如何将LSTM与Dense连接起来？
如何将LSTM与Dense连接起来？
提问于 2020-01-26 21:48:16
当尝试将LSTM与Dense连接时，它给出一个错误(在尝试训练时)：
input = Input(shape=(x_train.shape[1], None))
X = Embedding(num_words, max_article_len)(input)
X = LSTM(128, return_sequences=True, dropout = 0.5)(X)
X = LSTM(128)(X)
X = Dense(32, activation='softmax')(X)
model = Model(inputs=[input], outputs=[X])
...
>>> ValueError: Error when checking target: expected dense to have shape (32,) but got array with shape (1,)我尝试了不同的连接选项，但错误重复出现：
X, h, c = LSTM(128, return_sequences=False, return_state=True, dropout = 0.5)(X)
X = Dense(32, activation='softmax')(X)
>>> ValueError: Error when checking target: expected dense to have shape (32,) but got array with shape (1,)functional API / Sequential上有任何解决方案选项吗？
数据转换代码：
train = pd.read_csv('train.csv')
articles = train['text']
y_train = train['lang']
num_words = 50000
max_article_len = 20
tokenizer = Tokenizer(num_words=num_words)
tokenizer.fit_on_texts(articles)
sequences = tokenizer.texts_to_sequences(articles)
x_train = pad_sequences(sequences, maxlen=max_article_len, padding='post')
x_train.shape
>>> (18974, 100)
y_train.shape
>>> (18974,)回答 1
Stack Overflow用户
回答已采纳
发布于 2020-01-26 21:54:59
最后一个参数必须设置为False；
X = LSTM(128, return_sequences=True, dropout = 0.5)(X)
X = LSTM(128, return_sequences=False)(X)如果你仍然有问题，那么一定是你的输入形状有问题。
页面原文内容由Stack Overflow提供。腾讯云小微IT领域专用引擎提供翻译支持
原文链接：
https://stackoverflow.com/questions/59918948
复制相关文章
相似问题