module 'keras.utils' has no attribute 'to_categorical'

[namespacegit]

from keras import utils as np_utils
y_train = np_utils.to_categorical(y_train,num_classes=3)
y_test = np_utils.to_categorical(y_test,num_classes=3)

为什么会出现这个错误？

[Edward1Chou]

keras版本不同的原因，我在调试的时候也遇到这个问题，看不同版本的api解决把，是个坑...

[ResolveWang]

我用的Keras==2.1.3好像没这个问题，其他也有一些API不兼容，需要更新一下

[jeness]

Keras==2.1.5
gensim==3.4.0
jieba==0.38
我用的版本是这样，
然后增加from keras import utils as np_utils
120行附近，修改为：

 y_train = np_utils.to_categorical(y_train,num_classes=3)
 y_test = np_utils.to_categorical(y_test,num_classes=3)

107行附近，修改为：

model.train(combined,total_examples=model.corpus_count, epochs=model.iter)

72行附近，修改为：

gensim_dict.doc2bow(model.wv.vocab.keys(),
                            allow_update=True)

然后建立一个名字是lstm_data_test的directory
就能跑了

[skoundin]

Hi, I am getting the same issue, I was using cifar code from the location -
https://github.com/keras-team/keras/blob/master/examples/cifar10_cnn.py

Error I get is - module 'keras.utils' has no attribute 'to_categorical'.

Let me know if anyone was able to resolve this.
Thanks.

[DerekChia]

@skoundin I have no problem running it. What version of Keras and TF are you on? Mind sharing your environment text?

[yuhaoroy]

model.fit(x_train, y_train, batch_size=batch_size, epochs=n_epoch,verbose=1) ^ SyntaxError: invalid syntax
请问这里是怎么回事？ 没找到哪里有语法错误

[Edward-Joker]

Keras==2.1.5
gensim==3.4.0
jieba==0.38
我用的版本是这样，
然后增加from keras import utils as np_utils
120行附近，修改为：

 y_train = np_utils.to_categorical(y_train,num_classes=3)
 y_test = np_utils.to_categorical(y_test,num_classes=3)

107行附近，修改为：

model.train(combined,total_examples=model.corpus_count, epochs=model.iter)

72行附近，修改为：

gensim_dict.doc2bow(model.wv.vocab.keys(),
                            allow_update=True)

然后建立一个名字是lstm_data_test的directory
就能跑了

[Edward-Joker]

我试过了按照这种方法确实能用,感谢！

[Jhouny]

I tried the solution @Edward-Joker posted and it didn't work for me.

I'm running Tensorflow version 2.5.0. By trial and error I found that keras.utils.np_utils works. I guess they moved it into np_utils in some update, so with that .to_categorical works just fine.

Hope it helps someone!

[rituvermaCS]

TF version- 2.5.0
keras version -2.5.0

Resolving this issue, first import to_categorical()

from tensorflow.keras.utils import to_categorical

then write like

keras.utils.to_categorical()

It will work.

[13223486896]

I tried the solution @Edward-Joker posted and it didn't work for me.

I'm running Tensorflow version 2.5.0. By trial and error I found that keras.utils.np_utils works. I guess they moved it into np_utils in some update, so with that .to_categorical works just fine.

Hope it helps someone!

it is ok

[tennisee]

from keras import utils as np_utils y_train = np_utils.to_categorical(y_train,num_classes=3) y_test = np_utils.to_categorical(y_test,num_classes=3)

为什么会出现这个错误？

keras version -2.6.0

from keras.utils import np_utils
y_train = np_utils.to_categorical(y_train, num_classes)
y_test = np_utils.to_categorical(y_test, num_classes)

[sebastianTech]

I am not sure if this was solved, however, I encountered the same error and this solved it for me.

First import
from keras.utils import np_utils

Then copy the below code and you should be fine

y_train = keras.utils.np_utils.to_categorical(y_train,num_classes=3)
y_test =keras.utils.np_utils.to_categorical(y_test,num_classes=3)