tensorflow2代码积累

深度学习代码

保存与加载npz文件

数组形式

    acc = history.history['accuracy']
    val_acc = history.history['val_accuracy']

    loss = history.history['loss']
    val_loss = history.history['val_loss']

    np.savez('plot_npz/flower',acc=acc,val_acc=val_acc,loss=loss,val_loss=val_loss)

    data = np.load('plot_npz/flower.npz')
    # print(data.files)
    # print(data['acc'])
    acc = data['acc']
    val_acc = data['val_acc']

    loss = data['loss']
    val_loss = data['val_loss']

plot绘制一行两列子图精度和损失

    plt.figure(figsize=(16, 8))
    plt.subplot(1, 2, 1)
    plt.plot(epochs_range, acc, label='Training Accuracy')
    plt.plot(epochs_range, val_acc, label='Validation Accuracy')
    plt.legend(loc='lower right')
    plt.title('Training and Validation Accuracy')

    plt.subplot(1, 2, 2)
    plt.plot(epochs_range, loss, label='Training Loss')
    plt.plot(epochs_range, val_loss, label='Validation Loss')
    plt.legend(loc='upper right')
    plt.title('Training and Validation Loss')
    plt.show()

函数式模型

    if mode == 'functional':
        inputs = tf.keras.Input(shape=(28, 28, 1))
        x = tf.keras.layers.Flatten()(inputs)
        x = tf.keras.layers.Dense(units=100, activation=tf.nn.relu)(x)
        x = tf.keras.layers.Dense(units=10)(x)
        outputs = tf.keras.layers.Softmax()(x)
        model = tf.keras.Model(inputs=inputs, outputs=outputs)

序列式模型

model = Sequential([
    data_augmentation,
    layers.experimental.preprocessing.Rescaling(1. / 255),
    layers.Conv2D(16, 3, padding='same', activation='relu'),
    layers.MaxPooling2D(),
    layers.Conv2D(32, 3, padding='same', activation='relu'),
    layers.MaxPooling2D(),
    layers.Conv2D(64, 3, padding='same', activation='relu'),
    layers.MaxPooling2D(),
    layers.Dropout(0.2),
    layers.Flatten(),
    layers.Dense(128, activation='relu'),
    layers.Dense(num_classes)
])

data_augmentation = keras.Sequential(
    [
        layers.experimental.preprocessing.RandomFlip("horizontal",
                                                     input_shape=(img_height,
                                                                  img_width,
                                                                  3)),
        layers.experimental.preprocessing.RandomRotation(0.1),
        layers.experimental.preprocessing.RandomZoom(0.1),
    ]
)

绘制数据增强图片

    plt.figure(figsize=(10, 10))
    for images, _ in train_ds.take(1):
      for i in range(9):
        augmented_images = data_augmentation(images)
        ax = plt.subplot(3, 3, i + 1)
        plt.imshow(augmented_images[2].numpy().astype("uint8"))
        plt.axis("off")
    plt.show()

调用预处理层放缩图片到（0-255）->(0-1)

    # 第一种方法单独使用layer进行缩放，也可以放到模型里。我们采用第二种方法
    normalization_layer = layers.experimental.preprocessing.Rescaling(1./255)
    normalized_ds = train_ds.map(lambda x, y: (normalization_layer(x), y))
    image_batch, labels_batch = next(iter(normalized_ds))
    first_image = image_batch[0]
    # Notice the pixels values are now in `[0,1]`.
    # 经过标准缩放后 像素变为（0,1）之间
    print(np.min(first_image), np.max(first_image))
    # 第二种方法

函数式模型keras

def flower_model():
    inputs = tf.keras.Input(shape=(img_height, img_width, 3))
    x = layers.experimental.preprocessing.RandomFlip("horizontal")(inputs)
    x = layers.experimental.preprocessing.RandomRotation(0.1)(x)
    x = layers.experimental.preprocessing.RandomZoom(0.1)(x)
    x = layers.experimental.preprocessing.Rescaling(1. / 255)(x)
    x = layers.Conv2D(16, 3, padding='same', activation='relu')(x)
    x = layers.MaxPooling2D()(x)
    x = layers.Conv2D(32, 3, padding='same', activation='relu')(x)
    x = layers.MaxPooling2D()(x)
    x = layers.Conv2D(64, 3, padding='same', activation='relu')(x)
    x = layers.MaxPooling2D()(x)
    x = layers.Dropout(0.2)(x)
    x = layers.Flatten()(x)
    x = layers.Dense(128, activation='relu')(x)
    outputs = layers.Dense(num_classes)(x)
    model = tf.keras.Model(inputs=inputs, outputs=outputs)
    return model

加载测试数据集

    test_ds = tf.keras.preprocessing.image_dataset_from_directory(
        test_data_dir,
        validation_split=None,
        subset=None,
        seed=123,
        image_size=(img_height, img_width),
        batch_size=batch_size)

验证测试集evaluate

注意需要compile
而且可以加载checkpoint

def evaluate():
    global test_ds
    AUTOTUNE = tf.data.experimental.AUTOTUNE
    test_ds = test_ds.cache().prefetch(buffer_size=AUTOTUNE)
    model = flower_model()
    model.compile(optimizer='adam', loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
                  metrics=['accuracy'])
    checkpoint = tf.train.Checkpoint(myAwesomeModel=model)
    checkpoint.restore(tf.train.latest_checkpoint(checkpoint_path))
    loss, accuracy = model.evaluate(test_ds)
    print('Test accuracy :', accuracy)