Unrecognized keyword arguments passed to Embedding about tensorflow

[theraghavjuneja]

Instead of input length:10 I used input_shape=(10,) to reproduce the same effect. It worked for me. But when u read the documentation u won't find anything like input_shape you would only find input_length.

[theraghavjuneja]

Ok I tried converting the input_length to input_shape of(10,) formulating what it was trying to say in the documentation. (On searching I found this). But shouldn't that be updated in documentation as well?

[MujahidMalik819061]

ValueError: Unrecognized keyword arguments passed to Embedding: {'input_length': 10}
how to solve this error

[Venkat6871]

Hi @theraghavjuneja ,
Sorry for the delay, I tried to run your code on Colab using TF v2.16.1, nightly and faced the same issue. Please find the gist here for reference.

Thank you!

[Venkat6871]

Hi @theraghavjuneja ,
The error occurs because the Embedding layer in TensorFlow 2.x does not have an input_length argument. This argument was removed in version 2.0.0 of TensorFlow.

To fix this error, you need to remove the input_length argument from the Embedding layer. The input_length argument is used to specify the maximum length of the input sequences. However, in TensorFlow 2.x, the Embedding layer automatically infers the input length from the input data. Please find the gist here for reference.

[TomBerton]

Instead of input length:10 I used input_shape=(10,) to reproduce the same effect. It worked for me. But when u read the documentation u won't find anything like input_shape you would only find input_length.

I created an LSTM-based sequential model for text generation. I updated my keras to 3.1.1 and had the same issue. Using input_shape=(25,) worked for my model. Thanks.

[akramIOT]

Hi @TomBerton : I am hitting this same issue with TF==2.16 version for my CNN Word2Vector Embeddings

Define the CNN model

model = Sequential()
model.add(Embedding(vocab_size, 100, weights=[embedding_matrix],input_length=max_length, trainable=False))
model.add(Conv1D(128, 5, activation='relu'))

(DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 %
(DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 % pip list | grep Tensroflow
(DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 %
(DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 % pip list | grep Tensorflow
]% (DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 % ]
zsh: command not found: ]
(DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 % pip list | grep tensorflow
tensorflow 2.16.1
tensorflow-datasets 4.9.4
tensorflow-estimator 2.12.0
tensorflow-io-gcs-filesystem 0.36.0
tensorflow-metadata 1.14.0
(DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 %
(DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 % pip list | grep keras
keras 3.1.1
tf_keras 2.16.0
(DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 %

[akramIOT]

Hi @TomBerton : I am hitting this same issue with TF==2.16 version for my CNN Word2Vector Embeddings

Define the CNN model

model = Sequential() model.add(Embedding(vocab_size, 100, weights=[embedding_matrix],input_length=max_length, trainable=False)) model.add(Conv1D(128, 5, activation='relu'))

(DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 % (DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 % pip list | grep Tensroflow (DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 % (DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 % pip list | grep Tensorflow ]% (DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 % ] zsh: command not found: ] (DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 % pip list | grep tensorflow tensorflow 2.16.1 tensorflow-datasets 4.9.4 tensorflow-estimator 2.12.0 tensorflow-io-gcs-filesystem 0.36.0 tensorflow-metadata 1.14.0 (DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 % (DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 % pip list | grep keras keras 3.1.1 tf_keras 2.16.0 (DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 %

(DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 % python CNN_WordVec.py
Traceback (most recent call last):
File "/Users/akram_personal/AKRAM_CODE_FOLDER/DS_EXERCISES/DS_2024/CNN_WordVec.py", line 114, in
model.add(Embedding(vocab_size, 100, weights=[embedding_matrix], trainable=False))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/opt/anaconda3/envs/DS/lib/python3.11/site-packages/keras/src/layers/core/embedding.py", line 81, in init
super().init(**kwargs)
File "/opt/anaconda3/envs/DS/lib/python3.11/site-packages/keras/src/layers/layer.py", line 264, in init
raise ValueError(
ValueError: Unrecognized keyword arguments passed to Embedding: {'weights': [array([[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.],
...,
[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.]])]}
(DS) akram_personal@AKRAMs-MacBook-Pro DS_2024 %

[google-ml-butler]

Are you satisfied with the resolution of your issue?
Yes
No