_best.predict(待预测数据)