我是深度学习的新手，因此我正在此网站上进行教程： https://www.datacamp.com/community/tutorials/using-tensorflow-to-compose-music#key-transposition
设法完成自动编码器和VAE。但是本教程的最后一部分是LSTM，我遇到了这个问题，我尝试匹配尺寸，但找不到解决方案。 有人可以用模型摘要解释如何调试维度检查吗？
ValueError: Shapes (None, 1) and (None, 64, 32) are incompatible
这是相关的数据和维度：
#Convert data to numpy array of type float
trainChords = np.array(trainChords, np.float) #shape (3231,32)
trainDurations = np.array(trainDurations, np.float) #shape (3231,32)
targetChords = np.array(targetChords, np.float).reshape(-1,1) #shape (3231,1)
targetDurations = np.array(targetDurations, np.float).reshape(-1,1) #shape (3231,1)
#print(trainChords.shape, trainDurations.shape, targetChords.shape, targetDurations.shape)
# Define number of samples, notes and chords, and durations
nSamples = trainChords.shape[0]
nChords = trainChords.shape[1]
nDurations = trainDurations.shape[1]
# Set the input dimension
inputDim = nChords * sequenceLength
# Set the embedding layer dimension
embedDim = 64
# Define input layers
chordInput = tf.keras.layers.Input(shape = (None,))
durationInput = tf.keras.layers.Input(shape = (None,))
# Define embedding layers
chordEmbedding = tf.keras.layers.Embedding(nChords, embedDim, input_length = sequenceLength)(chordInput)
durationEmbedding = tf.keras.layers.Embedding(nDurations, embedDim, input_length = sequenceLength)(durationInput)
# Merge embedding layers using a concatenation layer
mergeLayer = tf.keras.layers.Concatenate(axis=1)([chordEmbedding, durationEmbedding])
# Define LSTM layer
lstmLayer = tf.keras.layers.LSTM(512, return_sequences=True)(mergeLayer)
# Define dense layer
denseLayer = tf.keras.layers.Dense(256)(lstmLayer)
# Define output layers
chordOutput = tf.keras.layers.Dense(nChords, activation = 'softmax')(denseLayer)
durationOutput = tf.keras.layers.Dense(nDurations, activation = 'softmax')(denseLayer)
# Define model
lstm = tf.keras.Model(inputs = [chordInput, durationInput], outputs = [chordOutput, durationOutput])
# Compile the model
lstm.compile(loss='categorical_crossentropy', optimizer='rmsprop')
# Train the model
lstm.fit([trainChords, trainDurations], [targetChords, targetDurations], epochs=100, batch_size=64)
模型摘要：
Model: "model_2"
__________________________________________________________________________________________________
Layer (type) Output Shape Param # Connected to
==================================================================================================
input_5 (InputLayer) [(None, None)] 0
__________________________________________________________________________________________________
input_6 (InputLayer) [(None, None)] 0
__________________________________________________________________________________________________
embedding_4 (Embedding) (None, None, 64) 2048 input_5[0][0]
__________________________________________________________________________________________________
embedding_5 (Embedding) (None, None, 64) 2048 input_6[0][0]
__________________________________________________________________________________________________
concatenate_2 (Concatenate) (None, None, 64) 0 embedding_4[0][0]
embedding_5[0][0]
__________________________________________________________________________________________________
lstm_2 (LSTM) (None, None, 512) 1181696 concatenate_2[0][0]
__________________________________________________________________________________________________
dense_6 (Dense) (None, None, 256) 131328 lstm_2[0][0]
__________________________________________________________________________________________________
dense_7 (Dense) (None, None, 32) 8224 dense_6[0][0]
__________________________________________________________________________________________________
dense_8 (Dense) (None, None, 32) 8224 dense_6[0][0]
==================================================================================================
Total params: 1,333,568
Trainable params: 1,333,568
Non-trainable params: 0
__________________________________________________________________________________________________