TypeError: Value passed to parameter 'input' has DataType int64 not in list of allowed values: float16, bfloat16, float32, float64

Question

我正在尝试运行一个模型并预测来自 mnist kaggle 数据集的测试数据。但是我在尝试 predict.What 时遇到错误，这是原因和解决方案吗？

    model = tf.keras.Sequential([
        tf.keras.layers.Conv2D(32, (3,3), padding='same', activation=tf.nn.relu,
                               input_shape=(28, 28, 1)),
        tf.keras.layers.MaxPooling2D((2, 2), strides=2),
        tf.keras.layers.Conv2D(64, (3,3), padding='same', activation=tf.nn.relu),
        tf.keras.layers.MaxPooling2D((2, 2), strides=2),
        tf.keras.layers.Flatten(input_shape=(28, 28, 1)),
        tf.keras.layers.Dense(128, activation=tf.nn.relu),
        tf.keras.layers.Dense(10,  activation=tf.nn.softmax)
    ])

test = pd.read_csv("test.csv") 
test.head()
CHANNELS = 1
IMAGE_SIZE = 28
IMAGE_WIDTH, IMAGE_HEIGHT = IMAGE_SIZE, IMAGE_SIZE
test = test.values.reshape(-1, IMAGE_WIDTH, IMAGE_HEIGHT, CHANNELS)
predictions = model.predict_classes(test, verbose=1)

TypeError: Value passed to parameter 'input' has DataType int64 not in list of allowed values: float16, bfloat16, float32, float64

Answer 1

正如 TypeError 所说，我认为 test 数据框包含 int 值，因此您必须像这样将类型更改为浮动：

test = test.astype('float')
test = test.values.reshape(-1, IMAGE_WIDTH, IMAGE_HEIGHT, CHANNELS)
predictions = model.predict_classes(test, verbose=1)

TypeError: Value passed to parameter 'input' has DataType int64 not in list of allowed values: float16, bfloat16, float32, float64

TypeError: Value passed to parameter 'input' has DataType int64 not in list of allowed values: float16, bfloat16, float32, float64

mnist

kaggle

python-3.7

tensorflow2.0