ImportError: cannot import name ‘pad_sequences‘ from ‘keras.preprocessing.sequence

运行步骤：

报错信息：

---------------------------------------------------------------------------
ImportError                               Traceback (most recent call last)
~\AppData\Local\Temp\ipykernel_20076\4123045159.py in <module>
      1 # 5) 划分数据集
----> 2 from keras.preprocessing.sequence import pad_sequences
      3 maxlen = 128
      4 vec_data = list(sequence.pad_sequences(data['vec'], maxlen=maxlen))
      5 x_train, x_test, y_train, y_test = train_test_split(vec_data, data['label'], test_size=0.2, random_state=0)

ImportError: cannot import name 'pad_sequences' from 'keras.preprocessing.sequence' (D:\Anaconda\lib\site-packages\keras\preprocessing\sequence.py)

报错原因： keras.preprocessing.sequence 模块中没有 pad_sequences 属性，在TensorFlow 2.x中，pad_sequences函数已经被整合到tensorflow.keras.preprocessing.sequence模块中。根据您提供的错误信息，您尝试从keras.preprocessing.sequence导入pad_sequences，但这个模块在TensorFlow 2.x中并不存在。您应该直接从tensorflow.keras.preprocessing.sequence导入pad_sequences。

解决办法：确保您的环境中安装了 TensorFlow 2.x，并且您的代码中没有其他导入 pad_sequences 的语句，以免发生冲突。

修改代码:

import tensorflow as tf
from tensorflow.keras.preprocessing.sequence import pad_sequences

maxlen = 128
vec_data = pad_sequences(data['vec'], maxlen=maxlen)
x_train, x_test, y_train, y_test = train_test_split(vec_data, data['label'], test_size=0.2, random_state=0)

问题解决!