如何使用Keras培训的嵌入式层？

Question

My model is:

        model = Sequential()
        model.add(Embedding(input_dim=vocab_size,
                            output_dim=1024, input_length=self.SEQ_LENGTH))

        model.add(LSTM(vocab_size))

        model.add(Dropout(rate=0.5))
        model.add(Dense(vocab_size - 1, activation='softmax'))

And I have it trained. But now during inference time, how can I use that embedding?

Answer 1

Your question is solved here . As skeleton you can use this code:

from tensorflow.python.keras.preprocessing.text import Tokenizer

tokenizer_obj = Tokenizer()
tokenizer_obj.fit_on_texts(your_dataset) 

...

max_length = max_number_words
X_test_tokens = tokenizer_obj.texts_to_sequences(X_test)
X_test_pad = pad_sequences(X_test_tokens, maxlen=max_length, padding='post')

score, acc = model.evaluate(X_test_pad, y_test, batch_size=128)