无法训练 LSTM 模型（ValueError: Data cardinality is ambiguous:)答案

【问题标题】：Unable to train LSTM model (ValueError: Data cardinality is ambiguous:)无法训练 LSTM 模型（ValueError: Data cardinality is ambiguous:)
【发布时间】：2021-02-10 17:26:01
【问题描述】：

我收到 LSTM 模型的此错误。数据有三列

句子（输入）
值（输出）
标签（输出）我正在使用 tensorflow2.3.0。我已按照建议尝试了 2.0.0，但出现依赖错误。

请帮我解决下面引号中的这个错误

"ValueError: 数据基数不明确：

x 尺寸：720

y 尺寸：89

请提供具有相同第一个维度的数据。”

### Create sequence
vocab_size = 20000
tokenizer = Tokenizer(num_words= vocab_size)
tokenizer.fit_on_texts(df['Sentence'])
sequences = tokenizer.texts_to_sequences(df['Sentence'])
data = pad_sequences(sequences, maxlen=100)

le = LabelEncoder()

df['label'] = le.fit_transform(df['label'])

X = df['Sentence']
y = df[['value','label']]


X_train, y_train, X_test, y_test = train_test_split(X, y, test_size = 0.1)

tokenizer = Tokenizer(num_words=5000)
tokenizer.fit_on_texts(X_train)

X_train = tokenizer.texts_to_sequences(X_train)
X_test = tokenizer.texts_to_sequences(X_test)

vocab_size = len(tokenizer.word_index) + 1

maxlen = 200

X_train = pad_sequences(X_train, padding='post', maxlen=maxlen)
X_test = pad_sequences(X_test, padding='post', maxlen=maxlen)


#print(X_train.shape, X_test.shape, y_train.shape, y_test.shape)

model = Sequential()
model.add(Embedding(vocab_size, 128))
model.add(LSTM(128, dropout=0.2, recurrent_dropout=0.2))
model.add(Flatten())
model.add(Dense(2, activation='sigmoid'))

# try using different optimizers and different optimizer configs
model.compile(loss='binary_crossentropy',
              optimizer='adam',
              metrics=['accuracy'])
print(model.summary())

model.fit(X_train, y_train, epochs=3,batch_size=8, validation_split=0.1)
accr = model.evaluate(X_test, y_test)
print('Test set\n  Loss: {:0.3f}\n  Accuracy: {:0.3f}'.format(accr[0], accr[1]))

【问题讨论】：

标签： python tensorflow lstm multilabel-classification

【解决方案1】：

您的数据有两个输出（值和标签列）。但是你的模型只有一个输出。

此代码有效：

X_train = tf.random.uniform([100, 100], 0, 100, dtype=tf.int32)
y_train = tf.random.uniform([100, 2])
model.fit(X_train, y_train, epochs=3,batch_size=8, validation_split=0.1)

检查 y_train 的形状。它应该是 [batch_size, 2]。

【讨论】：