How to set Conv2D parameters correctly?

Question

I'm trying to build CNN model on MNIST dataset but there is an error appear and I can not solve it. This is my code

import tensorflow as tf
from tensorflow.keras.datasets import mnist
from keras.utils import to_categorical
from keras.models import Sequential
from keras.layers import Dense, Flatten, Conv2D
from numpy import expand_dims
(x_train, y_train), (x_test, y_test) = mnist.load_data() 
x_train = expand_dims(x_train, 3)
x_test = expand_dims(x_test, 3)
y_train = to_categorical(y_train) 
y_test = to_categorical(y_test)
model = Sequential()
model.add(Conv2D(64, kernel_size=3, activation="relu", input_shape=(28, 28, 1)))
model.add(Conv2D(64, kernel_size=3, activation="relu"))
model.add(Flatten())
model.add(Dense(10, activation="softmax"))
model.compile(optimizer="adam", loss="categorical_crossentropy", metrics=["accuracy"])
model.fit(x_train, y_train, validation_data=(x_test, y_test), epochs=3)
model.save("model")
new_model = tf.keras.models.load_model("model")
predictions = new_model.predict([x_test])
print(predictions[10])

I got this error

TypeError: Value passed to parameter 'input' has DataType uint8 not in list of allowed values: float16, bfloat16, float32, float64

Any help?

Answer 1

You are getting this error because your input is uint8 type while your network has float values. You simply have to cast your inputs to float data type. Here, change your data part to this:

import numpy as np
(x_train, y_train), (x_test, y_test) = mnist.load_data()
x_train = expand_dims(x_train, -1)
x_test = expand_dims(x_test, -1)
x_train = x_train.astype(np.float32)
x_test = x_test.astype(np.float32)

Weirdly, I didn't the error you got but I should have and this is the solution.