InvalidArgumentError:接收到的标签值为3，超出了[0，3] tensorflow情感分析的有效范围

Question

所以我试着用tensorflow学习情感分析，我的数据集包含3 y_labels，即1 == 'negative,2=='neutral,3=='positive'。这是我的密码

tokenizer = Tokenizer(num_words=vocab_size,oov_token=oov_tok)
tokenizer.fit_on_texts(X)

word_index =tokenizer.word_index

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

training_padding = pad_sequences(training_sequence,maxlen=max_length,padding=padding_type)
test_padding = pad_sequences(training_sequence,maxlen=max_length,padding=padding_type)

training_padded = np.array(training_padding)
training_label = np.array(y_train)

test_padded = np.array(test_padding)
test_label = np.array(y_test)

model = tf.keras.Sequential([
    tf.keras.layers.Embedding(vocab_size,embedding_dim,input_length=max_length ),
    tf.keras.layers.GlobalMaxPool1D(),
    tf.keras.layers.Dense(20,activation='relu'),
    tf.keras.layers.Dense(40,activation='relu'),
    tf.keras.layers.Dense(3,activation='relu')

])
model.compile(loss='sparse_categorical_crossentropy',optimizer='adam',metrics=['accuracy'])
num_epoch = 200
history = model.fit(training_padded,training_label,epochs=num_epoch)

但是我有一个错误，我剪了一点，因为它太长了。

invalidArgumentError:  Received a label value of 3 which is outside the valid range of [0, 3).  Label values: 2 2 1 3 3 1 1 1 3 2 3 1 3 3 2 1 1 1 1 3 3 2 2 2 1 3 3 3 1 3 2 2
     [[node sparse_categorical_crossentropy/SparseSoftmaxCrossEntropyWithLogits/SparseSoftmaxCrossEntropyWithLogits
 (defined at c:\users\pc\pycharmprojects\deeplearning_sentiment_analysis\venv\lib\site-packages\keras\backend.py:5113)
]] [Op:__inference_train_function_3960]

Errors may have originated from an input operation.

在我看来，我所做的已经是正确的，谁能告诉我出了什么问题吗？据我所知，我们可以在sequental()中构建分类，但是所有的资源都只能用于二进制分类[0,1]，我们可以构建多个类吗？

这里是我的示例dataframe

​

​

Answer 1

我从Luke Borowy那里得到了这样的结论:在多类或多标签分类中，所有标签都应该从0开始。