텐서플로우 ValueError: Data cardinality is ambiguous 에러

[Chagon Son]

Model: "sequential_10"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
conv2d_9 (Conv2D) (None, 28, 28, 32) 320
_________________________________________________________________
max_pooling2d_9 (MaxPooling2 (None, 14, 14, 32) 0
_________________________________________________________________
flatten_10 (Flatten) (None, 6272) 0
_________________________________________________________________
dense_21 (Dense) (None, 64) 401472
_________________________________________________________________
dense_22 (Dense) (None, 10) 650
=================================================================
Total params: 402,442
Trainable params: 402,442
Non-trainable params: 0
_________________________________________________________________
AAAAAAAAAAAAAAAAAAAAAAAAAA

---------------------------------------------------------------------------

ValueError Traceback (most recent call last)

<ipython-input-11-968f22cba899> in <module>()
46 print("AAAAAAAAAAAAAAAAAAAAAAAAAA")
47
---> 48 model.fit(trainX, trainY, validation_data=(testX, testY), epochs=5)
49
50 score = model.evaluate(testX, testY)

3 frames
/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/engine/data_adapter.py in _check_data_cardinality(data)
1527 label, ", ".join(str(i.shape[0]) for i in nest.flatten(single_data)))
1528 msg += "Make sure all arrays contain the same number of samples."
-> 1529 raise ValueError(msg)
1530
1531

ValueError: Data cardinality is ambiguous:
x sizes: 60000
y sizes: 10000
Make sure all arrays contain the same number of samples.

 

에러발생을 알려주세요.

참고) COLAB에서 프로그램을 실행하였습니다.

 

프로그램 원본입니다.

import tensorflow as tf
import matplotlib.pyplot as plt
import numpy as np

(trainX, trainY), (testX, testY) = tf.keras.datasets.fashion_mnist.load_data()

trainX = trainX / 255.0
trainY = trainY / 255.0

trainX = trainX.reshape ( (trainX.shape[0], 28, 28, 1))
trainY = testX.reshape ( (testX.shape[0], 28, 28, 1))

class_names = ['T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat', 'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot']

#print(trainX[0])
#print(trainX.shape)

#print(testX[0])
#print(testX.shape)

#print(trainY)

#plt.imshow(trainX[1])
#plt.gray()
#plt.colorbar()
#plt.show()

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

])

model.summary()

model.compile(loss="sparse_categorical_crossentropy", optimizer="adam", metrics=['accuracy'])

print("AAAAAAAAAAAAAAAAAAAAAAAAAA")

model.fit(trainX, trainY, validation_data=(testX, testY), epochs=5)

score = model.evaluate(testX, testY)

[codingapple]

x데이터가 60000개 들어왔는데 y데이터는 왜 1만개밖에 없냐는 에러 같습니다 

trainX = trainX.reshape ( (trainX.shape[0], 28, 28, 1))
trainX = testX.reshape ( (testX.shape[0], 28, 28, 1))

이렇게 바꾸면 될듯합니다 

[Chagon Son]

감사합니다.  프로그램이 잘 돌아가네요.

[글쓴이]

글