Modelcheckpoint auto save model

-- coding: utf-8 --

import tensorflow as tf
import json
import tensorflow.keras.models
from tensorflow.keras.callbacks import *
# This is a separate package
import tensorflow.keras
mnist = tf.keras.datasets.mnist
(x_train, y_train), (x_test, y_test) = mnist.load_data()
x_train, x_test = x_train / 255.0, x_test / 255.0

model = tf.keras.models.Sequential([
tf.keras.layers.Flatten(),
tf.keras.layers.Dense(128, activation=‘relu’),
tf.keras.layers.Dense(10, activation=‘softmax’)
])

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

#Callbacks Used to specify in each epoch or batch What is the specific operation to be performed at the beginning and the end
‘’’
Parameters monitor You can set {‘val_accuracy’,‘val_loss’,‘accuracy’,‘loss’}.
If monitoring is set {‘val_accuracy’,‘accuracy’}, Then the accuracy of the model is greater than the maximum {‘val_accuracy’,‘accuracy’} The model will be saved when ;
If monitoring is set {‘val_loss’,‘loss’}, So the model loss Less than the minimum {‘val_loss’,‘loss’} The model will be saved when .
‘’’

ModelCheckpoint Used to automatically save the model

filepath You can set the model storage location and model information ,epoch Indicates the number of training cycles ,val_accuracy Indicates the exact value of the validation set

monitor Optional {‘val_accuracy’,‘val_loss’,‘accuracy’,‘loss’}, commonly ’val_accuracy’ Used a lot

verbose=1 It means printing information when saving the model

save_best_only=True Means save only >best_val_accuracy Model of

CSVLogger It's also callbacks, For generating model training log

‘’’
loss： Training set loss value

accuracy: Training set accuracy

val_loss: Test set loss value

val_accruacy: Test set accuracy
‘’’
# The saved variable name should be consistent with the variable name during training
callbacks = [
ModelCheckpoint(filepath=‘C:/Users/25360/Desktop/model/’+’{epoch:02d}-{sparse_categorical_accuracy:.4f}.h5’,
monitor=‘val_sparse_categorical_accuracy’,
verbose=1,
save_best_only=True)
]
model.fit(x_train, y_train, batch_size=32, epochs=5, validation_data=(x_test, y_test),
validation_freq=1,callbacks=callbacks)
model.summary()