tf2 + keras 使用卷积神经网络进行 fashion_mnist 数据集的识别

最新推荐文章于 2022-12-18 16:01:03 发布

怅然一梦，烟火星辰

最新推荐文章于 2022-12-18 16:01:03 发布

阅读量228

点赞数 1

文章标签：神经网络 keras tensorflow

本文链接：https://blog.csdn.net/weixin_46016713/article/details/120800109

版权

import tensorflow as tf
from tensorflow.keras import layers

# gpu = tf.config.list_physical_devices('GPU')
# 设置显卡显存按需求分配和增长
# tf.config.experimental.set_memory_growth(gpu[0], True)
# 设置显卡显存可用限度
# tf.config.experimental.set_virtual_device_configuration(
#     gpu[0],
#     [tf.config.experimental.VirtualDeviceConfiguration(memory_limit=4096)])


data = tf.keras.datasets.fashion_mnist


class myCallback(tf.keras.callbacks.Callback):
    def on_epoch_end(self, epoch, logs=None):
        if logs.get('accuracy') > 0.95:
            print("\n 已经到达 95% 的训练精度！")
            self.model.stop_training = True


callback = myCallback()

(training_images, training_labels), (test_images, test_labels) = data.load_data()

training_images = training_images.reshape(60000, 28, 28, 1)
training_images = training_images / 255.0
test_images = test_images.reshape(10000, 28, 28, 1)
test_images = test_images / 255.0

model = tf.keras.models.Sequential([
    layers.Conv2D(64, (3, 3), activation='relu', input_shape=(28, 28, 1)),
    layers.MaxPooling2D(2, 2),
    layers.Conv2D(64, (3, 3), activation='relu'),
    layers.MaxPooling2D(2, 2),
    layers.Flatten(),
    layers.Dense(128, activation=tf.nn.relu),
    layers.Dense(10, activation=tf.nn.softmax)
])
model.summary()
model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])
model.fit(training_images, training_labels, epochs=10, callbacks=callback)
model.evaluate(test_images, test_labels)